Research report

Food parenting practices and child dietary behavior Prospectiverelations and the moderating role of general parenting

Ester FC Sleddens a Stef PJ Kremers a Annette Stafleu b Pieter C Dagnelie cNanne K De Vries d Carel Thijs c

a Department of Health Promotion NUTRIM School for Nutrition Toxicology and Metabolism Maastricht University Medical Center+ PO Box 616 6200MD Maastricht The Netherlandsb TNO PO Box 360 3700 AJ Zeist The Netherlandsc Department of Epidemiology CAPHRI School of Public Health and Primary Care Maastricht University Medical Center+ PO Box 616 6200 MDMaastricht The Netherlandsd Department of Health Promotion CAPHRI School of Public Health and Primary Care NUTRIM School for Nutrition Toxicology and Metabolism MaastrichtUniversity Medical Center+ PO Box 616 6200 MD Maastricht The Netherlands

A R T I C L E I N F O

Article historyReceived 27 March 2013Received in revised form 1 April 2014Accepted 3 April 2014Available online 13 April 2014

KeywordsGeneral parentingFood parenting practicesChild dietary behavior

A B S T R A C T

Research on parenting practices has focused on individual behaviors while largely failing to consider thecontext of their use ie general parenting We examined the extent to which food parenting practicespredict childrenrsquos dietary behavior (classified as unhealthy snacking sugar-sweetened beverage andhealthy water and fruit intake) Furthermore we tested the moderating role of general parenting on thisrelationship Within the KOALA Birth Cohort Study in the Netherlands questionnaire data were collect-ed at 6 and 8 years (N = 1654) Correlations were computed to assess the association between food par-enting practices and general parenting (ie nurturance behavioral control structure coercive controland overprotection) Linear regression models were fitted to assess whether food parenting practices predictdietary behavior Instrumental and emotional feeding and pressure to eat were found to have associa-tions with undesirable child dietary behavior (increased unhealthy intakedecreased healthy intake) whereasassociations were in the desirable direction for covert control encouragement and restriction Modera-tion analyses were performed by evaluating interactions with general parenting The associations of en-couragement and covert control with desirable child dietary behaviors were found to be stronger for childrenwho were reared in a positive parenting context Future research should assess the influence of contex-tual parenting factors moderating the relationships between food parenting and child dietary behavioras the basis for the development of more effective family-based interventions

copy 2014 Elsevier Ltd All rights reserved

Introduction

It is well-known that childhood overweight has a tendency topersist into adulthood (Singh Mulder Twisk Van Mechelen ampChinapaw 2008) and is associated with multiple adverse health out-

comes both in the short term (Daniels 2009) and longer term (Reillyamp Kelly 2011)

While it is generally accepted that parenting has importantinfluences on childrenrsquos dietary intake and ultimately on weightstatus the exact conditions of its influence are still not well definedIn the feeding domain previous reviews showed the relationshipsof parentsrsquo use of food parenting practices (ie content-specificacts of parenting (Darling amp Steinberg 1993)) on child eatingand weight status (eg Clark Goyder Bissell Blank amp Peters2007 Faith Scanlon Birch Francis amp Sherry 2004 Ventura ampBirch 2008 Wardle amp Carnell 2007) Most of the included studiesfocus on highly controlling food parenting practices (restricting thetype and amount of food and using food as a reward) failing toassess other practices such as encouragement and guidanceThese (mainly cross-sectional) studies have yielded contradictoryresults and have in common that they did not assess the largercontext in which these behaviors occurred (ie general parentingstyles)

Abbreviations CFQ Child Feeding Questionnaire CGPQ Comprehensive GeneralParenting Questionnaire FFQ Food Frequency Questionnaire PFSQ Parental FeedingStyle Questionnaire

Acknowledgments We are grateful for the help of the KOALA participants for par-

ticipating in this study To our knowledge no conflict of interest financial or otherexists ES SK and CT designed the research ES conducted research wrote the paperand had primary responsibility for final content All authors read commented onand approved the final manuscript The authors declare that they have no compet-ing interests This research was funded by the Netherlands Heart Foundation (projectnumber 2008B112)

Corresponding authorE-mail address estersleddensmaastrichtuniversitynl (EFC Sleddens)

httpdxdoiorg101016jappet2014040040195-6663copy 2014 Elsevier Ltd All rights reserved

Appetite 79 (2014) 42ndash50

Contents lists available at ScienceDirect

Appetite

journal homepage wwwelseviercom locate appet

General parenting is independent of specific socialization contentit has commonly been defined as the approach parents use to raisetheir child and is a function of parentrsquos attitudes beliefs and be-haviors creating a family emotional climate (Darling amp Steinberg1993) In a recent review Sleddens Gerards Thijs De Vries andKremers (2011) reported that authoritative forms of parenting (char-acterized by parental warmth and guidance) were associated withmore positive weight-related outcomes than more permissive or co-ercive forms of parenting This was confirmed in some recent cor-relational studies (Fuemmeler et al 2012 Johnson WelkSaint-Maurice amp Ihmels 2012 Rodenburg Kremers Oenema amp Vande Mheen 2011 2013 Topham et al 2011) However these effectsof general parenting on weight-related behaviors were generally in-direct and weak compared with the effects of more proximalbehavior-specific parenting practices (Cislak Safron Pratt Gasparamp Luszczynska 2012)

Studies assessing both general parenting and food parenting prac-tices are scarce (eg Blissett amp Haycraft 2008 Blissett Meyer ampHaycraft 2011 Hennessy Hughes Goldberg Hyatt amp Economos2010 Hubbs-Tait Kennedy Page Topham amp Harrist 2008Musher-Eizenman amp Holub 2006 Rodenburg et al 2013 TaylorWilson Slater amp Mohr 2011 Van der Horst et al 2007 VereeckenRovner amp Maes 2010) and only few tested whether general par-enting influences the impact of food parenting practices (Hennessyet al 2010 Musher-Eizenman amp Holub 2006 Rodenburg et al 2013Taylor et al 2011 Van der Horst et al 2007 Vereecken et al 2010)Two of these studies found that restrictive food parenting was as-sociated with a decrease in sugar-sweetened beverage consump-tion (Van der Horst et al 2007) and caloric intake (Musher-Eizenmanamp Holub 2006) when parents reported higher levels of authorita-tive parenting The contextual influence of general parenting is likelyto be more profound than its direct relationship with weight statusor related behavior (dietary intake physical activity sedentary be-havior) (Sleddens et al 2011) Therefore general parenting can beconsidered as a crucial contextual factor in understanding the impactof food related parenting practices

Current interventions to prevent childhood overweight andobesity have been largely ineffective due to a lack of understand-ing of how family characteristics influence child behaviors (HingleOrsquoConnor Dave amp Baranowski 2010 OrsquoConnor Jago amp Baranowski2009) Childhood interventions with the best outcomes (ie in-creased healthy eating and physical activity) have engaged parents(Golley Hendrie Slater amp Corsini 2011) and interventions target-ing specific parenting practices have improved by being modifiedto include general parenting skills (Gerards Sleddens Dagnelie DeVries amp Kremers 2011) In order to develop effective interven-tions that can incorporate general parenting to tailor information

on optimal parenting practices we need to further unravel the exactmechanisms of its influence

To our knowledge the current study is one of the first attemptsto assess interactions between general parenting and food parent-ing practices in shaping childrenrsquos dietary behavior applying a lon-gitudinal study design (see Fig 1) With few exceptions empiricalstudies have dealt with one or a few isolated determinants at a timeWe investigated the impact of specific food parenting practices onchild dietary behavior in the context of broader parent level factorslike general parenting with the aim to execute research that extendsbeyond isolated associative approaches The first objective of thisstudy was to examine the extent to which food parenting prac-tices predict child dietary behavior from the childrsquos age of 6ndash8 yearsThe age span encompasses a critical period for the development ofoverweight among children (eg Pryor Tremblay Boivin Touchetteamp Dubois 2011 Ventura Loken amp Birch 2009) The second objec-tive was to explore the moderating role of general parenting on therelationship between food parenting practices and childrenrsquos dietarybehavior This contextual higher-order moderation approach is ad-vocated to have significant value in understanding the complexprocess of parentndashchild interactions in the area of childhood over-weight (Kremers et al 2013) We predicted using foods as rewardsand coercing children to eat to be positively related to unhealthychild dietary behaviors and negatively to healthy child dietary be-haviors On the other hand we predicted encouraging and control-ling food parenting practices to be positively related to healthydietary behaviors and negatively related to unhealthy dietary be-haviors Importantly we hypothesized that these relationships wouldbe strengthened andor weakened depending on general parent-ing For instance we hypothesized that the favorable associationswould be strengthened for children who were reared in a positiveparenting context The current study takes one of the first steps inunraveling complex environmentndashbehavior interactions in the re-search area of food parenting

Subjects and methods

Respondents and procedure

Data were collected within the ongoing KOALA Birth Cohort Studyin the Netherlands The study design has been previously de-scribed in detail (Kummeling et al 2005) Briefly from 2000onwards healthy pregnant women with a conventional lifestyle(N = 2343) were recruited from an existing cohort study onpregnancy-related pelvic girdle pain In addition pregnant womenwith an lsquoalternativersquo lifestyle (N = 491) with regard to dietary habits(eg preferring organic foods) vaccination programs andor

Food parenting practices(Measured at age 6)

EncouragementControl

Instrumental feedingEmotional feeding

Covert controlPressure to eat

RestrictionMonitoring

Child dietary behavior(Measured at age 6 and 8)

SnackingSugar-sweetened drinks

Fruit

General parentingNurturanceStructure

Behavioral controlCoercive controlOverprotection

B

A

C

Fig 1 Hypothesized model of the parentingndashchild dietary behavior relationship

43EFC Sleddens et alAppetite 79 (2014) 42ndash50

antibiotics use (most often refusal of the latter two) were re-cruited through several lsquoalternativersquo circles like anthroposophical phy-sician offices and midwives and organic food shops All participants(N = 2834) were enrolled between 14 and 18 weeks of gestationand were asked to complete questionnaires during pregnancy andat regular intervals after birth Informed consents were signed byall parents and ethical approval was obtained from the Maas-tricht UniversityUniversity Hospital Maastricht medical ethicscommittee

Measures

When the children were around 6 years old (mean age = 661SD = 65) parents (mostly mothers) completed a questionnaire re-garding their food parenting practices and dietary intake A total of1828 questionnaires (76) were returned After removing the secondborn of twins (N = 18) removing cases who did not complete anyof the food parenting practices scales (N = 1) and cases who did notcomplete any of the dietary items (N = 156) 1654 cases remainedfor the analyses At around age 8 (mean age = 860 SD = 66) anotherquestionnaire was administered to parents of the KOALA study as-sessing general parenting and childrenrsquos dietary intake A total of1853 questionnaires (79) were returned Of those 1657 cases alsohad completed the previous measurement but only 1654 cases re-mained for the analyses (see above)

Food parenting practicesDifferent validated instruments were used to measure food par-

enting practices at the childrsquos age of 6 the Parental Feeding StyleQuestionnaire (PFSQ) (Wardle Sanderson Guthrie Rapoport ampPlomin 2002) parental covert control over eating (Ogden Reynoldsamp Smith 2006) and the Child Feeding Questionnaire (CFQ) (Birchet al 2001) Mean scores were computed for each subscale pro-vided that at least 60 of the items were completed

The Dutch validated translation of the PFSQ (Sleddens KremersDe Vries amp Thijs 2010) was used to assess the following foursubscales lsquoinstrumental feedingrsquo (four items eg lsquoIf my child mis-behaves I withhold hisher favorite foodrsquo) lsquoemotional feedingrsquo (fiveitems eg lsquoI give my child something to eat to make himher feelbetter when heshe is feeling upsetrsquo) lsquocontrol over eatingrsquo (10 itemseg lsquoI decide how many snacks my child should haversquo) and lsquoen-couragement to eatrsquo (eight items eg lsquoI encourage my child to enjoyhisher foodrsquo) Parents were asked to rate their feeding behavior ona five-point Likert scale from 1 (never) to 5 (always) In our sampleCronbachrsquos α for each of the subscales were 65 82 72 and 75respectively

Parental covert control has been defined as forms of controlrelated to eating that cannot be detected by the child (eg Ogdenet al 2006) It was assessed using three items adapted from the5-item covert control scale developed by Ogden et al (2006) as theseitems were better suited for our study population The items witha five-point Likert scale ranging from 1 (totally disagree) to 5 (totallyagree) were lsquoI try not to eat unhealthy food when my child isaroundrsquo lsquoI avoid buying unhealthy foods at grocery storesrsquo and lsquoItry not to buy foods that I would like because I do not want my childto have themrsquo The Cronbachrsquos α was 65 in our sample

The measures of restriction pressure to eat and monitoring werebased on the CFQ (Birch et al 2001) (1) lsquorestrictionrsquo (two itemsPearsonrsquos r = 31) the extent to which parents restrict their chil-drenrsquos access to unhealthy foods corresponding to the dietary be-haviors assessed in the questionnaire (lsquoI have to be sure that my childdoes not eat too many unhealthy foodsrsquo and lsquoI have to be sure thatmy child does not drink too many sugar-sweetened beveragesrsquo) (2)lsquopressure to eatrsquo (four items eg lsquoMy child should always eat all ofthe food on hisher plate) and (3) lsquomonitoringrsquo (four items eg lsquoHowmuch do you keep track of the snackssweets that your child eatsrsquo)

Three additional items not originally in the CFQ asked parents toreport on monitoring of their childrsquos sugar-sweetened beverage andfruit intake and the amount of foods their child consumes duringbreakfast (ie lsquoHow much do you keep track of the sugar-sweetenedbeveragefruit that your child consumesrsquo lsquoHow much do you keeptrack of the foods that your child consumes during breakfastrsquo) Theresponse format consisted of a five-point Likert scale from 1 (never)to 5 (always) Cronbachrsquos α for the pressure to eat and monitoringscales were 53 and 75 respectively in our sample

General parentingThe Comprehensive General Parenting Questionnaire (CGPQ) ad-

ministered at the childrsquos age of 8 years is a recently developed andvalidated parental-reported measure consisting of 85 items to assessthe five key constructs of general parenting nurturance (eg lsquoI en-courage my child to be curious to explore and to question thingsrsquo)structure (eg lsquoI use clear and consistent messages when I tell mychild to do somethingrsquo) behavioral control (eg lsquoI am aware of mychildrsquos choice of friends who they are what they are likersquo) coer-cive control (eg lsquoI do not allow my child to question my deci-sionsrsquo) and overprotection (eg lsquoevery free minute I have I spendwith my childrsquo) (Sleddens et al 2014) The first three constructs canbe regarded as the more lsquopositive aspectsrsquo of parenting or author-itative behaviors (Sleddens et al 2014) The measure was devel-oped as consensually identified parenting questionnaires of highquality are currently lacking (eg most instruments only assess-ing limited aspects of parenting) (Sleddens et al 2014) When com-pleting the CGPQ parents have to indicate on a five-point Likert scalehow much they agree with statements regarding parenting rangingfrom 1 (strongly disagree) to 5 (strongly agree) The questionnaire con-sists of 17 subscales with five items each (Sleddens et al 2014) Meanscores were computed for each subscale provided that at least 60of the items were completed The Cronbachrsquos α for each of the fivehigher-order constructs in our sample were 82 76 74 74 and 69respectively Based on poor psychometric properties of the con-struct behavioral control in this study (Cronbachrsquos α = 53) wedecided to delete the newly developed subconstruct of lsquoconsider-ing child inputrsquo potentially belonging to this construct (Sleddens et al2014) Further validation of the CGPQ is necessary to assess the psy-chometrics of this subconstruct

Child dietary behaviorDietary behavior of children was assessed using 10 items from

a validated Food Frequency Questionnaire (FFQ) administered at thechildrsquos age of 6 and 8 years designed to accurately assess energyintake of Dutch children aged 2ndash12 years (Brants Stafleu Ter DoestHulshof amp Thijs 2006 Dutman et al 2010) The 10 items werechosen as we were interested in the consumption of snacks fruitsugar-sweetened drinks and water These dietary behaviors havebeen found to be related to the development and prevention of child-hood obesity (eg Daniels amp Popkin 2010 Hu 2013 Ledoux Hingleamp Baranowski 2011 Rosenheck 2008) Additionally the partici-pant burden was too high if we administered the total FFQ also as-sessing other dietary behaviors The original FFQ (71 items) has beenvalidated using the doubly labeled water method and shows thatmean energy intake as calculated from these items did not differfrom the mean energy expenditure as assessed with the doublylabeled water method The Pearson correlation coefficient betweenenergy intake and energy expenditure was 62 (Dutman et al 2010)For all of the eating and drinking variables the following responsecategories were used never less than 1 day a week 1 day a week2ndash3 days a week 4ndash5 days a week and 6ndash7 days a week

Childrenrsquos snacking frequency of several sugar-sweetened andenergy-dense food products (between meals) was assessed with fouritems derived from this FFQ Parents were asked to indicate howmany days a week (normal week) their children consumed the

44 EFC Sleddens et alAppetite 79 (2014) 42ndash50

following snacks in between meals (1) fried snacks (2) potato crispssalted biscuits and peanuts (3) cake or large biscuits and (4) piepastry candy bars and chocolates A single score was calculated forthe number of snacking occasions (between meals) by adding re-ported frequency (in days a week) of the different snacks (ie range0ndash28)

Both childrenrsquos sugar-sweetened soft drink consumption andwater consumption were assessed with two items Parents wereasked to indicate on how many days a week their child consumedthese drinks Additionally parents were asked to indicate the numberof glasses their child consumed of these drinks on such a day Oneglass equals 250 ml Frequency and amount of soft drinks and waterconsumed were multiplied to obtain an average score of glasses ofsoft drinks and water consumed a week

Fruit consumption was assessed by asking parents to indicate howmany days a week (normal week) their child consumed fruit Ad-ditionally parents were asked to indicate the number of servingstheir child consumed on such a day corresponding with earlier val-idation studies (Bogers Van Assema Kester Westerterp amp Dagnelie2004) One apple or one pear counted as one serving two manda-rins counted as one serving and one bunch of grapes counted asone serving as well Multiplying frequency with the reported usualamount computed an average score for the number of fruit serv-ings consumed per week

Parental background characteristicsFor educational level seven categories were distinguished which

were recoded into three levels (low medium and high) in line withinternational classification systems (Eurostat 2007) Country ofbirth was recoded into lsquoNetherlandsrsquo versus lsquoother countryrsquo In ad-dition maternal age at birth of the child and recruitment group (con-ventional versus lsquoalternativersquo lifestyle) was used in the currentanalyses

Data analyses

Partial correlations were computed for a general assessment ofthe correlations between food parenting practices and the poten-tial moderating factors (ie five general parenting constructs) ad-justed for several covariates listed at the end of this paragraphThereafter separate linear regression analyses were performed toexamine the relationship between food parenting practices and childdietary behavior (ie snacking sugar-sweetened beverage intakefruit intake and water consumption) For each of the four diet-related outcomes the analyses were performed twice using differ-ent dependent variables one predicting dietary behavior at age 8the other predicting dietary behavior change from age 6 to 8 (ad-justing dietary behavior at age 8 for dietary behavior at age 6) Allfood parenting practices were entered simultaneously correctingfor potential confounding by the other variables Finally in orderto examine whether general parenting moderated the association

between food parenting practices and childrenrsquos dietary behaviorwe calculated interaction terms between the food parenting andgeneral parenting variables The interaction terms were added tofour different regression models (ie in each model one of the fourdietary behaviors was set as the dependent variable) in the fourthstep using a forward entering procedure The threshold for includ-ing interaction terms in the stepwise regression was set at p lt 10suggested by Stone-Romero and Liakhovitski (2002) For each of thethree models we added all main effects of the potential moderat-ing variables together in the third step Stratified linear regressionanalyses were performed in order to examine the associationbetween food parenting practices and child dietary behavior in thedifferent strata of the moderator variables (ie general parenting)For each of the moderators three groups were created based on halfa standard deviation from the mean score (less than minus5 SD minus5 SDto 5 SD and higher than 5 SD from the mean) to obtain roughlysimilar group sizes All analyses were adjusted for the influence ofseveral potential covariates recruitment group (conventional versuslsquoalternativersquo lifestyle) educational level (low medium and highhighest educational level attained within a family) country of birthof mother and father (Netherlands versus other) maternal age atbirth and gender and age of the child The hierarchical multiple re-gression analyses were additionally controlled for child baseline BMIz-score (at age 6) because parentsrsquo food parenting practices maydepend on to their weight perceptions of their child (eg WebberCooke Hill amp Wardle 2010)

Results

Of the 1654 children eligible for the current study 513 were boysand 487 were girls The majority of the mothers (968) and fathers(958) were born in the Netherlands Moreover most of the fami-lies were characterized by a conventional lifestyle (813) com-pared with an lsquoalternativersquo lifestyle (187) Educational level washigh for 666 of the families medium for 287 and low for 48Average maternal age at the time of their child birth was 3231 years(SD = 372)

Children consumed snacks 493 (SD = 283) times a week at age8 Childrenrsquos mean intake of sugar-sweetened beverage and waterat age 8 was 220 (SD = 426) and 975 (SD = 860) glasses per weekrespectively The mean number of servings of fruit per week at age8 was 775 (SD = 432)

Associations between general parenting and food parenting(A in Fig 1)

General parenting practices correlated with most of the food par-enting practices although modestly at most (r = 24 or lower Table 1)Nurturing and structuring parents applied less instrumental andemotional food parenting practices and more encouragement andcontrolling practices Parents scoring high on behavioral control more

Table 1Partial correlations between food parenting practices and general parenting

Mean (SD) Food parenting practices

Instrumentalfeeding

Emotionalfeeding

Control Encouragement Covertcontrol

Pressureto eat

Restriction Monitoring

General parentingNurturance 450 (32) minus16 minus11 15 31 08 02 18 18Structure 408 (37) minus21 minus19 24 23 08 01 18 18Behavioral control 401 (39) 06 02 19 17 07 21 19 19Coercive control 195 (41) 23 17 05 minus09 02 18 minus04 02Overprotection 269 (52) 02 10 03 07 17 15 04 14

Note n = 1624 Partial rank correlation coefficients adjusted for recruitment group parental educational level country of birth maternal age at birth and child gender andage CGPQCFQ answering scale 1ndash5 p le 05 p le 01 p le 001

45EFC Sleddens et alAppetite 79 (2014) 42ndash50

often used encouraging and controlling food parenting practices Co-ercive control was positively related to instrumental and emotion-al feeding and overprotection was related to emotional feedingOverprotective parents more often indicated to keep foods out ofreach of their children (covert control) The controlling general par-enting constructs (ie behavioral control coercive control and over-protection) were positively related to pressure to eat

Associations between food parenting and child dietary behavior(B in Fig 1)

We examined the extent to which food parenting practices predictchild dietary behavior The strength of associations was generallylow as indicated by the size of the standardized regression coeffi-cients between minus19 and 13 (see Table 2) The associations betweenfood parenting practices and childrsquos dietary behavior at age 8 weremostly attenuated by controlling for baseline dietary behavior andBMI z-score at age 6 (compare models 1 and 2 Table 2) Instrumen-tal feeding at age 6 was related to less fruit intake at age 8 This as-sociation was also present with fruit intake change from age 6 to 8as an outcome Emotional feeding was related to increased childsnack intake both cross-sectionally and longitudinally Encourage-ment was positively related to fruit intake when not controlling forbaseline fruit intake and BMI z-score Pressure to eat was positive-ly associated with unhealthy behavior (ie snacking and consump-tion of sugar-sweetened beverages) Monitoring was not significantlyrelated to dietary behavior The results for the controlling food par-enting practices (eg PFSQ control CFQ restriction and covertcontrol) were mixed Control as measured by the PFSQ was not as-sociated with any of the child eating outcomes Parental restric-tion had a desirable effect on the consumption of sugar-sweeteneddrinks and water (leading to a lower and higher intake respective-ly) Parents who kept unhealthy foods out of reach from their chil-dren (covert control) were more likely to have children who areeating healthily (ie less snacking and sugar-sweetened beverageintake and more fruit and water intake)

General parenting constructs moderating the food parentingndashchilddietary behavior relationship (C in Fig 1)

We explored the moderating role of general parenting on the re-lationship between food parenting practices and childrenrsquos dietarybehavior In Table 3 we provide all significant interaction terms

between the general parenting and food parenting practices in ex-plaining child dietary behavior at age 8 and change from age 6 to8 and the results of the stratified analyses In Fig 2 we graphical-ly display these stratified analyses only for associations between thefood parenting practices and the outcomes that were statisticallysignificant in one or more of the strata of the moderator variable(12 graphs in total)

In total 22 interactions were found between food parenting prac-tices and general parenting in predicting child dietary behavior Gen-erally we found that the parenting practices of encouragement(Fig 21) and covert control (Fig 22) had more desirable effects whenparents scored higher on positive (ie behavioral control) and loweron negative (ie overprotection) general parenting constructs re-spectively One exception for all levels of overprotection the as-sociation between covert controlling parenting and child snackingwas desirable For restriction (Fig 23) more desirable effects werefound when parents scored low on behavioral control (more fruitintake) and high on coercive control (more water intake) Instru-mental feeding (Fig 24) had stronger relations with undesirable childdietary behavior (ie less fruit intake) in less controlled and over-protective home environments When parents scored intermedi-ate on behavioral control instrumental feeding was related to lowersugar-sweetened beverage intake whereas when parents scored highon behavioral control instrumental feeding was related to highersugar-sweetened beverage intake Additionally the negative rela-tion between instrumental feeding and water consumption wasstronger among children of parents who scored high on nurturanceLastly pressure to eat (Fig 25) indicated to have an undesirable effecton child dietary behavior (ie more sugar-sweetened beverageintake) among children of parents scoring intermediate and highon behavioral control For the remaining 10 interactions the strati-fied analyses were not significant so they are not considered further

Discussion

Many studies on food parenting fail to consider the larger familycontext The present study provides evidence for effective and in-effective food parenting practices as well as for child- and parent-related contexts that moderate their impact The relatively weakmain effects could be summarized as follows instrumental andemotional feeding and pressure to eat were found to have detri-mental associations with child dietary behavior (ie increased un-healthy intakedecreased healthy intake) whereas covert control

Table 2Associations between food parenting practices and child dietary behavior

Mean (SD) Child dietary behavior

Snacking Sugar-sweetened drinks Fruit Water

Model 1n = 1390Beta

Model 2n = 1378Beta

Model 1n = 1159Beta

Model 2n = 1044Beta

Model 1n = 1159Beta

Model 2n = 1078Beta

Model 1n = 1108Beta

Model 2n = 982Beta

BMI z-score minus05 minus05 03 00 04 01 07 04Baseline eating minus minus 54 minus 49 minus 61 minus 64Food parenting

Instrumental feeding 172 (57) 01 minus01 minus01 minus01 minus09 minus09 minus03 03Emotional feeding 142 (46) 13 07 02 minus01 04 05 minus05 minus05Control 436 (36) minus03 01 minus05 minus03 minus02 minus01 minus02 00Encouragement 388 (47) minus02 minus03 minus01 minus03 08 05 06 minus02Covert control 293 (93) minus19 minus09 minus08 minus02 09 04 12 04Pressure to eat 351 (75) 07 03 11 07 minus05 00 minus05 minus01Restriction 452 (62) minus04 00 minus13 minus07 06 02 10 04Monitoring 431 (56) minus05 minus04 04 05 03 03 minus02 00

Note Beta standardized linear regression coefficient Model 1 dependent variable is child dietary behavior around the age of 8 years which is additionally adjusted for childBMI z-score at age 6 Model 2 dependent variable is additionally adjusted for baseline child dietary behavior (snacking or sugar-sweetened drink intake or fruit intake orwater intake) at age 6 and child BMI z-score at age 6 both models adjusted for recruitment group parental educational level country of birth maternal age at birth andchild gender and age food parenting practices answering scale 1ndash5 p le 05 p le 01 p le 001

46 EFC Sleddens et alAppetite 79 (2014) 42ndash50

encouragement and restriction were found to have favorable asso-ciations with child dietary behavior (ie increased healthy intakedecreased unhealthy intake) This is confirming our hypotheses andmost of these relationships have been demonstrated previously (egBrown Ogden Voumlgele amp Gibson 2008 Gubbels et al 2009Murashima Hoerr Hughes amp Kaplowitz 2012 Rodenburg et al2013 Sleddens et al 2010 Van der Horst et al 2007) These as-sociations were stronger for some subgroups of the moderating vari-ables as expected The associations of encouragement and covertcontrol with desirable child dietary behaviors were found to be stron-ger for children who were reared in a positive parenting context

Associations between food parenting and general parenting

The more lsquopositiversquo constructs of general parenting (ienurturance and structure) (Sleddens et al 2014) were associatedwith lower use of instrumental and emotional feeding Reversed as-sociations were found for the more firmly controlling parents Thusthose parents tend to give their child more foods in response togood behavior and to soothe their child The other scales exceptpressure to eat were positively related to the authoritative aspectsof parenting For pressure to eat we found positive associations withcontrolling parenting styles This finding is partially supported byprevious studies (Hennessy et al 2010 Hubbs-Tait et al 2008)

Associations between food parenting and child dietary behavior

We confirmed some of the results of previous studies in whichnondirective child-centered food practices were related to consum-ing healthier diets (Murashima et al 2012 Vereecken et al 2010)and parental reward of food was related to unhealthy food (Kroumllleramp Warschburger 2009 Sleddens et al 2010) and soft drink intake(Kroumlller amp Warschburger 2009) Parental restriction was negative-ly related to sugar-sweetened beverage intake and positively relatedto water intake in contrast to previous studies that showed that high

levels of restriction can lead to increases in calorie intake and likingfor the restricted food (eg Fisher amp Birch 1999 Jansen Mulkensamp Jansen 2007) However the latter types of studies were mainlyexperimental looking at what the child eats when the parent is notaround with free access to food In line with a previous study withinthe KOALA birth cohort by Gubbels et al (2009) we failed to confirmthe paradoxical effect of increased intake of unhealthy foods fol-lowing parental restriction although we should be cautious in in-terpreting these findings as parent-reported child dietary intake onlyaccounts for the dietary behaviors that parents see and are awareof Coercing children to eat was associated with unhealthy eatingHigher levels of pressure in child feeding could have detrimentaleffects on childrenrsquos change in healthy dietary behavior as chil-dren are focused away from internal cues to hunger and satiety(Francis Hofer amp Birch 2001) leading to decreases in preferenceand intake of the healthy foods and subsequent increases in con-sumption of unhealthy foods

General parenting constructs modifying the food parentingndashchilddietary behavior relationship

Moderating effects of general parenting on the relationshipbetween food parenting practices and child eating behavior werefound For children who were reared in a positive parenting contextencouragement and covert control were found to work better (ieincreased healthy intakedecreased unhealthy intake) However someconflicting findings were found The one concept standing out mostwas pressure to eat which was related to sugar-sweetened bever-age intake for children of parents scoring high on behavioral controlPressure to eat often occurs when parents feel that their child iseating insufficient amounts of food or in response to their childrsquosunderweight (Francis et al 2001) Secondary analyses confirmedthis assumption as we found that parents of children who were un-derweight scored significantly higher on pressure to eat com-pared with parents of children who were overweight or obese (data

Table 3Interaction terms and stratified analyses food parentingndashchild dietary behavior relationship

Dietary behavior Year Interaction term(general parenting times food parenting)

Groups

Low Intermediate High

Beta P Beta P Beta P Beta P

Snacking 6ndash8 Behavioral control times Encouragement minus62 034 02 624 00 994 minus11 010Structure times Pressure to eat 53 045 03 533 minus02 641 08 053

8 Overprotection times Covert control minus27 099 minus14 005 minus21 000 minus19 000Sugar-sweetened beverage 6ndash8 Behavioral control times Instrumental feeding 77 007 minus04 509 minus11 048 12 036

Behavioral control times Encouragement minus91 009 03 539 01 788 minus14 0058 Behavioral control times Pressure to eat 82 018 03 578 15 003 13 021

Behavioral control times Encouragement minus116 005 07 218 minus01 899 minus08 135Behavioral control times Control 96 054 minus07 221 minus07 149 minus02 688

Fruit 6ndash8 Nurturance times Monitoring 89 034 minus05 344 05 308 07 081Behavioral control times Instrumental feeding 52 049 minus16 003 minus08 080 minus01 785

8 Behavioral control times Monitoring 125 004 minus04 444 03 553 08 169Behavioral control times Emotional feeding 66 027 minus05 439 07 220 09 163Behavioral control times Restriction minus69 086 15 012 10 064 minus05 437Overprotection times Instrumental feeding 85 031 minus02 800 minus13 028 minus14 043Overprotection times Encouragement minus52 091 14 011 06 231 06 312

Water 6ndash8 Nurturance times Instrumental feeding minus70 024 01 874 08 067 minus04 467Overprotection times Monitoring minus41 070 03 536 00 977 minus04 445

8 Nurturance times Instrumental feeding minus110 004 06 416 02 709 minus15 013Coercive control times Encouragement minus55 047 09 112 07 157 minus01 881Coercive control times Restriction 47 083 05 459 10 047 16 013Overprotection times Covert Control minus34 062 17 003 07 170 09 184Overprotection times Pressure to eat 53 014 minus06 320 minus09 091 02 712

Note Interactions for which the association between the food parenting practice and the outcome was statistically significant in one or more groups of the moderator vari-able Beta standardized regression coefficient from linear regression analysis stratified for the moderator (ie low intermediate high) model adjusted for recruitment groupparental educational level country of birth maternal age at birth child gender and age child BMI z-score at age 6 and all other food parenting practices (note that therewere two different models (1) dependent variable is additionally adjusted for baseline child dietary behavior at age 6 and adjusted for child BMI z-score at age 6 (2) de-pendent variable is child dietary behavior at age 8 adjusted for child BMI z-score at age 6

47EFC Sleddens et alAppetite 79 (2014) 42ndash50

nsns

nsns

ns ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

ASnacking Sugar sweetened

beverageFruit

Behavioral control OverprotectionBehavioral control

6-8 yr 6-8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 21 Food parenting practice lsquoencouragementrsquo

nsns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Snacking Water

OverprotectionOverprotection

8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 22 Food parenting practice lsquocovert controlrsquo

ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Fruit Water

Behavioral control Coercive control

8 yr 6-8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 23 Food parenting practice lsquorestrictionrsquo

ns

ns

ns ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidrad nats(no itaicoss

A

Sugar sweetened beverage

Fruit WaterFruit

Behavioral control Behavioral control Overprotection Nurturance

6-8 yr 6-8 yr 8-yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 24 Food parenting practice lsquoinstrumental feedingrsquo

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Sugar sweetened beverage

Behavioral control

8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 25 Food parenting practice lsquopressure to eatrsquo

Fig 2 Stratified analyses the food parentingndashchild dietary behavior relationship moderated by general parenting Note Red bars represent significant associations betweenfood parenting and undesirable behavior (ie increased unhealthy intakedecreased healthy intake) green bars represent significant associations between food parentingand desirable behavior (ie increased healthy intakedecreased unhealthy intake) white represent nonsignificant associations between food parenting and child dietaryintake bars within each cluster represent the stratified levels of the moderator (iegeneral parenting constructs) p le 05 p le 01 p le 001

48 EFC Sleddens et alAppetite 79 (2014) 42ndash50

not reported) Low weight status has been proven to be related withmore food avoidant behaviors (Jansen et al 2012 Sleddens Kremersamp Thijs 2008) As a result parents could apply coercing food par-enting practices trying to increase their childrsquos calorie intake andprobably also increasing their childrsquos unhealthy dietary behavior (iesnacking and sugar-sweetened beverage intake) Our moderationresults should be interpreted with caution and more studies areneeded before drawing definite conclusions

Study strengths and limitations

This study benefitted from a longitudinal design with mea-sures of child dietary behavior repeated at age 6 and 8 The in-cluded variables (food parenting practices and general parenting)were measured with instruments validated for the Dutch context(Bogers et al 2004 Brants et al 2006 Dutman et al 2010 Sleddenset al 2010 2014) Moderation analysis was possible thanks to alarge sample size and sufficient diversity within the study Weconfirmed the operation of higher-order moderation processesimplying that parenting factors at higher more distal levels alterthe impact of food parenting practices at more proximal levels Werecommend future studies to include the theory-based examina-tion of possible moderation effects and ensure sufficient study sizeto do so or examine specific hypotheses in smaller scale studies withcareful selection of the contextual situation

There were also some limitations that need special attentionHighly educated parents and parents with an lsquoalternativersquo lifestylewere overrepresented in our sample partially due to the choice ofrecruitment methods (Gubbels et al 2009) although mean valuesfor the parenting variables (ie general parenting and food parent-ing) were comparable with previous studies (eg Sleddens et al2010 2014) We therefore adjusted all analyses for highest educa-tion level attained and recruitment channel Additionally all of themeasured variables were reported by parents This may evoke socialdesirability The associations between food parenting practices andchildrenrsquos dietary behavior should therefore be interpreted withcaution Cronbachrsquos alpha for the pressure to eat scale of the CFQwas quite low although this finding is in accordance with some pre-vious studies using this scale (eg Boles et al 2010 Gregory Paxtonamp Brozovic 2010) In spite of the longitudinal analysis causality isdifficult to establish since part of the associations (and interac-tions) may be modified by parental adaptations to unwanted be-havior Snacking and intake of sugar-sweetened drinks were studiedas unhealthy behaviors and fruit and water intake as healthy be-haviors Of course other behaviors such as breakfast and vegeta-ble consumption are important as well to determine dietary intakeFinally any choice of single food groups as healthy or unhealthy isdeemed to be debatable due to complexities such as substitution(eg between fruit natural fruit drinks sweetened fruit drinks andsoft drinks tea water and milk drinks) and the ambiguities of re-lations with specific health indicators such as nutritional imbal-ances dental health and overweight development Further studiesare needed with specific health outcomes to evaluate whether mod-eration by contextual factors as shown in our study with health be-haviors as outcome also translate to health outcomes such asoverweight development

Conclusion

Our results show that food parenting practices are important de-terminants in explaining child dietary behavior and that general par-enting behaviors moderate this association Future research effortsshould continue to focus on testing the influence of factors (eggeneral parenting but also child characteristics such as tempera-ment and eating style) impacting on the food parentingndashchild eatingandor weight relationship in order to gain insights into relevant con-

textual factors that need to be taken into account in designinginterventions

References

Birch L L Fisher J O Grimm-Thomas K Markey C N Sawyer R amp JohnsonS L (2001) Confirmatory factor analysis of the Child Feeding Questionnaire Ameasure of parental attitudes beliefs and practices about child feeding andobesity proneness Appetite 36 201ndash210

Blissett J amp Haycraft E (2008) Are parenting style and controlling feeding practicesrelated Appetite 50 477ndash485

Blissett J Meyer C amp Haycraft E (2011) The role of parenting in the relationshipbetween childhood eating problems and broader behaviour problems Child CareHealth and Development 37 642ndash648

Bogers R P Van Assema P Kester A D M Westerterp K R amp Dagnelie P C (2004)Reproducibility validity and responsiveness to change of a short questionnairefor measuring fruit and vegetable intake American Journal of Epidemiology 159900ndash909

Boles R E Nelson T D Chamberlin L A Valenzuela J M Sherman S N JohnsonS L et al (2010) Confirmatory factor analysis of the Child Feeding Questionnaireamong low-income African American families of preschool children Appetite 54402ndash405

Brants H Stafleu A Ter Doest D Hulshof K amp Thijs C (2006) Ontwikkeling vaneen voedselfrequentievragenlijst Energie-inneming van kinderen van 2 tot enmet 12 jaar [Development of a food frequency questionnaire Energy intake ofchildren 2-12 years of age] Voeding Nu 2 25ndash28

Brown K A Ogden J Voumlgele C amp Gibson E L (2008) The role of parental controlpractices in explaining childrenrsquos diet and BMI Appetite 50 252ndash259

Cislak A Safron M Pratt M Gaspar T amp Luszczynska A (2012) Family-relatedpredictors of body weight and weight-related behaviours among children andadolescents A systematic umbrella review Child Care Health and Development38 321ndash331

Clark H R Goyder E Bissell P Blank L amp Peters J (2007) How do parentsrsquochild-feeding behaviours influence child weight Implications for childhoodobesity policy Journal of Public Health 29 132ndash141

Daniels M C amp Popkin B M (2010) Impact of water intake on energy intake andweight status A systematic review Nutrition Reviews 68 505ndash521

Daniels S R (2009) Complications of obesity in children and adolescentsInternational Journal of Obesity 33(s1) 60ndash65

Darling N amp Steinberg L (1993) Parenting style as context An integrative modelPsychological Bulletin 113 487ndash496

Dutman A E Stafleu A Kruizinga A Brants H A M Westerterp K R KistemakerC et al (2010) Validation of an FFQ and options for data processing using thedoubly labelled water method in children Public Health Nutrition 14 410ndash417

Eurostat (2007) Task force on core social variables Final report Luxembourg EuropeanCommunitees

Faith M S Scanlon K S Birch L L Francis L A amp Sherry B (2004) Parent-childfeeding strategies and their relationships to child eating and weight status ObesityResearch 12 1711ndash1722

Fisher J O amp Birch L L (1999) Restricting access to foods and childrenrsquos eatingAppetite 32 405ndash419

Francis L A Hofer S M amp Birch L L (2001) Predictors of maternal child-feedingstyle Maternal and child characteristics Appetite 37 231ndash243

Fuemmeler B F Yang C Costanzo P Hoyle R H Siegler I C Williams R B et al(2012) Parenting styles and body mass index trajectories from adolescence toadulthood Health Psychology 31 441ndash449

Gerards S M P L Sleddens E F C Dagnelie P C De Vries N K amp Kremers S PJ (2011) Interventions addressing general parenting to prevent or treat childhoodobesity International Journal of Pediatric Obesity 6 e28ndashe45

Golley R K Hendrie G A Slater A amp Corsini N (2011) Interventions thatinvolve parents to improve childrenrsquos weight-related nutrition intake andactivity patterns What nutrition and activity targets and behaviour changetechniques are associated with intervention effectiveness Obesity Reviews 12114ndash130

Gregory J E Paxton S J amp Brozovic A M (2010) Maternal feeding practices childeating behaviour and body mass index in preschool-aged children A prospectiveanalysis International Journal of Behavioral Nutrition and Physical Activity 7 55

Gubbels J S Kremers S P J Stafleu A Dagnelie P C Goldbohm R A De VriesN K et al (2009) Diet-related restrictive parenting practices Impact on dietaryintake of 2-year-old children and interactions with child characteristics Appetite52 423ndash429

Hennessy E Hughes S O Goldberg J P Hyatt R R amp Economos C D (2010) Parentbehavior and child weight status among a diverse group of underserved ruralfamilies Appetite 54 369ndash377

Hingle M D OrsquoConnor T M Dave J M amp Baranowski T (2010) Parentalinvolvement in interventions to improve child dietary intake A systematic reviewPreventive Medicine 51 103ndash111

Hu F B (2013) Resolved There is sufficient scientific evidence that decreasingsugar-sweetened beverage consumption will reduce the prevalence of obesityand obesity-related diseases Obesity Reviews 14 606ndash619

Hubbs-Tait L Kennedy T S Page M C Topham G L amp Harrist A W(2008) Parental feeding practices predict authoritative authoritarian andpermissive parenting styles Journal of the American Dietetic Association 1081154ndash1161

49EFC Sleddens et alAppetite 79 (2014) 42ndash50

antibiotics use (most often refusal of the latter two) were re-cruited through several lsquoalternativersquo circles like anthroposophical phy-sician offices and midwives and organic food shops All participants(N = 2834) were enrolled between 14 and 18 weeks of gestationand were asked to complete questionnaires during pregnancy andat regular intervals after birth Informed consents were signed byall parents and ethical approval was obtained from the Maas-tricht UniversityUniversity Hospital Maastricht medical ethicscommittee

Measures

When the children were around 6 years old (mean age = 661SD = 65) parents (mostly mothers) completed a questionnaire re-garding their food parenting practices and dietary intake A total of1828 questionnaires (76) were returned After removing the secondborn of twins (N = 18) removing cases who did not complete anyof the food parenting practices scales (N = 1) and cases who did notcomplete any of the dietary items (N = 156) 1654 cases remainedfor the analyses At around age 8 (mean age = 860 SD = 66) anotherquestionnaire was administered to parents of the KOALA study as-sessing general parenting and childrenrsquos dietary intake A total of1853 questionnaires (79) were returned Of those 1657 cases alsohad completed the previous measurement but only 1654 cases re-mained for the analyses (see above)

Food parenting practicesDifferent validated instruments were used to measure food par-

enting practices at the childrsquos age of 6 the Parental Feeding StyleQuestionnaire (PFSQ) (Wardle Sanderson Guthrie Rapoport ampPlomin 2002) parental covert control over eating (Ogden Reynoldsamp Smith 2006) and the Child Feeding Questionnaire (CFQ) (Birchet al 2001) Mean scores were computed for each subscale pro-vided that at least 60 of the items were completed

The Dutch validated translation of the PFSQ (Sleddens KremersDe Vries amp Thijs 2010) was used to assess the following foursubscales lsquoinstrumental feedingrsquo (four items eg lsquoIf my child mis-behaves I withhold hisher favorite foodrsquo) lsquoemotional feedingrsquo (fiveitems eg lsquoI give my child something to eat to make himher feelbetter when heshe is feeling upsetrsquo) lsquocontrol over eatingrsquo (10 itemseg lsquoI decide how many snacks my child should haversquo) and lsquoen-couragement to eatrsquo (eight items eg lsquoI encourage my child to enjoyhisher foodrsquo) Parents were asked to rate their feeding behavior ona five-point Likert scale from 1 (never) to 5 (always) In our sampleCronbachrsquos α for each of the subscales were 65 82 72 and 75respectively

Parental covert control has been defined as forms of controlrelated to eating that cannot be detected by the child (eg Ogdenet al 2006) It was assessed using three items adapted from the5-item covert control scale developed by Ogden et al (2006) as theseitems were better suited for our study population The items witha five-point Likert scale ranging from 1 (totally disagree) to 5 (totallyagree) were lsquoI try not to eat unhealthy food when my child isaroundrsquo lsquoI avoid buying unhealthy foods at grocery storesrsquo and lsquoItry not to buy foods that I would like because I do not want my childto have themrsquo The Cronbachrsquos α was 65 in our sample

The measures of restriction pressure to eat and monitoring werebased on the CFQ (Birch et al 2001) (1) lsquorestrictionrsquo (two itemsPearsonrsquos r = 31) the extent to which parents restrict their chil-drenrsquos access to unhealthy foods corresponding to the dietary be-haviors assessed in the questionnaire (lsquoI have to be sure that my childdoes not eat too many unhealthy foodsrsquo and lsquoI have to be sure thatmy child does not drink too many sugar-sweetened beveragesrsquo) (2)lsquopressure to eatrsquo (four items eg lsquoMy child should always eat all ofthe food on hisher plate) and (3) lsquomonitoringrsquo (four items eg lsquoHowmuch do you keep track of the snackssweets that your child eatsrsquo)

Three additional items not originally in the CFQ asked parents toreport on monitoring of their childrsquos sugar-sweetened beverage andfruit intake and the amount of foods their child consumes duringbreakfast (ie lsquoHow much do you keep track of the sugar-sweetenedbeveragefruit that your child consumesrsquo lsquoHow much do you keeptrack of the foods that your child consumes during breakfastrsquo) Theresponse format consisted of a five-point Likert scale from 1 (never)to 5 (always) Cronbachrsquos α for the pressure to eat and monitoringscales were 53 and 75 respectively in our sample

General parentingThe Comprehensive General Parenting Questionnaire (CGPQ) ad-

ministered at the childrsquos age of 8 years is a recently developed andvalidated parental-reported measure consisting of 85 items to assessthe five key constructs of general parenting nurturance (eg lsquoI en-courage my child to be curious to explore and to question thingsrsquo)structure (eg lsquoI use clear and consistent messages when I tell mychild to do somethingrsquo) behavioral control (eg lsquoI am aware of mychildrsquos choice of friends who they are what they are likersquo) coer-cive control (eg lsquoI do not allow my child to question my deci-sionsrsquo) and overprotection (eg lsquoevery free minute I have I spendwith my childrsquo) (Sleddens et al 2014) The first three constructs canbe regarded as the more lsquopositive aspectsrsquo of parenting or author-itative behaviors (Sleddens et al 2014) The measure was devel-oped as consensually identified parenting questionnaires of highquality are currently lacking (eg most instruments only assess-ing limited aspects of parenting) (Sleddens et al 2014) When com-pleting the CGPQ parents have to indicate on a five-point Likert scalehow much they agree with statements regarding parenting rangingfrom 1 (strongly disagree) to 5 (strongly agree) The questionnaire con-sists of 17 subscales with five items each (Sleddens et al 2014) Meanscores were computed for each subscale provided that at least 60of the items were completed The Cronbachrsquos α for each of the fivehigher-order constructs in our sample were 82 76 74 74 and 69respectively Based on poor psychometric properties of the con-struct behavioral control in this study (Cronbachrsquos α = 53) wedecided to delete the newly developed subconstruct of lsquoconsider-ing child inputrsquo potentially belonging to this construct (Sleddens et al2014) Further validation of the CGPQ is necessary to assess the psy-chometrics of this subconstruct

Child dietary behaviorDietary behavior of children was assessed using 10 items from

a validated Food Frequency Questionnaire (FFQ) administered at thechildrsquos age of 6 and 8 years designed to accurately assess energyintake of Dutch children aged 2ndash12 years (Brants Stafleu Ter DoestHulshof amp Thijs 2006 Dutman et al 2010) The 10 items werechosen as we were interested in the consumption of snacks fruitsugar-sweetened drinks and water These dietary behaviors havebeen found to be related to the development and prevention of child-hood obesity (eg Daniels amp Popkin 2010 Hu 2013 Ledoux Hingleamp Baranowski 2011 Rosenheck 2008) Additionally the partici-pant burden was too high if we administered the total FFQ also as-sessing other dietary behaviors The original FFQ (71 items) has beenvalidated using the doubly labeled water method and shows thatmean energy intake as calculated from these items did not differfrom the mean energy expenditure as assessed with the doublylabeled water method The Pearson correlation coefficient betweenenergy intake and energy expenditure was 62 (Dutman et al 2010)For all of the eating and drinking variables the following responsecategories were used never less than 1 day a week 1 day a week2ndash3 days a week 4ndash5 days a week and 6ndash7 days a week

Childrenrsquos snacking frequency of several sugar-sweetened andenergy-dense food products (between meals) was assessed with fouritems derived from this FFQ Parents were asked to indicate howmany days a week (normal week) their children consumed the

44 EFC Sleddens et alAppetite 79 (2014) 42ndash50

following snacks in between meals (1) fried snacks (2) potato crispssalted biscuits and peanuts (3) cake or large biscuits and (4) piepastry candy bars and chocolates A single score was calculated forthe number of snacking occasions (between meals) by adding re-ported frequency (in days a week) of the different snacks (ie range0ndash28)

Both childrenrsquos sugar-sweetened soft drink consumption andwater consumption were assessed with two items Parents wereasked to indicate on how many days a week their child consumedthese drinks Additionally parents were asked to indicate the numberof glasses their child consumed of these drinks on such a day Oneglass equals 250 ml Frequency and amount of soft drinks and waterconsumed were multiplied to obtain an average score of glasses ofsoft drinks and water consumed a week

Fruit consumption was assessed by asking parents to indicate howmany days a week (normal week) their child consumed fruit Ad-ditionally parents were asked to indicate the number of servingstheir child consumed on such a day corresponding with earlier val-idation studies (Bogers Van Assema Kester Westerterp amp Dagnelie2004) One apple or one pear counted as one serving two manda-rins counted as one serving and one bunch of grapes counted asone serving as well Multiplying frequency with the reported usualamount computed an average score for the number of fruit serv-ings consumed per week

Parental background characteristicsFor educational level seven categories were distinguished which

were recoded into three levels (low medium and high) in line withinternational classification systems (Eurostat 2007) Country ofbirth was recoded into lsquoNetherlandsrsquo versus lsquoother countryrsquo In ad-dition maternal age at birth of the child and recruitment group (con-ventional versus lsquoalternativersquo lifestyle) was used in the currentanalyses

Data analyses

Partial correlations were computed for a general assessment ofthe correlations between food parenting practices and the poten-tial moderating factors (ie five general parenting constructs) ad-justed for several covariates listed at the end of this paragraphThereafter separate linear regression analyses were performed toexamine the relationship between food parenting practices and childdietary behavior (ie snacking sugar-sweetened beverage intakefruit intake and water consumption) For each of the four diet-related outcomes the analyses were performed twice using differ-ent dependent variables one predicting dietary behavior at age 8the other predicting dietary behavior change from age 6 to 8 (ad-justing dietary behavior at age 8 for dietary behavior at age 6) Allfood parenting practices were entered simultaneously correctingfor potential confounding by the other variables Finally in orderto examine whether general parenting moderated the association

between food parenting practices and childrenrsquos dietary behaviorwe calculated interaction terms between the food parenting andgeneral parenting variables The interaction terms were added tofour different regression models (ie in each model one of the fourdietary behaviors was set as the dependent variable) in the fourthstep using a forward entering procedure The threshold for includ-ing interaction terms in the stepwise regression was set at p lt 10suggested by Stone-Romero and Liakhovitski (2002) For each of thethree models we added all main effects of the potential moderat-ing variables together in the third step Stratified linear regressionanalyses were performed in order to examine the associationbetween food parenting practices and child dietary behavior in thedifferent strata of the moderator variables (ie general parenting)For each of the moderators three groups were created based on halfa standard deviation from the mean score (less than minus5 SD minus5 SDto 5 SD and higher than 5 SD from the mean) to obtain roughlysimilar group sizes All analyses were adjusted for the influence ofseveral potential covariates recruitment group (conventional versuslsquoalternativersquo lifestyle) educational level (low medium and highhighest educational level attained within a family) country of birthof mother and father (Netherlands versus other) maternal age atbirth and gender and age of the child The hierarchical multiple re-gression analyses were additionally controlled for child baseline BMIz-score (at age 6) because parentsrsquo food parenting practices maydepend on to their weight perceptions of their child (eg WebberCooke Hill amp Wardle 2010)

Results

Of the 1654 children eligible for the current study 513 were boysand 487 were girls The majority of the mothers (968) and fathers(958) were born in the Netherlands Moreover most of the fami-lies were characterized by a conventional lifestyle (813) com-pared with an lsquoalternativersquo lifestyle (187) Educational level washigh for 666 of the families medium for 287 and low for 48Average maternal age at the time of their child birth was 3231 years(SD = 372)

Children consumed snacks 493 (SD = 283) times a week at age8 Childrenrsquos mean intake of sugar-sweetened beverage and waterat age 8 was 220 (SD = 426) and 975 (SD = 860) glasses per weekrespectively The mean number of servings of fruit per week at age8 was 775 (SD = 432)

Associations between general parenting and food parenting(A in Fig 1)

General parenting practices correlated with most of the food par-enting practices although modestly at most (r = 24 or lower Table 1)Nurturing and structuring parents applied less instrumental andemotional food parenting practices and more encouragement andcontrolling practices Parents scoring high on behavioral control more

Table 1Partial correlations between food parenting practices and general parenting

Mean (SD) Food parenting practices

Instrumentalfeeding

Emotionalfeeding

Control Encouragement Covertcontrol

Pressureto eat

Restriction Monitoring

General parentingNurturance 450 (32) minus16 minus11 15 31 08 02 18 18Structure 408 (37) minus21 minus19 24 23 08 01 18 18Behavioral control 401 (39) 06 02 19 17 07 21 19 19Coercive control 195 (41) 23 17 05 minus09 02 18 minus04 02Overprotection 269 (52) 02 10 03 07 17 15 04 14

Note n = 1624 Partial rank correlation coefficients adjusted for recruitment group parental educational level country of birth maternal age at birth and child gender andage CGPQCFQ answering scale 1ndash5 p le 05 p le 01 p le 001

45EFC Sleddens et alAppetite 79 (2014) 42ndash50

often used encouraging and controlling food parenting practices Co-ercive control was positively related to instrumental and emotion-al feeding and overprotection was related to emotional feedingOverprotective parents more often indicated to keep foods out ofreach of their children (covert control) The controlling general par-enting constructs (ie behavioral control coercive control and over-protection) were positively related to pressure to eat

Associations between food parenting and child dietary behavior(B in Fig 1)

We examined the extent to which food parenting practices predictchild dietary behavior The strength of associations was generallylow as indicated by the size of the standardized regression coeffi-cients between minus19 and 13 (see Table 2) The associations betweenfood parenting practices and childrsquos dietary behavior at age 8 weremostly attenuated by controlling for baseline dietary behavior andBMI z-score at age 6 (compare models 1 and 2 Table 2) Instrumen-tal feeding at age 6 was related to less fruit intake at age 8 This as-sociation was also present with fruit intake change from age 6 to 8as an outcome Emotional feeding was related to increased childsnack intake both cross-sectionally and longitudinally Encourage-ment was positively related to fruit intake when not controlling forbaseline fruit intake and BMI z-score Pressure to eat was positive-ly associated with unhealthy behavior (ie snacking and consump-tion of sugar-sweetened beverages) Monitoring was not significantlyrelated to dietary behavior The results for the controlling food par-enting practices (eg PFSQ control CFQ restriction and covertcontrol) were mixed Control as measured by the PFSQ was not as-sociated with any of the child eating outcomes Parental restric-tion had a desirable effect on the consumption of sugar-sweeteneddrinks and water (leading to a lower and higher intake respective-ly) Parents who kept unhealthy foods out of reach from their chil-dren (covert control) were more likely to have children who areeating healthily (ie less snacking and sugar-sweetened beverageintake and more fruit and water intake)

General parenting constructs moderating the food parentingndashchilddietary behavior relationship (C in Fig 1)

We explored the moderating role of general parenting on the re-lationship between food parenting practices and childrenrsquos dietarybehavior In Table 3 we provide all significant interaction terms

between the general parenting and food parenting practices in ex-plaining child dietary behavior at age 8 and change from age 6 to8 and the results of the stratified analyses In Fig 2 we graphical-ly display these stratified analyses only for associations between thefood parenting practices and the outcomes that were statisticallysignificant in one or more of the strata of the moderator variable(12 graphs in total)

In total 22 interactions were found between food parenting prac-tices and general parenting in predicting child dietary behavior Gen-erally we found that the parenting practices of encouragement(Fig 21) and covert control (Fig 22) had more desirable effects whenparents scored higher on positive (ie behavioral control) and loweron negative (ie overprotection) general parenting constructs re-spectively One exception for all levels of overprotection the as-sociation between covert controlling parenting and child snackingwas desirable For restriction (Fig 23) more desirable effects werefound when parents scored low on behavioral control (more fruitintake) and high on coercive control (more water intake) Instru-mental feeding (Fig 24) had stronger relations with undesirable childdietary behavior (ie less fruit intake) in less controlled and over-protective home environments When parents scored intermedi-ate on behavioral control instrumental feeding was related to lowersugar-sweetened beverage intake whereas when parents scored highon behavioral control instrumental feeding was related to highersugar-sweetened beverage intake Additionally the negative rela-tion between instrumental feeding and water consumption wasstronger among children of parents who scored high on nurturanceLastly pressure to eat (Fig 25) indicated to have an undesirable effecton child dietary behavior (ie more sugar-sweetened beverageintake) among children of parents scoring intermediate and highon behavioral control For the remaining 10 interactions the strati-fied analyses were not significant so they are not considered further

Discussion

Many studies on food parenting fail to consider the larger familycontext The present study provides evidence for effective and in-effective food parenting practices as well as for child- and parent-related contexts that moderate their impact The relatively weakmain effects could be summarized as follows instrumental andemotional feeding and pressure to eat were found to have detri-mental associations with child dietary behavior (ie increased un-healthy intakedecreased healthy intake) whereas covert control

Table 2Associations between food parenting practices and child dietary behavior

Mean (SD) Child dietary behavior

Snacking Sugar-sweetened drinks Fruit Water

Model 1n = 1390Beta

Model 2n = 1378Beta

Model 1n = 1159Beta

Model 2n = 1044Beta

Model 1n = 1159Beta

Model 2n = 1078Beta

Model 1n = 1108Beta

Model 2n = 982Beta

BMI z-score minus05 minus05 03 00 04 01 07 04Baseline eating minus minus 54 minus 49 minus 61 minus 64Food parenting

Instrumental feeding 172 (57) 01 minus01 minus01 minus01 minus09 minus09 minus03 03Emotional feeding 142 (46) 13 07 02 minus01 04 05 minus05 minus05Control 436 (36) minus03 01 minus05 minus03 minus02 minus01 minus02 00Encouragement 388 (47) minus02 minus03 minus01 minus03 08 05 06 minus02Covert control 293 (93) minus19 minus09 minus08 minus02 09 04 12 04Pressure to eat 351 (75) 07 03 11 07 minus05 00 minus05 minus01Restriction 452 (62) minus04 00 minus13 minus07 06 02 10 04Monitoring 431 (56) minus05 minus04 04 05 03 03 minus02 00

Note Beta standardized linear regression coefficient Model 1 dependent variable is child dietary behavior around the age of 8 years which is additionally adjusted for childBMI z-score at age 6 Model 2 dependent variable is additionally adjusted for baseline child dietary behavior (snacking or sugar-sweetened drink intake or fruit intake orwater intake) at age 6 and child BMI z-score at age 6 both models adjusted for recruitment group parental educational level country of birth maternal age at birth andchild gender and age food parenting practices answering scale 1ndash5 p le 05 p le 01 p le 001

46 EFC Sleddens et alAppetite 79 (2014) 42ndash50

encouragement and restriction were found to have favorable asso-ciations with child dietary behavior (ie increased healthy intakedecreased unhealthy intake) This is confirming our hypotheses andmost of these relationships have been demonstrated previously (egBrown Ogden Voumlgele amp Gibson 2008 Gubbels et al 2009Murashima Hoerr Hughes amp Kaplowitz 2012 Rodenburg et al2013 Sleddens et al 2010 Van der Horst et al 2007) These as-sociations were stronger for some subgroups of the moderating vari-ables as expected The associations of encouragement and covertcontrol with desirable child dietary behaviors were found to be stron-ger for children who were reared in a positive parenting context

Associations between food parenting and general parenting

The more lsquopositiversquo constructs of general parenting (ienurturance and structure) (Sleddens et al 2014) were associatedwith lower use of instrumental and emotional feeding Reversed as-sociations were found for the more firmly controlling parents Thusthose parents tend to give their child more foods in response togood behavior and to soothe their child The other scales exceptpressure to eat were positively related to the authoritative aspectsof parenting For pressure to eat we found positive associations withcontrolling parenting styles This finding is partially supported byprevious studies (Hennessy et al 2010 Hubbs-Tait et al 2008)

Associations between food parenting and child dietary behavior

We confirmed some of the results of previous studies in whichnondirective child-centered food practices were related to consum-ing healthier diets (Murashima et al 2012 Vereecken et al 2010)and parental reward of food was related to unhealthy food (Kroumllleramp Warschburger 2009 Sleddens et al 2010) and soft drink intake(Kroumlller amp Warschburger 2009) Parental restriction was negative-ly related to sugar-sweetened beverage intake and positively relatedto water intake in contrast to previous studies that showed that high

levels of restriction can lead to increases in calorie intake and likingfor the restricted food (eg Fisher amp Birch 1999 Jansen Mulkensamp Jansen 2007) However the latter types of studies were mainlyexperimental looking at what the child eats when the parent is notaround with free access to food In line with a previous study withinthe KOALA birth cohort by Gubbels et al (2009) we failed to confirmthe paradoxical effect of increased intake of unhealthy foods fol-lowing parental restriction although we should be cautious in in-terpreting these findings as parent-reported child dietary intake onlyaccounts for the dietary behaviors that parents see and are awareof Coercing children to eat was associated with unhealthy eatingHigher levels of pressure in child feeding could have detrimentaleffects on childrenrsquos change in healthy dietary behavior as chil-dren are focused away from internal cues to hunger and satiety(Francis Hofer amp Birch 2001) leading to decreases in preferenceand intake of the healthy foods and subsequent increases in con-sumption of unhealthy foods

General parenting constructs modifying the food parentingndashchilddietary behavior relationship

Moderating effects of general parenting on the relationshipbetween food parenting practices and child eating behavior werefound For children who were reared in a positive parenting contextencouragement and covert control were found to work better (ieincreased healthy intakedecreased unhealthy intake) However someconflicting findings were found The one concept standing out mostwas pressure to eat which was related to sugar-sweetened bever-age intake for children of parents scoring high on behavioral controlPressure to eat often occurs when parents feel that their child iseating insufficient amounts of food or in response to their childrsquosunderweight (Francis et al 2001) Secondary analyses confirmedthis assumption as we found that parents of children who were un-derweight scored significantly higher on pressure to eat com-pared with parents of children who were overweight or obese (data

Table 3Interaction terms and stratified analyses food parentingndashchild dietary behavior relationship

Dietary behavior Year Interaction term(general parenting times food parenting)

Groups

Low Intermediate High

Beta P Beta P Beta P Beta P

Snacking 6ndash8 Behavioral control times Encouragement minus62 034 02 624 00 994 minus11 010Structure times Pressure to eat 53 045 03 533 minus02 641 08 053

8 Overprotection times Covert control minus27 099 minus14 005 minus21 000 minus19 000Sugar-sweetened beverage 6ndash8 Behavioral control times Instrumental feeding 77 007 minus04 509 minus11 048 12 036

Behavioral control times Encouragement minus91 009 03 539 01 788 minus14 0058 Behavioral control times Pressure to eat 82 018 03 578 15 003 13 021

Behavioral control times Encouragement minus116 005 07 218 minus01 899 minus08 135Behavioral control times Control 96 054 minus07 221 minus07 149 minus02 688

Fruit 6ndash8 Nurturance times Monitoring 89 034 minus05 344 05 308 07 081Behavioral control times Instrumental feeding 52 049 minus16 003 minus08 080 minus01 785

8 Behavioral control times Monitoring 125 004 minus04 444 03 553 08 169Behavioral control times Emotional feeding 66 027 minus05 439 07 220 09 163Behavioral control times Restriction minus69 086 15 012 10 064 minus05 437Overprotection times Instrumental feeding 85 031 minus02 800 minus13 028 minus14 043Overprotection times Encouragement minus52 091 14 011 06 231 06 312

Water 6ndash8 Nurturance times Instrumental feeding minus70 024 01 874 08 067 minus04 467Overprotection times Monitoring minus41 070 03 536 00 977 minus04 445

8 Nurturance times Instrumental feeding minus110 004 06 416 02 709 minus15 013Coercive control times Encouragement minus55 047 09 112 07 157 minus01 881Coercive control times Restriction 47 083 05 459 10 047 16 013Overprotection times Covert Control minus34 062 17 003 07 170 09 184Overprotection times Pressure to eat 53 014 minus06 320 minus09 091 02 712

Note Interactions for which the association between the food parenting practice and the outcome was statistically significant in one or more groups of the moderator vari-able Beta standardized regression coefficient from linear regression analysis stratified for the moderator (ie low intermediate high) model adjusted for recruitment groupparental educational level country of birth maternal age at birth child gender and age child BMI z-score at age 6 and all other food parenting practices (note that therewere two different models (1) dependent variable is additionally adjusted for baseline child dietary behavior at age 6 and adjusted for child BMI z-score at age 6 (2) de-pendent variable is child dietary behavior at age 8 adjusted for child BMI z-score at age 6

47EFC Sleddens et alAppetite 79 (2014) 42ndash50

nsns

nsns

ns ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

ASnacking Sugar sweetened

beverageFruit

Behavioral control OverprotectionBehavioral control

6-8 yr 6-8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 21 Food parenting practice lsquoencouragementrsquo

nsns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Snacking Water

OverprotectionOverprotection

8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 22 Food parenting practice lsquocovert controlrsquo

ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Fruit Water

Behavioral control Coercive control

8 yr 6-8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 23 Food parenting practice lsquorestrictionrsquo

ns

ns

ns ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidrad nats(no itaicoss

A

Sugar sweetened beverage

Fruit WaterFruit

Behavioral control Behavioral control Overprotection Nurturance

6-8 yr 6-8 yr 8-yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 24 Food parenting practice lsquoinstrumental feedingrsquo

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Sugar sweetened beverage

Behavioral control

8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 25 Food parenting practice lsquopressure to eatrsquo

Fig 2 Stratified analyses the food parentingndashchild dietary behavior relationship moderated by general parenting Note Red bars represent significant associations betweenfood parenting and undesirable behavior (ie increased unhealthy intakedecreased healthy intake) green bars represent significant associations between food parentingand desirable behavior (ie increased healthy intakedecreased unhealthy intake) white represent nonsignificant associations between food parenting and child dietaryintake bars within each cluster represent the stratified levels of the moderator (iegeneral parenting constructs) p le 05 p le 01 p le 001

48 EFC Sleddens et alAppetite 79 (2014) 42ndash50

not reported) Low weight status has been proven to be related withmore food avoidant behaviors (Jansen et al 2012 Sleddens Kremersamp Thijs 2008) As a result parents could apply coercing food par-enting practices trying to increase their childrsquos calorie intake andprobably also increasing their childrsquos unhealthy dietary behavior (iesnacking and sugar-sweetened beverage intake) Our moderationresults should be interpreted with caution and more studies areneeded before drawing definite conclusions

Study strengths and limitations

This study benefitted from a longitudinal design with mea-sures of child dietary behavior repeated at age 6 and 8 The in-cluded variables (food parenting practices and general parenting)were measured with instruments validated for the Dutch context(Bogers et al 2004 Brants et al 2006 Dutman et al 2010 Sleddenset al 2010 2014) Moderation analysis was possible thanks to alarge sample size and sufficient diversity within the study Weconfirmed the operation of higher-order moderation processesimplying that parenting factors at higher more distal levels alterthe impact of food parenting practices at more proximal levels Werecommend future studies to include the theory-based examina-tion of possible moderation effects and ensure sufficient study sizeto do so or examine specific hypotheses in smaller scale studies withcareful selection of the contextual situation

There were also some limitations that need special attentionHighly educated parents and parents with an lsquoalternativersquo lifestylewere overrepresented in our sample partially due to the choice ofrecruitment methods (Gubbels et al 2009) although mean valuesfor the parenting variables (ie general parenting and food parent-ing) were comparable with previous studies (eg Sleddens et al2010 2014) We therefore adjusted all analyses for highest educa-tion level attained and recruitment channel Additionally all of themeasured variables were reported by parents This may evoke socialdesirability The associations between food parenting practices andchildrenrsquos dietary behavior should therefore be interpreted withcaution Cronbachrsquos alpha for the pressure to eat scale of the CFQwas quite low although this finding is in accordance with some pre-vious studies using this scale (eg Boles et al 2010 Gregory Paxtonamp Brozovic 2010) In spite of the longitudinal analysis causality isdifficult to establish since part of the associations (and interac-tions) may be modified by parental adaptations to unwanted be-havior Snacking and intake of sugar-sweetened drinks were studiedas unhealthy behaviors and fruit and water intake as healthy be-haviors Of course other behaviors such as breakfast and vegeta-ble consumption are important as well to determine dietary intakeFinally any choice of single food groups as healthy or unhealthy isdeemed to be debatable due to complexities such as substitution(eg between fruit natural fruit drinks sweetened fruit drinks andsoft drinks tea water and milk drinks) and the ambiguities of re-lations with specific health indicators such as nutritional imbal-ances dental health and overweight development Further studiesare needed with specific health outcomes to evaluate whether mod-eration by contextual factors as shown in our study with health be-haviors as outcome also translate to health outcomes such asoverweight development

Conclusion

Our results show that food parenting practices are important de-terminants in explaining child dietary behavior and that general par-enting behaviors moderate this association Future research effortsshould continue to focus on testing the influence of factors (eggeneral parenting but also child characteristics such as tempera-ment and eating style) impacting on the food parentingndashchild eatingandor weight relationship in order to gain insights into relevant con-

textual factors that need to be taken into account in designinginterventions

References

Birch L L Fisher J O Grimm-Thomas K Markey C N Sawyer R amp JohnsonS L (2001) Confirmatory factor analysis of the Child Feeding Questionnaire Ameasure of parental attitudes beliefs and practices about child feeding andobesity proneness Appetite 36 201ndash210

Blissett J amp Haycraft E (2008) Are parenting style and controlling feeding practicesrelated Appetite 50 477ndash485

Blissett J Meyer C amp Haycraft E (2011) The role of parenting in the relationshipbetween childhood eating problems and broader behaviour problems Child CareHealth and Development 37 642ndash648

Bogers R P Van Assema P Kester A D M Westerterp K R amp Dagnelie P C (2004)Reproducibility validity and responsiveness to change of a short questionnairefor measuring fruit and vegetable intake American Journal of Epidemiology 159900ndash909

Boles R E Nelson T D Chamberlin L A Valenzuela J M Sherman S N JohnsonS L et al (2010) Confirmatory factor analysis of the Child Feeding Questionnaireamong low-income African American families of preschool children Appetite 54402ndash405

Brants H Stafleu A Ter Doest D Hulshof K amp Thijs C (2006) Ontwikkeling vaneen voedselfrequentievragenlijst Energie-inneming van kinderen van 2 tot enmet 12 jaar [Development of a food frequency questionnaire Energy intake ofchildren 2-12 years of age] Voeding Nu 2 25ndash28

Brown K A Ogden J Voumlgele C amp Gibson E L (2008) The role of parental controlpractices in explaining childrenrsquos diet and BMI Appetite 50 252ndash259

Cislak A Safron M Pratt M Gaspar T amp Luszczynska A (2012) Family-relatedpredictors of body weight and weight-related behaviours among children andadolescents A systematic umbrella review Child Care Health and Development38 321ndash331

Clark H R Goyder E Bissell P Blank L amp Peters J (2007) How do parentsrsquochild-feeding behaviours influence child weight Implications for childhoodobesity policy Journal of Public Health 29 132ndash141

Daniels M C amp Popkin B M (2010) Impact of water intake on energy intake andweight status A systematic review Nutrition Reviews 68 505ndash521

Daniels S R (2009) Complications of obesity in children and adolescentsInternational Journal of Obesity 33(s1) 60ndash65

Darling N amp Steinberg L (1993) Parenting style as context An integrative modelPsychological Bulletin 113 487ndash496

Dutman A E Stafleu A Kruizinga A Brants H A M Westerterp K R KistemakerC et al (2010) Validation of an FFQ and options for data processing using thedoubly labelled water method in children Public Health Nutrition 14 410ndash417

Eurostat (2007) Task force on core social variables Final report Luxembourg EuropeanCommunitees

Faith M S Scanlon K S Birch L L Francis L A amp Sherry B (2004) Parent-childfeeding strategies and their relationships to child eating and weight status ObesityResearch 12 1711ndash1722

Fisher J O amp Birch L L (1999) Restricting access to foods and childrenrsquos eatingAppetite 32 405ndash419

Francis L A Hofer S M amp Birch L L (2001) Predictors of maternal child-feedingstyle Maternal and child characteristics Appetite 37 231ndash243

Fuemmeler B F Yang C Costanzo P Hoyle R H Siegler I C Williams R B et al(2012) Parenting styles and body mass index trajectories from adolescence toadulthood Health Psychology 31 441ndash449

Gerards S M P L Sleddens E F C Dagnelie P C De Vries N K amp Kremers S PJ (2011) Interventions addressing general parenting to prevent or treat childhoodobesity International Journal of Pediatric Obesity 6 e28ndashe45

Golley R K Hendrie G A Slater A amp Corsini N (2011) Interventions thatinvolve parents to improve childrenrsquos weight-related nutrition intake andactivity patterns What nutrition and activity targets and behaviour changetechniques are associated with intervention effectiveness Obesity Reviews 12114ndash130

Gregory J E Paxton S J amp Brozovic A M (2010) Maternal feeding practices childeating behaviour and body mass index in preschool-aged children A prospectiveanalysis International Journal of Behavioral Nutrition and Physical Activity 7 55

Gubbels J S Kremers S P J Stafleu A Dagnelie P C Goldbohm R A De VriesN K et al (2009) Diet-related restrictive parenting practices Impact on dietaryintake of 2-year-old children and interactions with child characteristics Appetite52 423ndash429

Hennessy E Hughes S O Goldberg J P Hyatt R R amp Economos C D (2010) Parentbehavior and child weight status among a diverse group of underserved ruralfamilies Appetite 54 369ndash377

Hingle M D OrsquoConnor T M Dave J M amp Baranowski T (2010) Parentalinvolvement in interventions to improve child dietary intake A systematic reviewPreventive Medicine 51 103ndash111

Hu F B (2013) Resolved There is sufficient scientific evidence that decreasingsugar-sweetened beverage consumption will reduce the prevalence of obesityand obesity-related diseases Obesity Reviews 14 606ndash619

Hubbs-Tait L Kennedy T S Page M C Topham G L amp Harrist A W(2008) Parental feeding practices predict authoritative authoritarian andpermissive parenting styles Journal of the American Dietetic Association 1081154ndash1161

49EFC Sleddens et alAppetite 79 (2014) 42ndash50

following snacks in between meals (1) fried snacks (2) potato crispssalted biscuits and peanuts (3) cake or large biscuits and (4) piepastry candy bars and chocolates A single score was calculated forthe number of snacking occasions (between meals) by adding re-ported frequency (in days a week) of the different snacks (ie range0ndash28)

Both childrenrsquos sugar-sweetened soft drink consumption andwater consumption were assessed with two items Parents wereasked to indicate on how many days a week their child consumedthese drinks Additionally parents were asked to indicate the numberof glasses their child consumed of these drinks on such a day Oneglass equals 250 ml Frequency and amount of soft drinks and waterconsumed were multiplied to obtain an average score of glasses ofsoft drinks and water consumed a week

Fruit consumption was assessed by asking parents to indicate howmany days a week (normal week) their child consumed fruit Ad-ditionally parents were asked to indicate the number of servingstheir child consumed on such a day corresponding with earlier val-idation studies (Bogers Van Assema Kester Westerterp amp Dagnelie2004) One apple or one pear counted as one serving two manda-rins counted as one serving and one bunch of grapes counted asone serving as well Multiplying frequency with the reported usualamount computed an average score for the number of fruit serv-ings consumed per week

Parental background characteristicsFor educational level seven categories were distinguished which

were recoded into three levels (low medium and high) in line withinternational classification systems (Eurostat 2007) Country ofbirth was recoded into lsquoNetherlandsrsquo versus lsquoother countryrsquo In ad-dition maternal age at birth of the child and recruitment group (con-ventional versus lsquoalternativersquo lifestyle) was used in the currentanalyses

Data analyses

Partial correlations were computed for a general assessment ofthe correlations between food parenting practices and the poten-tial moderating factors (ie five general parenting constructs) ad-justed for several covariates listed at the end of this paragraphThereafter separate linear regression analyses were performed toexamine the relationship between food parenting practices and childdietary behavior (ie snacking sugar-sweetened beverage intakefruit intake and water consumption) For each of the four diet-related outcomes the analyses were performed twice using differ-ent dependent variables one predicting dietary behavior at age 8the other predicting dietary behavior change from age 6 to 8 (ad-justing dietary behavior at age 8 for dietary behavior at age 6) Allfood parenting practices were entered simultaneously correctingfor potential confounding by the other variables Finally in orderto examine whether general parenting moderated the association

between food parenting practices and childrenrsquos dietary behaviorwe calculated interaction terms between the food parenting andgeneral parenting variables The interaction terms were added tofour different regression models (ie in each model one of the fourdietary behaviors was set as the dependent variable) in the fourthstep using a forward entering procedure The threshold for includ-ing interaction terms in the stepwise regression was set at p lt 10suggested by Stone-Romero and Liakhovitski (2002) For each of thethree models we added all main effects of the potential moderat-ing variables together in the third step Stratified linear regressionanalyses were performed in order to examine the associationbetween food parenting practices and child dietary behavior in thedifferent strata of the moderator variables (ie general parenting)For each of the moderators three groups were created based on halfa standard deviation from the mean score (less than minus5 SD minus5 SDto 5 SD and higher than 5 SD from the mean) to obtain roughlysimilar group sizes All analyses were adjusted for the influence ofseveral potential covariates recruitment group (conventional versuslsquoalternativersquo lifestyle) educational level (low medium and highhighest educational level attained within a family) country of birthof mother and father (Netherlands versus other) maternal age atbirth and gender and age of the child The hierarchical multiple re-gression analyses were additionally controlled for child baseline BMIz-score (at age 6) because parentsrsquo food parenting practices maydepend on to their weight perceptions of their child (eg WebberCooke Hill amp Wardle 2010)

Results

Of the 1654 children eligible for the current study 513 were boysand 487 were girls The majority of the mothers (968) and fathers(958) were born in the Netherlands Moreover most of the fami-lies were characterized by a conventional lifestyle (813) com-pared with an lsquoalternativersquo lifestyle (187) Educational level washigh for 666 of the families medium for 287 and low for 48Average maternal age at the time of their child birth was 3231 years(SD = 372)

Children consumed snacks 493 (SD = 283) times a week at age8 Childrenrsquos mean intake of sugar-sweetened beverage and waterat age 8 was 220 (SD = 426) and 975 (SD = 860) glasses per weekrespectively The mean number of servings of fruit per week at age8 was 775 (SD = 432)

Associations between general parenting and food parenting(A in Fig 1)

General parenting practices correlated with most of the food par-enting practices although modestly at most (r = 24 or lower Table 1)Nurturing and structuring parents applied less instrumental andemotional food parenting practices and more encouragement andcontrolling practices Parents scoring high on behavioral control more

Table 1Partial correlations between food parenting practices and general parenting

Mean (SD) Food parenting practices

Instrumentalfeeding

Emotionalfeeding

Control Encouragement Covertcontrol

Pressureto eat

Restriction Monitoring

General parentingNurturance 450 (32) minus16 minus11 15 31 08 02 18 18Structure 408 (37) minus21 minus19 24 23 08 01 18 18Behavioral control 401 (39) 06 02 19 17 07 21 19 19Coercive control 195 (41) 23 17 05 minus09 02 18 minus04 02Overprotection 269 (52) 02 10 03 07 17 15 04 14

Note n = 1624 Partial rank correlation coefficients adjusted for recruitment group parental educational level country of birth maternal age at birth and child gender andage CGPQCFQ answering scale 1ndash5 p le 05 p le 01 p le 001

45EFC Sleddens et alAppetite 79 (2014) 42ndash50

often used encouraging and controlling food parenting practices Co-ercive control was positively related to instrumental and emotion-al feeding and overprotection was related to emotional feedingOverprotective parents more often indicated to keep foods out ofreach of their children (covert control) The controlling general par-enting constructs (ie behavioral control coercive control and over-protection) were positively related to pressure to eat

Associations between food parenting and child dietary behavior(B in Fig 1)

We examined the extent to which food parenting practices predictchild dietary behavior The strength of associations was generallylow as indicated by the size of the standardized regression coeffi-cients between minus19 and 13 (see Table 2) The associations betweenfood parenting practices and childrsquos dietary behavior at age 8 weremostly attenuated by controlling for baseline dietary behavior andBMI z-score at age 6 (compare models 1 and 2 Table 2) Instrumen-tal feeding at age 6 was related to less fruit intake at age 8 This as-sociation was also present with fruit intake change from age 6 to 8as an outcome Emotional feeding was related to increased childsnack intake both cross-sectionally and longitudinally Encourage-ment was positively related to fruit intake when not controlling forbaseline fruit intake and BMI z-score Pressure to eat was positive-ly associated with unhealthy behavior (ie snacking and consump-tion of sugar-sweetened beverages) Monitoring was not significantlyrelated to dietary behavior The results for the controlling food par-enting practices (eg PFSQ control CFQ restriction and covertcontrol) were mixed Control as measured by the PFSQ was not as-sociated with any of the child eating outcomes Parental restric-tion had a desirable effect on the consumption of sugar-sweeteneddrinks and water (leading to a lower and higher intake respective-ly) Parents who kept unhealthy foods out of reach from their chil-dren (covert control) were more likely to have children who areeating healthily (ie less snacking and sugar-sweetened beverageintake and more fruit and water intake)

General parenting constructs moderating the food parentingndashchilddietary behavior relationship (C in Fig 1)

We explored the moderating role of general parenting on the re-lationship between food parenting practices and childrenrsquos dietarybehavior In Table 3 we provide all significant interaction terms

between the general parenting and food parenting practices in ex-plaining child dietary behavior at age 8 and change from age 6 to8 and the results of the stratified analyses In Fig 2 we graphical-ly display these stratified analyses only for associations between thefood parenting practices and the outcomes that were statisticallysignificant in one or more of the strata of the moderator variable(12 graphs in total)

In total 22 interactions were found between food parenting prac-tices and general parenting in predicting child dietary behavior Gen-erally we found that the parenting practices of encouragement(Fig 21) and covert control (Fig 22) had more desirable effects whenparents scored higher on positive (ie behavioral control) and loweron negative (ie overprotection) general parenting constructs re-spectively One exception for all levels of overprotection the as-sociation between covert controlling parenting and child snackingwas desirable For restriction (Fig 23) more desirable effects werefound when parents scored low on behavioral control (more fruitintake) and high on coercive control (more water intake) Instru-mental feeding (Fig 24) had stronger relations with undesirable childdietary behavior (ie less fruit intake) in less controlled and over-protective home environments When parents scored intermedi-ate on behavioral control instrumental feeding was related to lowersugar-sweetened beverage intake whereas when parents scored highon behavioral control instrumental feeding was related to highersugar-sweetened beverage intake Additionally the negative rela-tion between instrumental feeding and water consumption wasstronger among children of parents who scored high on nurturanceLastly pressure to eat (Fig 25) indicated to have an undesirable effecton child dietary behavior (ie more sugar-sweetened beverageintake) among children of parents scoring intermediate and highon behavioral control For the remaining 10 interactions the strati-fied analyses were not significant so they are not considered further

Discussion

Many studies on food parenting fail to consider the larger familycontext The present study provides evidence for effective and in-effective food parenting practices as well as for child- and parent-related contexts that moderate their impact The relatively weakmain effects could be summarized as follows instrumental andemotional feeding and pressure to eat were found to have detri-mental associations with child dietary behavior (ie increased un-healthy intakedecreased healthy intake) whereas covert control

Table 2Associations between food parenting practices and child dietary behavior

Mean (SD) Child dietary behavior

Snacking Sugar-sweetened drinks Fruit Water

Model 1n = 1390Beta

Model 2n = 1378Beta

Model 1n = 1159Beta

Model 2n = 1044Beta

Model 1n = 1159Beta

Model 2n = 1078Beta

Model 1n = 1108Beta

Model 2n = 982Beta

BMI z-score minus05 minus05 03 00 04 01 07 04Baseline eating minus minus 54 minus 49 minus 61 minus 64Food parenting

Instrumental feeding 172 (57) 01 minus01 minus01 minus01 minus09 minus09 minus03 03Emotional feeding 142 (46) 13 07 02 minus01 04 05 minus05 minus05Control 436 (36) minus03 01 minus05 minus03 minus02 minus01 minus02 00Encouragement 388 (47) minus02 minus03 minus01 minus03 08 05 06 minus02Covert control 293 (93) minus19 minus09 minus08 minus02 09 04 12 04Pressure to eat 351 (75) 07 03 11 07 minus05 00 minus05 minus01Restriction 452 (62) minus04 00 minus13 minus07 06 02 10 04Monitoring 431 (56) minus05 minus04 04 05 03 03 minus02 00

Note Beta standardized linear regression coefficient Model 1 dependent variable is child dietary behavior around the age of 8 years which is additionally adjusted for childBMI z-score at age 6 Model 2 dependent variable is additionally adjusted for baseline child dietary behavior (snacking or sugar-sweetened drink intake or fruit intake orwater intake) at age 6 and child BMI z-score at age 6 both models adjusted for recruitment group parental educational level country of birth maternal age at birth andchild gender and age food parenting practices answering scale 1ndash5 p le 05 p le 01 p le 001

46 EFC Sleddens et alAppetite 79 (2014) 42ndash50

encouragement and restriction were found to have favorable asso-ciations with child dietary behavior (ie increased healthy intakedecreased unhealthy intake) This is confirming our hypotheses andmost of these relationships have been demonstrated previously (egBrown Ogden Voumlgele amp Gibson 2008 Gubbels et al 2009Murashima Hoerr Hughes amp Kaplowitz 2012 Rodenburg et al2013 Sleddens et al 2010 Van der Horst et al 2007) These as-sociations were stronger for some subgroups of the moderating vari-ables as expected The associations of encouragement and covertcontrol with desirable child dietary behaviors were found to be stron-ger for children who were reared in a positive parenting context

Associations between food parenting and general parenting

The more lsquopositiversquo constructs of general parenting (ienurturance and structure) (Sleddens et al 2014) were associatedwith lower use of instrumental and emotional feeding Reversed as-sociations were found for the more firmly controlling parents Thusthose parents tend to give their child more foods in response togood behavior and to soothe their child The other scales exceptpressure to eat were positively related to the authoritative aspectsof parenting For pressure to eat we found positive associations withcontrolling parenting styles This finding is partially supported byprevious studies (Hennessy et al 2010 Hubbs-Tait et al 2008)

Associations between food parenting and child dietary behavior

We confirmed some of the results of previous studies in whichnondirective child-centered food practices were related to consum-ing healthier diets (Murashima et al 2012 Vereecken et al 2010)and parental reward of food was related to unhealthy food (Kroumllleramp Warschburger 2009 Sleddens et al 2010) and soft drink intake(Kroumlller amp Warschburger 2009) Parental restriction was negative-ly related to sugar-sweetened beverage intake and positively relatedto water intake in contrast to previous studies that showed that high

levels of restriction can lead to increases in calorie intake and likingfor the restricted food (eg Fisher amp Birch 1999 Jansen Mulkensamp Jansen 2007) However the latter types of studies were mainlyexperimental looking at what the child eats when the parent is notaround with free access to food In line with a previous study withinthe KOALA birth cohort by Gubbels et al (2009) we failed to confirmthe paradoxical effect of increased intake of unhealthy foods fol-lowing parental restriction although we should be cautious in in-terpreting these findings as parent-reported child dietary intake onlyaccounts for the dietary behaviors that parents see and are awareof Coercing children to eat was associated with unhealthy eatingHigher levels of pressure in child feeding could have detrimentaleffects on childrenrsquos change in healthy dietary behavior as chil-dren are focused away from internal cues to hunger and satiety(Francis Hofer amp Birch 2001) leading to decreases in preferenceand intake of the healthy foods and subsequent increases in con-sumption of unhealthy foods

General parenting constructs modifying the food parentingndashchilddietary behavior relationship

Moderating effects of general parenting on the relationshipbetween food parenting practices and child eating behavior werefound For children who were reared in a positive parenting contextencouragement and covert control were found to work better (ieincreased healthy intakedecreased unhealthy intake) However someconflicting findings were found The one concept standing out mostwas pressure to eat which was related to sugar-sweetened bever-age intake for children of parents scoring high on behavioral controlPressure to eat often occurs when parents feel that their child iseating insufficient amounts of food or in response to their childrsquosunderweight (Francis et al 2001) Secondary analyses confirmedthis assumption as we found that parents of children who were un-derweight scored significantly higher on pressure to eat com-pared with parents of children who were overweight or obese (data

Table 3Interaction terms and stratified analyses food parentingndashchild dietary behavior relationship

Dietary behavior Year Interaction term(general parenting times food parenting)

Groups

Low Intermediate High

Beta P Beta P Beta P Beta P

Snacking 6ndash8 Behavioral control times Encouragement minus62 034 02 624 00 994 minus11 010Structure times Pressure to eat 53 045 03 533 minus02 641 08 053

8 Overprotection times Covert control minus27 099 minus14 005 minus21 000 minus19 000Sugar-sweetened beverage 6ndash8 Behavioral control times Instrumental feeding 77 007 minus04 509 minus11 048 12 036

Behavioral control times Encouragement minus91 009 03 539 01 788 minus14 0058 Behavioral control times Pressure to eat 82 018 03 578 15 003 13 021

Behavioral control times Encouragement minus116 005 07 218 minus01 899 minus08 135Behavioral control times Control 96 054 minus07 221 minus07 149 minus02 688

Fruit 6ndash8 Nurturance times Monitoring 89 034 minus05 344 05 308 07 081Behavioral control times Instrumental feeding 52 049 minus16 003 minus08 080 minus01 785

8 Behavioral control times Monitoring 125 004 minus04 444 03 553 08 169Behavioral control times Emotional feeding 66 027 minus05 439 07 220 09 163Behavioral control times Restriction minus69 086 15 012 10 064 minus05 437Overprotection times Instrumental feeding 85 031 minus02 800 minus13 028 minus14 043Overprotection times Encouragement minus52 091 14 011 06 231 06 312

Water 6ndash8 Nurturance times Instrumental feeding minus70 024 01 874 08 067 minus04 467Overprotection times Monitoring minus41 070 03 536 00 977 minus04 445

8 Nurturance times Instrumental feeding minus110 004 06 416 02 709 minus15 013Coercive control times Encouragement minus55 047 09 112 07 157 minus01 881Coercive control times Restriction 47 083 05 459 10 047 16 013Overprotection times Covert Control minus34 062 17 003 07 170 09 184Overprotection times Pressure to eat 53 014 minus06 320 minus09 091 02 712

Note Interactions for which the association between the food parenting practice and the outcome was statistically significant in one or more groups of the moderator vari-able Beta standardized regression coefficient from linear regression analysis stratified for the moderator (ie low intermediate high) model adjusted for recruitment groupparental educational level country of birth maternal age at birth child gender and age child BMI z-score at age 6 and all other food parenting practices (note that therewere two different models (1) dependent variable is additionally adjusted for baseline child dietary behavior at age 6 and adjusted for child BMI z-score at age 6 (2) de-pendent variable is child dietary behavior at age 8 adjusted for child BMI z-score at age 6

47EFC Sleddens et alAppetite 79 (2014) 42ndash50

nsns

nsns

ns ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

ASnacking Sugar sweetened

beverageFruit

Behavioral control OverprotectionBehavioral control

6-8 yr 6-8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 21 Food parenting practice lsquoencouragementrsquo

nsns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Snacking Water

OverprotectionOverprotection

8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 22 Food parenting practice lsquocovert controlrsquo

ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Fruit Water

Behavioral control Coercive control

8 yr 6-8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 23 Food parenting practice lsquorestrictionrsquo

ns

ns

ns ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidrad nats(no itaicoss

A

Sugar sweetened beverage

Fruit WaterFruit

Behavioral control Behavioral control Overprotection Nurturance

6-8 yr 6-8 yr 8-yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 24 Food parenting practice lsquoinstrumental feedingrsquo

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Sugar sweetened beverage

Behavioral control

8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 25 Food parenting practice lsquopressure to eatrsquo

Fig 2 Stratified analyses the food parentingndashchild dietary behavior relationship moderated by general parenting Note Red bars represent significant associations betweenfood parenting and undesirable behavior (ie increased unhealthy intakedecreased healthy intake) green bars represent significant associations between food parentingand desirable behavior (ie increased healthy intakedecreased unhealthy intake) white represent nonsignificant associations between food parenting and child dietaryintake bars within each cluster represent the stratified levels of the moderator (iegeneral parenting constructs) p le 05 p le 01 p le 001

48 EFC Sleddens et alAppetite 79 (2014) 42ndash50

not reported) Low weight status has been proven to be related withmore food avoidant behaviors (Jansen et al 2012 Sleddens Kremersamp Thijs 2008) As a result parents could apply coercing food par-enting practices trying to increase their childrsquos calorie intake andprobably also increasing their childrsquos unhealthy dietary behavior (iesnacking and sugar-sweetened beverage intake) Our moderationresults should be interpreted with caution and more studies areneeded before drawing definite conclusions

Study strengths and limitations

This study benefitted from a longitudinal design with mea-sures of child dietary behavior repeated at age 6 and 8 The in-cluded variables (food parenting practices and general parenting)were measured with instruments validated for the Dutch context(Bogers et al 2004 Brants et al 2006 Dutman et al 2010 Sleddenset al 2010 2014) Moderation analysis was possible thanks to alarge sample size and sufficient diversity within the study Weconfirmed the operation of higher-order moderation processesimplying that parenting factors at higher more distal levels alterthe impact of food parenting practices at more proximal levels Werecommend future studies to include the theory-based examina-tion of possible moderation effects and ensure sufficient study sizeto do so or examine specific hypotheses in smaller scale studies withcareful selection of the contextual situation

There were also some limitations that need special attentionHighly educated parents and parents with an lsquoalternativersquo lifestylewere overrepresented in our sample partially due to the choice ofrecruitment methods (Gubbels et al 2009) although mean valuesfor the parenting variables (ie general parenting and food parent-ing) were comparable with previous studies (eg Sleddens et al2010 2014) We therefore adjusted all analyses for highest educa-tion level attained and recruitment channel Additionally all of themeasured variables were reported by parents This may evoke socialdesirability The associations between food parenting practices andchildrenrsquos dietary behavior should therefore be interpreted withcaution Cronbachrsquos alpha for the pressure to eat scale of the CFQwas quite low although this finding is in accordance with some pre-vious studies using this scale (eg Boles et al 2010 Gregory Paxtonamp Brozovic 2010) In spite of the longitudinal analysis causality isdifficult to establish since part of the associations (and interac-tions) may be modified by parental adaptations to unwanted be-havior Snacking and intake of sugar-sweetened drinks were studiedas unhealthy behaviors and fruit and water intake as healthy be-haviors Of course other behaviors such as breakfast and vegeta-ble consumption are important as well to determine dietary intakeFinally any choice of single food groups as healthy or unhealthy isdeemed to be debatable due to complexities such as substitution(eg between fruit natural fruit drinks sweetened fruit drinks andsoft drinks tea water and milk drinks) and the ambiguities of re-lations with specific health indicators such as nutritional imbal-ances dental health and overweight development Further studiesare needed with specific health outcomes to evaluate whether mod-eration by contextual factors as shown in our study with health be-haviors as outcome also translate to health outcomes such asoverweight development

Conclusion

Our results show that food parenting practices are important de-terminants in explaining child dietary behavior and that general par-enting behaviors moderate this association Future research effortsshould continue to focus on testing the influence of factors (eggeneral parenting but also child characteristics such as tempera-ment and eating style) impacting on the food parentingndashchild eatingandor weight relationship in order to gain insights into relevant con-

textual factors that need to be taken into account in designinginterventions

References

Birch L L Fisher J O Grimm-Thomas K Markey C N Sawyer R amp JohnsonS L (2001) Confirmatory factor analysis of the Child Feeding Questionnaire Ameasure of parental attitudes beliefs and practices about child feeding andobesity proneness Appetite 36 201ndash210

Blissett J amp Haycraft E (2008) Are parenting style and controlling feeding practicesrelated Appetite 50 477ndash485

Blissett J Meyer C amp Haycraft E (2011) The role of parenting in the relationshipbetween childhood eating problems and broader behaviour problems Child CareHealth and Development 37 642ndash648

Bogers R P Van Assema P Kester A D M Westerterp K R amp Dagnelie P C (2004)Reproducibility validity and responsiveness to change of a short questionnairefor measuring fruit and vegetable intake American Journal of Epidemiology 159900ndash909

Boles R E Nelson T D Chamberlin L A Valenzuela J M Sherman S N JohnsonS L et al (2010) Confirmatory factor analysis of the Child Feeding Questionnaireamong low-income African American families of preschool children Appetite 54402ndash405

Brants H Stafleu A Ter Doest D Hulshof K amp Thijs C (2006) Ontwikkeling vaneen voedselfrequentievragenlijst Energie-inneming van kinderen van 2 tot enmet 12 jaar [Development of a food frequency questionnaire Energy intake ofchildren 2-12 years of age] Voeding Nu 2 25ndash28

Brown K A Ogden J Voumlgele C amp Gibson E L (2008) The role of parental controlpractices in explaining childrenrsquos diet and BMI Appetite 50 252ndash259

Cislak A Safron M Pratt M Gaspar T amp Luszczynska A (2012) Family-relatedpredictors of body weight and weight-related behaviours among children andadolescents A systematic umbrella review Child Care Health and Development38 321ndash331

Clark H R Goyder E Bissell P Blank L amp Peters J (2007) How do parentsrsquochild-feeding behaviours influence child weight Implications for childhoodobesity policy Journal of Public Health 29 132ndash141

Daniels M C amp Popkin B M (2010) Impact of water intake on energy intake andweight status A systematic review Nutrition Reviews 68 505ndash521

Daniels S R (2009) Complications of obesity in children and adolescentsInternational Journal of Obesity 33(s1) 60ndash65

Darling N amp Steinberg L (1993) Parenting style as context An integrative modelPsychological Bulletin 113 487ndash496

Dutman A E Stafleu A Kruizinga A Brants H A M Westerterp K R KistemakerC et al (2010) Validation of an FFQ and options for data processing using thedoubly labelled water method in children Public Health Nutrition 14 410ndash417

Eurostat (2007) Task force on core social variables Final report Luxembourg EuropeanCommunitees

Faith M S Scanlon K S Birch L L Francis L A amp Sherry B (2004) Parent-childfeeding strategies and their relationships to child eating and weight status ObesityResearch 12 1711ndash1722

Fisher J O amp Birch L L (1999) Restricting access to foods and childrenrsquos eatingAppetite 32 405ndash419

Francis L A Hofer S M amp Birch L L (2001) Predictors of maternal child-feedingstyle Maternal and child characteristics Appetite 37 231ndash243

Fuemmeler B F Yang C Costanzo P Hoyle R H Siegler I C Williams R B et al(2012) Parenting styles and body mass index trajectories from adolescence toadulthood Health Psychology 31 441ndash449

Gerards S M P L Sleddens E F C Dagnelie P C De Vries N K amp Kremers S PJ (2011) Interventions addressing general parenting to prevent or treat childhoodobesity International Journal of Pediatric Obesity 6 e28ndashe45

Golley R K Hendrie G A Slater A amp Corsini N (2011) Interventions thatinvolve parents to improve childrenrsquos weight-related nutrition intake andactivity patterns What nutrition and activity targets and behaviour changetechniques are associated with intervention effectiveness Obesity Reviews 12114ndash130

Gregory J E Paxton S J amp Brozovic A M (2010) Maternal feeding practices childeating behaviour and body mass index in preschool-aged children A prospectiveanalysis International Journal of Behavioral Nutrition and Physical Activity 7 55

Gubbels J S Kremers S P J Stafleu A Dagnelie P C Goldbohm R A De VriesN K et al (2009) Diet-related restrictive parenting practices Impact on dietaryintake of 2-year-old children and interactions with child characteristics Appetite52 423ndash429

Hennessy E Hughes S O Goldberg J P Hyatt R R amp Economos C D (2010) Parentbehavior and child weight status among a diverse group of underserved ruralfamilies Appetite 54 369ndash377

Hingle M D OrsquoConnor T M Dave J M amp Baranowski T (2010) Parentalinvolvement in interventions to improve child dietary intake A systematic reviewPreventive Medicine 51 103ndash111

Hu F B (2013) Resolved There is sufficient scientific evidence that decreasingsugar-sweetened beverage consumption will reduce the prevalence of obesityand obesity-related diseases Obesity Reviews 14 606ndash619

Hubbs-Tait L Kennedy T S Page M C Topham G L amp Harrist A W(2008) Parental feeding practices predict authoritative authoritarian andpermissive parenting styles Journal of the American Dietetic Association 1081154ndash1161

49EFC Sleddens et alAppetite 79 (2014) 42ndash50

often used encouraging and controlling food parenting practices Co-ercive control was positively related to instrumental and emotion-al feeding and overprotection was related to emotional feedingOverprotective parents more often indicated to keep foods out ofreach of their children (covert control) The controlling general par-enting constructs (ie behavioral control coercive control and over-protection) were positively related to pressure to eat

Associations between food parenting and child dietary behavior(B in Fig 1)

We examined the extent to which food parenting practices predictchild dietary behavior The strength of associations was generallylow as indicated by the size of the standardized regression coeffi-cients between minus19 and 13 (see Table 2) The associations betweenfood parenting practices and childrsquos dietary behavior at age 8 weremostly attenuated by controlling for baseline dietary behavior andBMI z-score at age 6 (compare models 1 and 2 Table 2) Instrumen-tal feeding at age 6 was related to less fruit intake at age 8 This as-sociation was also present with fruit intake change from age 6 to 8as an outcome Emotional feeding was related to increased childsnack intake both cross-sectionally and longitudinally Encourage-ment was positively related to fruit intake when not controlling forbaseline fruit intake and BMI z-score Pressure to eat was positive-ly associated with unhealthy behavior (ie snacking and consump-tion of sugar-sweetened beverages) Monitoring was not significantlyrelated to dietary behavior The results for the controlling food par-enting practices (eg PFSQ control CFQ restriction and covertcontrol) were mixed Control as measured by the PFSQ was not as-sociated with any of the child eating outcomes Parental restric-tion had a desirable effect on the consumption of sugar-sweeteneddrinks and water (leading to a lower and higher intake respective-ly) Parents who kept unhealthy foods out of reach from their chil-dren (covert control) were more likely to have children who areeating healthily (ie less snacking and sugar-sweetened beverageintake and more fruit and water intake)

General parenting constructs moderating the food parentingndashchilddietary behavior relationship (C in Fig 1)

We explored the moderating role of general parenting on the re-lationship between food parenting practices and childrenrsquos dietarybehavior In Table 3 we provide all significant interaction terms

between the general parenting and food parenting practices in ex-plaining child dietary behavior at age 8 and change from age 6 to8 and the results of the stratified analyses In Fig 2 we graphical-ly display these stratified analyses only for associations between thefood parenting practices and the outcomes that were statisticallysignificant in one or more of the strata of the moderator variable(12 graphs in total)

In total 22 interactions were found between food parenting prac-tices and general parenting in predicting child dietary behavior Gen-erally we found that the parenting practices of encouragement(Fig 21) and covert control (Fig 22) had more desirable effects whenparents scored higher on positive (ie behavioral control) and loweron negative (ie overprotection) general parenting constructs re-spectively One exception for all levels of overprotection the as-sociation between covert controlling parenting and child snackingwas desirable For restriction (Fig 23) more desirable effects werefound when parents scored low on behavioral control (more fruitintake) and high on coercive control (more water intake) Instru-mental feeding (Fig 24) had stronger relations with undesirable childdietary behavior (ie less fruit intake) in less controlled and over-protective home environments When parents scored intermedi-ate on behavioral control instrumental feeding was related to lowersugar-sweetened beverage intake whereas when parents scored highon behavioral control instrumental feeding was related to highersugar-sweetened beverage intake Additionally the negative rela-tion between instrumental feeding and water consumption wasstronger among children of parents who scored high on nurturanceLastly pressure to eat (Fig 25) indicated to have an undesirable effecton child dietary behavior (ie more sugar-sweetened beverageintake) among children of parents scoring intermediate and highon behavioral control For the remaining 10 interactions the strati-fied analyses were not significant so they are not considered further

Discussion

Many studies on food parenting fail to consider the larger familycontext The present study provides evidence for effective and in-effective food parenting practices as well as for child- and parent-related contexts that moderate their impact The relatively weakmain effects could be summarized as follows instrumental andemotional feeding and pressure to eat were found to have detri-mental associations with child dietary behavior (ie increased un-healthy intakedecreased healthy intake) whereas covert control

Table 2Associations between food parenting practices and child dietary behavior

Mean (SD) Child dietary behavior

Snacking Sugar-sweetened drinks Fruit Water

Model 1n = 1390Beta

Model 2n = 1378Beta

Model 1n = 1159Beta

Model 2n = 1044Beta

Model 1n = 1159Beta

Model 2n = 1078Beta

Model 1n = 1108Beta

Model 2n = 982Beta

BMI z-score minus05 minus05 03 00 04 01 07 04Baseline eating minus minus 54 minus 49 minus 61 minus 64Food parenting

Instrumental feeding 172 (57) 01 minus01 minus01 minus01 minus09 minus09 minus03 03Emotional feeding 142 (46) 13 07 02 minus01 04 05 minus05 minus05Control 436 (36) minus03 01 minus05 minus03 minus02 minus01 minus02 00Encouragement 388 (47) minus02 minus03 minus01 minus03 08 05 06 minus02Covert control 293 (93) minus19 minus09 minus08 minus02 09 04 12 04Pressure to eat 351 (75) 07 03 11 07 minus05 00 minus05 minus01Restriction 452 (62) minus04 00 minus13 minus07 06 02 10 04Monitoring 431 (56) minus05 minus04 04 05 03 03 minus02 00

Note Beta standardized linear regression coefficient Model 1 dependent variable is child dietary behavior around the age of 8 years which is additionally adjusted for childBMI z-score at age 6 Model 2 dependent variable is additionally adjusted for baseline child dietary behavior (snacking or sugar-sweetened drink intake or fruit intake orwater intake) at age 6 and child BMI z-score at age 6 both models adjusted for recruitment group parental educational level country of birth maternal age at birth andchild gender and age food parenting practices answering scale 1ndash5 p le 05 p le 01 p le 001

46 EFC Sleddens et alAppetite 79 (2014) 42ndash50

encouragement and restriction were found to have favorable asso-ciations with child dietary behavior (ie increased healthy intakedecreased unhealthy intake) This is confirming our hypotheses andmost of these relationships have been demonstrated previously (egBrown Ogden Voumlgele amp Gibson 2008 Gubbels et al 2009Murashima Hoerr Hughes amp Kaplowitz 2012 Rodenburg et al2013 Sleddens et al 2010 Van der Horst et al 2007) These as-sociations were stronger for some subgroups of the moderating vari-ables as expected The associations of encouragement and covertcontrol with desirable child dietary behaviors were found to be stron-ger for children who were reared in a positive parenting context

Associations between food parenting and general parenting

The more lsquopositiversquo constructs of general parenting (ienurturance and structure) (Sleddens et al 2014) were associatedwith lower use of instrumental and emotional feeding Reversed as-sociations were found for the more firmly controlling parents Thusthose parents tend to give their child more foods in response togood behavior and to soothe their child The other scales exceptpressure to eat were positively related to the authoritative aspectsof parenting For pressure to eat we found positive associations withcontrolling parenting styles This finding is partially supported byprevious studies (Hennessy et al 2010 Hubbs-Tait et al 2008)

Associations between food parenting and child dietary behavior

We confirmed some of the results of previous studies in whichnondirective child-centered food practices were related to consum-ing healthier diets (Murashima et al 2012 Vereecken et al 2010)and parental reward of food was related to unhealthy food (Kroumllleramp Warschburger 2009 Sleddens et al 2010) and soft drink intake(Kroumlller amp Warschburger 2009) Parental restriction was negative-ly related to sugar-sweetened beverage intake and positively relatedto water intake in contrast to previous studies that showed that high

levels of restriction can lead to increases in calorie intake and likingfor the restricted food (eg Fisher amp Birch 1999 Jansen Mulkensamp Jansen 2007) However the latter types of studies were mainlyexperimental looking at what the child eats when the parent is notaround with free access to food In line with a previous study withinthe KOALA birth cohort by Gubbels et al (2009) we failed to confirmthe paradoxical effect of increased intake of unhealthy foods fol-lowing parental restriction although we should be cautious in in-terpreting these findings as parent-reported child dietary intake onlyaccounts for the dietary behaviors that parents see and are awareof Coercing children to eat was associated with unhealthy eatingHigher levels of pressure in child feeding could have detrimentaleffects on childrenrsquos change in healthy dietary behavior as chil-dren are focused away from internal cues to hunger and satiety(Francis Hofer amp Birch 2001) leading to decreases in preferenceand intake of the healthy foods and subsequent increases in con-sumption of unhealthy foods

General parenting constructs modifying the food parentingndashchilddietary behavior relationship

Moderating effects of general parenting on the relationshipbetween food parenting practices and child eating behavior werefound For children who were reared in a positive parenting contextencouragement and covert control were found to work better (ieincreased healthy intakedecreased unhealthy intake) However someconflicting findings were found The one concept standing out mostwas pressure to eat which was related to sugar-sweetened bever-age intake for children of parents scoring high on behavioral controlPressure to eat often occurs when parents feel that their child iseating insufficient amounts of food or in response to their childrsquosunderweight (Francis et al 2001) Secondary analyses confirmedthis assumption as we found that parents of children who were un-derweight scored significantly higher on pressure to eat com-pared with parents of children who were overweight or obese (data

Table 3Interaction terms and stratified analyses food parentingndashchild dietary behavior relationship

Dietary behavior Year Interaction term(general parenting times food parenting)

Groups

Low Intermediate High

Beta P Beta P Beta P Beta P

Snacking 6ndash8 Behavioral control times Encouragement minus62 034 02 624 00 994 minus11 010Structure times Pressure to eat 53 045 03 533 minus02 641 08 053

8 Overprotection times Covert control minus27 099 minus14 005 minus21 000 minus19 000Sugar-sweetened beverage 6ndash8 Behavioral control times Instrumental feeding 77 007 minus04 509 minus11 048 12 036

Behavioral control times Encouragement minus91 009 03 539 01 788 minus14 0058 Behavioral control times Pressure to eat 82 018 03 578 15 003 13 021

Behavioral control times Encouragement minus116 005 07 218 minus01 899 minus08 135Behavioral control times Control 96 054 minus07 221 minus07 149 minus02 688

Fruit 6ndash8 Nurturance times Monitoring 89 034 minus05 344 05 308 07 081Behavioral control times Instrumental feeding 52 049 minus16 003 minus08 080 minus01 785

8 Behavioral control times Monitoring 125 004 minus04 444 03 553 08 169Behavioral control times Emotional feeding 66 027 minus05 439 07 220 09 163Behavioral control times Restriction minus69 086 15 012 10 064 minus05 437Overprotection times Instrumental feeding 85 031 minus02 800 minus13 028 minus14 043Overprotection times Encouragement minus52 091 14 011 06 231 06 312

Water 6ndash8 Nurturance times Instrumental feeding minus70 024 01 874 08 067 minus04 467Overprotection times Monitoring minus41 070 03 536 00 977 minus04 445

8 Nurturance times Instrumental feeding minus110 004 06 416 02 709 minus15 013Coercive control times Encouragement minus55 047 09 112 07 157 minus01 881Coercive control times Restriction 47 083 05 459 10 047 16 013Overprotection times Covert Control minus34 062 17 003 07 170 09 184Overprotection times Pressure to eat 53 014 minus06 320 minus09 091 02 712

Note Interactions for which the association between the food parenting practice and the outcome was statistically significant in one or more groups of the moderator vari-able Beta standardized regression coefficient from linear regression analysis stratified for the moderator (ie low intermediate high) model adjusted for recruitment groupparental educational level country of birth maternal age at birth child gender and age child BMI z-score at age 6 and all other food parenting practices (note that therewere two different models (1) dependent variable is additionally adjusted for baseline child dietary behavior at age 6 and adjusted for child BMI z-score at age 6 (2) de-pendent variable is child dietary behavior at age 8 adjusted for child BMI z-score at age 6

47EFC Sleddens et alAppetite 79 (2014) 42ndash50

nsns

nsns

ns ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

ASnacking Sugar sweetened

beverageFruit

Behavioral control OverprotectionBehavioral control

6-8 yr 6-8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 21 Food parenting practice lsquoencouragementrsquo

nsns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Snacking Water

OverprotectionOverprotection

8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 22 Food parenting practice lsquocovert controlrsquo

ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Fruit Water

Behavioral control Coercive control

8 yr 6-8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 23 Food parenting practice lsquorestrictionrsquo

ns

ns

ns ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidrad nats(no itaicoss

A

Sugar sweetened beverage

Fruit WaterFruit

Behavioral control Behavioral control Overprotection Nurturance

6-8 yr 6-8 yr 8-yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 24 Food parenting practice lsquoinstrumental feedingrsquo

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Sugar sweetened beverage

Behavioral control

8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 25 Food parenting practice lsquopressure to eatrsquo

Fig 2 Stratified analyses the food parentingndashchild dietary behavior relationship moderated by general parenting Note Red bars represent significant associations betweenfood parenting and undesirable behavior (ie increased unhealthy intakedecreased healthy intake) green bars represent significant associations between food parentingand desirable behavior (ie increased healthy intakedecreased unhealthy intake) white represent nonsignificant associations between food parenting and child dietaryintake bars within each cluster represent the stratified levels of the moderator (iegeneral parenting constructs) p le 05 p le 01 p le 001

48 EFC Sleddens et alAppetite 79 (2014) 42ndash50

not reported) Low weight status has been proven to be related withmore food avoidant behaviors (Jansen et al 2012 Sleddens Kremersamp Thijs 2008) As a result parents could apply coercing food par-enting practices trying to increase their childrsquos calorie intake andprobably also increasing their childrsquos unhealthy dietary behavior (iesnacking and sugar-sweetened beverage intake) Our moderationresults should be interpreted with caution and more studies areneeded before drawing definite conclusions

Study strengths and limitations

This study benefitted from a longitudinal design with mea-sures of child dietary behavior repeated at age 6 and 8 The in-cluded variables (food parenting practices and general parenting)were measured with instruments validated for the Dutch context(Bogers et al 2004 Brants et al 2006 Dutman et al 2010 Sleddenset al 2010 2014) Moderation analysis was possible thanks to alarge sample size and sufficient diversity within the study Weconfirmed the operation of higher-order moderation processesimplying that parenting factors at higher more distal levels alterthe impact of food parenting practices at more proximal levels Werecommend future studies to include the theory-based examina-tion of possible moderation effects and ensure sufficient study sizeto do so or examine specific hypotheses in smaller scale studies withcareful selection of the contextual situation

There were also some limitations that need special attentionHighly educated parents and parents with an lsquoalternativersquo lifestylewere overrepresented in our sample partially due to the choice ofrecruitment methods (Gubbels et al 2009) although mean valuesfor the parenting variables (ie general parenting and food parent-ing) were comparable with previous studies (eg Sleddens et al2010 2014) We therefore adjusted all analyses for highest educa-tion level attained and recruitment channel Additionally all of themeasured variables were reported by parents This may evoke socialdesirability The associations between food parenting practices andchildrenrsquos dietary behavior should therefore be interpreted withcaution Cronbachrsquos alpha for the pressure to eat scale of the CFQwas quite low although this finding is in accordance with some pre-vious studies using this scale (eg Boles et al 2010 Gregory Paxtonamp Brozovic 2010) In spite of the longitudinal analysis causality isdifficult to establish since part of the associations (and interac-tions) may be modified by parental adaptations to unwanted be-havior Snacking and intake of sugar-sweetened drinks were studiedas unhealthy behaviors and fruit and water intake as healthy be-haviors Of course other behaviors such as breakfast and vegeta-ble consumption are important as well to determine dietary intakeFinally any choice of single food groups as healthy or unhealthy isdeemed to be debatable due to complexities such as substitution(eg between fruit natural fruit drinks sweetened fruit drinks andsoft drinks tea water and milk drinks) and the ambiguities of re-lations with specific health indicators such as nutritional imbal-ances dental health and overweight development Further studiesare needed with specific health outcomes to evaluate whether mod-eration by contextual factors as shown in our study with health be-haviors as outcome also translate to health outcomes such asoverweight development

Conclusion

Our results show that food parenting practices are important de-terminants in explaining child dietary behavior and that general par-enting behaviors moderate this association Future research effortsshould continue to focus on testing the influence of factors (eggeneral parenting but also child characteristics such as tempera-ment and eating style) impacting on the food parentingndashchild eatingandor weight relationship in order to gain insights into relevant con-

textual factors that need to be taken into account in designinginterventions

References

Birch L L Fisher J O Grimm-Thomas K Markey C N Sawyer R amp JohnsonS L (2001) Confirmatory factor analysis of the Child Feeding Questionnaire Ameasure of parental attitudes beliefs and practices about child feeding andobesity proneness Appetite 36 201ndash210

Blissett J amp Haycraft E (2008) Are parenting style and controlling feeding practicesrelated Appetite 50 477ndash485

Blissett J Meyer C amp Haycraft E (2011) The role of parenting in the relationshipbetween childhood eating problems and broader behaviour problems Child CareHealth and Development 37 642ndash648

Bogers R P Van Assema P Kester A D M Westerterp K R amp Dagnelie P C (2004)Reproducibility validity and responsiveness to change of a short questionnairefor measuring fruit and vegetable intake American Journal of Epidemiology 159900ndash909

Boles R E Nelson T D Chamberlin L A Valenzuela J M Sherman S N JohnsonS L et al (2010) Confirmatory factor analysis of the Child Feeding Questionnaireamong low-income African American families of preschool children Appetite 54402ndash405

Brants H Stafleu A Ter Doest D Hulshof K amp Thijs C (2006) Ontwikkeling vaneen voedselfrequentievragenlijst Energie-inneming van kinderen van 2 tot enmet 12 jaar [Development of a food frequency questionnaire Energy intake ofchildren 2-12 years of age] Voeding Nu 2 25ndash28

Brown K A Ogden J Voumlgele C amp Gibson E L (2008) The role of parental controlpractices in explaining childrenrsquos diet and BMI Appetite 50 252ndash259

Cislak A Safron M Pratt M Gaspar T amp Luszczynska A (2012) Family-relatedpredictors of body weight and weight-related behaviours among children andadolescents A systematic umbrella review Child Care Health and Development38 321ndash331

Clark H R Goyder E Bissell P Blank L amp Peters J (2007) How do parentsrsquochild-feeding behaviours influence child weight Implications for childhoodobesity policy Journal of Public Health 29 132ndash141

Daniels M C amp Popkin B M (2010) Impact of water intake on energy intake andweight status A systematic review Nutrition Reviews 68 505ndash521

Daniels S R (2009) Complications of obesity in children and adolescentsInternational Journal of Obesity 33(s1) 60ndash65

Darling N amp Steinberg L (1993) Parenting style as context An integrative modelPsychological Bulletin 113 487ndash496

Dutman A E Stafleu A Kruizinga A Brants H A M Westerterp K R KistemakerC et al (2010) Validation of an FFQ and options for data processing using thedoubly labelled water method in children Public Health Nutrition 14 410ndash417

Eurostat (2007) Task force on core social variables Final report Luxembourg EuropeanCommunitees

Faith M S Scanlon K S Birch L L Francis L A amp Sherry B (2004) Parent-childfeeding strategies and their relationships to child eating and weight status ObesityResearch 12 1711ndash1722

Fisher J O amp Birch L L (1999) Restricting access to foods and childrenrsquos eatingAppetite 32 405ndash419

Francis L A Hofer S M amp Birch L L (2001) Predictors of maternal child-feedingstyle Maternal and child characteristics Appetite 37 231ndash243

Fuemmeler B F Yang C Costanzo P Hoyle R H Siegler I C Williams R B et al(2012) Parenting styles and body mass index trajectories from adolescence toadulthood Health Psychology 31 441ndash449

Gerards S M P L Sleddens E F C Dagnelie P C De Vries N K amp Kremers S PJ (2011) Interventions addressing general parenting to prevent or treat childhoodobesity International Journal of Pediatric Obesity 6 e28ndashe45

Golley R K Hendrie G A Slater A amp Corsini N (2011) Interventions thatinvolve parents to improve childrenrsquos weight-related nutrition intake andactivity patterns What nutrition and activity targets and behaviour changetechniques are associated with intervention effectiveness Obesity Reviews 12114ndash130

Gregory J E Paxton S J amp Brozovic A M (2010) Maternal feeding practices childeating behaviour and body mass index in preschool-aged children A prospectiveanalysis International Journal of Behavioral Nutrition and Physical Activity 7 55

Gubbels J S Kremers S P J Stafleu A Dagnelie P C Goldbohm R A De VriesN K et al (2009) Diet-related restrictive parenting practices Impact on dietaryintake of 2-year-old children and interactions with child characteristics Appetite52 423ndash429

Hennessy E Hughes S O Goldberg J P Hyatt R R amp Economos C D (2010) Parentbehavior and child weight status among a diverse group of underserved ruralfamilies Appetite 54 369ndash377

Hingle M D OrsquoConnor T M Dave J M amp Baranowski T (2010) Parentalinvolvement in interventions to improve child dietary intake A systematic reviewPreventive Medicine 51 103ndash111

Hu F B (2013) Resolved There is sufficient scientific evidence that decreasingsugar-sweetened beverage consumption will reduce the prevalence of obesityand obesity-related diseases Obesity Reviews 14 606ndash619

Hubbs-Tait L Kennedy T S Page M C Topham G L amp Harrist A W(2008) Parental feeding practices predict authoritative authoritarian andpermissive parenting styles Journal of the American Dietetic Association 1081154ndash1161

49EFC Sleddens et alAppetite 79 (2014) 42ndash50

encouragement and restriction were found to have favorable asso-ciations with child dietary behavior (ie increased healthy intakedecreased unhealthy intake) This is confirming our hypotheses andmost of these relationships have been demonstrated previously (egBrown Ogden Voumlgele amp Gibson 2008 Gubbels et al 2009Murashima Hoerr Hughes amp Kaplowitz 2012 Rodenburg et al2013 Sleddens et al 2010 Van der Horst et al 2007) These as-sociations were stronger for some subgroups of the moderating vari-ables as expected The associations of encouragement and covertcontrol with desirable child dietary behaviors were found to be stron-ger for children who were reared in a positive parenting context

Associations between food parenting and general parenting

The more lsquopositiversquo constructs of general parenting (ienurturance and structure) (Sleddens et al 2014) were associatedwith lower use of instrumental and emotional feeding Reversed as-sociations were found for the more firmly controlling parents Thusthose parents tend to give their child more foods in response togood behavior and to soothe their child The other scales exceptpressure to eat were positively related to the authoritative aspectsof parenting For pressure to eat we found positive associations withcontrolling parenting styles This finding is partially supported byprevious studies (Hennessy et al 2010 Hubbs-Tait et al 2008)

Associations between food parenting and child dietary behavior

We confirmed some of the results of previous studies in whichnondirective child-centered food practices were related to consum-ing healthier diets (Murashima et al 2012 Vereecken et al 2010)and parental reward of food was related to unhealthy food (Kroumllleramp Warschburger 2009 Sleddens et al 2010) and soft drink intake(Kroumlller amp Warschburger 2009) Parental restriction was negative-ly related to sugar-sweetened beverage intake and positively relatedto water intake in contrast to previous studies that showed that high

levels of restriction can lead to increases in calorie intake and likingfor the restricted food (eg Fisher amp Birch 1999 Jansen Mulkensamp Jansen 2007) However the latter types of studies were mainlyexperimental looking at what the child eats when the parent is notaround with free access to food In line with a previous study withinthe KOALA birth cohort by Gubbels et al (2009) we failed to confirmthe paradoxical effect of increased intake of unhealthy foods fol-lowing parental restriction although we should be cautious in in-terpreting these findings as parent-reported child dietary intake onlyaccounts for the dietary behaviors that parents see and are awareof Coercing children to eat was associated with unhealthy eatingHigher levels of pressure in child feeding could have detrimentaleffects on childrenrsquos change in healthy dietary behavior as chil-dren are focused away from internal cues to hunger and satiety(Francis Hofer amp Birch 2001) leading to decreases in preferenceand intake of the healthy foods and subsequent increases in con-sumption of unhealthy foods

General parenting constructs modifying the food parentingndashchilddietary behavior relationship

Moderating effects of general parenting on the relationshipbetween food parenting practices and child eating behavior werefound For children who were reared in a positive parenting contextencouragement and covert control were found to work better (ieincreased healthy intakedecreased unhealthy intake) However someconflicting findings were found The one concept standing out mostwas pressure to eat which was related to sugar-sweetened bever-age intake for children of parents scoring high on behavioral controlPressure to eat often occurs when parents feel that their child iseating insufficient amounts of food or in response to their childrsquosunderweight (Francis et al 2001) Secondary analyses confirmedthis assumption as we found that parents of children who were un-derweight scored significantly higher on pressure to eat com-pared with parents of children who were overweight or obese (data

Table 3Interaction terms and stratified analyses food parentingndashchild dietary behavior relationship

Dietary behavior Year Interaction term(general parenting times food parenting)

Groups

Low Intermediate High

Beta P Beta P Beta P Beta P

Snacking 6ndash8 Behavioral control times Encouragement minus62 034 02 624 00 994 minus11 010Structure times Pressure to eat 53 045 03 533 minus02 641 08 053

8 Overprotection times Covert control minus27 099 minus14 005 minus21 000 minus19 000Sugar-sweetened beverage 6ndash8 Behavioral control times Instrumental feeding 77 007 minus04 509 minus11 048 12 036

Behavioral control times Encouragement minus91 009 03 539 01 788 minus14 0058 Behavioral control times Pressure to eat 82 018 03 578 15 003 13 021

Behavioral control times Encouragement minus116 005 07 218 minus01 899 minus08 135Behavioral control times Control 96 054 minus07 221 minus07 149 minus02 688

Fruit 6ndash8 Nurturance times Monitoring 89 034 minus05 344 05 308 07 081Behavioral control times Instrumental feeding 52 049 minus16 003 minus08 080 minus01 785

8 Behavioral control times Monitoring 125 004 minus04 444 03 553 08 169Behavioral control times Emotional feeding 66 027 minus05 439 07 220 09 163Behavioral control times Restriction minus69 086 15 012 10 064 minus05 437Overprotection times Instrumental feeding 85 031 minus02 800 minus13 028 minus14 043Overprotection times Encouragement minus52 091 14 011 06 231 06 312

Water 6ndash8 Nurturance times Instrumental feeding minus70 024 01 874 08 067 minus04 467Overprotection times Monitoring minus41 070 03 536 00 977 minus04 445

8 Nurturance times Instrumental feeding minus110 004 06 416 02 709 minus15 013Coercive control times Encouragement minus55 047 09 112 07 157 minus01 881Coercive control times Restriction 47 083 05 459 10 047 16 013Overprotection times Covert Control minus34 062 17 003 07 170 09 184Overprotection times Pressure to eat 53 014 minus06 320 minus09 091 02 712

Note Interactions for which the association between the food parenting practice and the outcome was statistically significant in one or more groups of the moderator vari-able Beta standardized regression coefficient from linear regression analysis stratified for the moderator (ie low intermediate high) model adjusted for recruitment groupparental educational level country of birth maternal age at birth child gender and age child BMI z-score at age 6 and all other food parenting practices (note that therewere two different models (1) dependent variable is additionally adjusted for baseline child dietary behavior at age 6 and adjusted for child BMI z-score at age 6 (2) de-pendent variable is child dietary behavior at age 8 adjusted for child BMI z-score at age 6

47EFC Sleddens et alAppetite 79 (2014) 42ndash50

nsns

nsns

ns ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

ASnacking Sugar sweetened

beverageFruit

Behavioral control OverprotectionBehavioral control

6-8 yr 6-8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 21 Food parenting practice lsquoencouragementrsquo

nsns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Snacking Water

OverprotectionOverprotection

8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 22 Food parenting practice lsquocovert controlrsquo

ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Fruit Water

Behavioral control Coercive control

8 yr 6-8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 23 Food parenting practice lsquorestrictionrsquo

ns

ns

ns ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidrad nats(no itaicoss

A

Sugar sweetened beverage

Fruit WaterFruit

Behavioral control Behavioral control Overprotection Nurturance

6-8 yr 6-8 yr 8-yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 24 Food parenting practice lsquoinstrumental feedingrsquo

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Sugar sweetened beverage

Behavioral control

8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 25 Food parenting practice lsquopressure to eatrsquo

Fig 2 Stratified analyses the food parentingndashchild dietary behavior relationship moderated by general parenting Note Red bars represent significant associations betweenfood parenting and undesirable behavior (ie increased unhealthy intakedecreased healthy intake) green bars represent significant associations between food parentingand desirable behavior (ie increased healthy intakedecreased unhealthy intake) white represent nonsignificant associations between food parenting and child dietaryintake bars within each cluster represent the stratified levels of the moderator (iegeneral parenting constructs) p le 05 p le 01 p le 001

48 EFC Sleddens et alAppetite 79 (2014) 42ndash50

not reported) Low weight status has been proven to be related withmore food avoidant behaviors (Jansen et al 2012 Sleddens Kremersamp Thijs 2008) As a result parents could apply coercing food par-enting practices trying to increase their childrsquos calorie intake andprobably also increasing their childrsquos unhealthy dietary behavior (iesnacking and sugar-sweetened beverage intake) Our moderationresults should be interpreted with caution and more studies areneeded before drawing definite conclusions

Study strengths and limitations

This study benefitted from a longitudinal design with mea-sures of child dietary behavior repeated at age 6 and 8 The in-cluded variables (food parenting practices and general parenting)were measured with instruments validated for the Dutch context(Bogers et al 2004 Brants et al 2006 Dutman et al 2010 Sleddenset al 2010 2014) Moderation analysis was possible thanks to alarge sample size and sufficient diversity within the study Weconfirmed the operation of higher-order moderation processesimplying that parenting factors at higher more distal levels alterthe impact of food parenting practices at more proximal levels Werecommend future studies to include the theory-based examina-tion of possible moderation effects and ensure sufficient study sizeto do so or examine specific hypotheses in smaller scale studies withcareful selection of the contextual situation

There were also some limitations that need special attentionHighly educated parents and parents with an lsquoalternativersquo lifestylewere overrepresented in our sample partially due to the choice ofrecruitment methods (Gubbels et al 2009) although mean valuesfor the parenting variables (ie general parenting and food parent-ing) were comparable with previous studies (eg Sleddens et al2010 2014) We therefore adjusted all analyses for highest educa-tion level attained and recruitment channel Additionally all of themeasured variables were reported by parents This may evoke socialdesirability The associations between food parenting practices andchildrenrsquos dietary behavior should therefore be interpreted withcaution Cronbachrsquos alpha for the pressure to eat scale of the CFQwas quite low although this finding is in accordance with some pre-vious studies using this scale (eg Boles et al 2010 Gregory Paxtonamp Brozovic 2010) In spite of the longitudinal analysis causality isdifficult to establish since part of the associations (and interac-tions) may be modified by parental adaptations to unwanted be-havior Snacking and intake of sugar-sweetened drinks were studiedas unhealthy behaviors and fruit and water intake as healthy be-haviors Of course other behaviors such as breakfast and vegeta-ble consumption are important as well to determine dietary intakeFinally any choice of single food groups as healthy or unhealthy isdeemed to be debatable due to complexities such as substitution(eg between fruit natural fruit drinks sweetened fruit drinks andsoft drinks tea water and milk drinks) and the ambiguities of re-lations with specific health indicators such as nutritional imbal-ances dental health and overweight development Further studiesare needed with specific health outcomes to evaluate whether mod-eration by contextual factors as shown in our study with health be-haviors as outcome also translate to health outcomes such asoverweight development

Conclusion

Our results show that food parenting practices are important de-terminants in explaining child dietary behavior and that general par-enting behaviors moderate this association Future research effortsshould continue to focus on testing the influence of factors (eggeneral parenting but also child characteristics such as tempera-ment and eating style) impacting on the food parentingndashchild eatingandor weight relationship in order to gain insights into relevant con-

textual factors that need to be taken into account in designinginterventions

References

Birch L L Fisher J O Grimm-Thomas K Markey C N Sawyer R amp JohnsonS L (2001) Confirmatory factor analysis of the Child Feeding Questionnaire Ameasure of parental attitudes beliefs and practices about child feeding andobesity proneness Appetite 36 201ndash210

Blissett J amp Haycraft E (2008) Are parenting style and controlling feeding practicesrelated Appetite 50 477ndash485

Blissett J Meyer C amp Haycraft E (2011) The role of parenting in the relationshipbetween childhood eating problems and broader behaviour problems Child CareHealth and Development 37 642ndash648

Bogers R P Van Assema P Kester A D M Westerterp K R amp Dagnelie P C (2004)Reproducibility validity and responsiveness to change of a short questionnairefor measuring fruit and vegetable intake American Journal of Epidemiology 159900ndash909

Boles R E Nelson T D Chamberlin L A Valenzuela J M Sherman S N JohnsonS L et al (2010) Confirmatory factor analysis of the Child Feeding Questionnaireamong low-income African American families of preschool children Appetite 54402ndash405

Brants H Stafleu A Ter Doest D Hulshof K amp Thijs C (2006) Ontwikkeling vaneen voedselfrequentievragenlijst Energie-inneming van kinderen van 2 tot enmet 12 jaar [Development of a food frequency questionnaire Energy intake ofchildren 2-12 years of age] Voeding Nu 2 25ndash28

Brown K A Ogden J Voumlgele C amp Gibson E L (2008) The role of parental controlpractices in explaining childrenrsquos diet and BMI Appetite 50 252ndash259

Cislak A Safron M Pratt M Gaspar T amp Luszczynska A (2012) Family-relatedpredictors of body weight and weight-related behaviours among children andadolescents A systematic umbrella review Child Care Health and Development38 321ndash331

Clark H R Goyder E Bissell P Blank L amp Peters J (2007) How do parentsrsquochild-feeding behaviours influence child weight Implications for childhoodobesity policy Journal of Public Health 29 132ndash141

Daniels M C amp Popkin B M (2010) Impact of water intake on energy intake andweight status A systematic review Nutrition Reviews 68 505ndash521

Daniels S R (2009) Complications of obesity in children and adolescentsInternational Journal of Obesity 33(s1) 60ndash65

Darling N amp Steinberg L (1993) Parenting style as context An integrative modelPsychological Bulletin 113 487ndash496

Dutman A E Stafleu A Kruizinga A Brants H A M Westerterp K R KistemakerC et al (2010) Validation of an FFQ and options for data processing using thedoubly labelled water method in children Public Health Nutrition 14 410ndash417

Eurostat (2007) Task force on core social variables Final report Luxembourg EuropeanCommunitees

Faith M S Scanlon K S Birch L L Francis L A amp Sherry B (2004) Parent-childfeeding strategies and their relationships to child eating and weight status ObesityResearch 12 1711ndash1722

Fisher J O amp Birch L L (1999) Restricting access to foods and childrenrsquos eatingAppetite 32 405ndash419

Francis L A Hofer S M amp Birch L L (2001) Predictors of maternal child-feedingstyle Maternal and child characteristics Appetite 37 231ndash243

Fuemmeler B F Yang C Costanzo P Hoyle R H Siegler I C Williams R B et al(2012) Parenting styles and body mass index trajectories from adolescence toadulthood Health Psychology 31 441ndash449

Gerards S M P L Sleddens E F C Dagnelie P C De Vries N K amp Kremers S PJ (2011) Interventions addressing general parenting to prevent or treat childhoodobesity International Journal of Pediatric Obesity 6 e28ndashe45

Golley R K Hendrie G A Slater A amp Corsini N (2011) Interventions thatinvolve parents to improve childrenrsquos weight-related nutrition intake andactivity patterns What nutrition and activity targets and behaviour changetechniques are associated with intervention effectiveness Obesity Reviews 12114ndash130

Gregory J E Paxton S J amp Brozovic A M (2010) Maternal feeding practices childeating behaviour and body mass index in preschool-aged children A prospectiveanalysis International Journal of Behavioral Nutrition and Physical Activity 7 55

Gubbels J S Kremers S P J Stafleu A Dagnelie P C Goldbohm R A De VriesN K et al (2009) Diet-related restrictive parenting practices Impact on dietaryintake of 2-year-old children and interactions with child characteristics Appetite52 423ndash429

Hennessy E Hughes S O Goldberg J P Hyatt R R amp Economos C D (2010) Parentbehavior and child weight status among a diverse group of underserved ruralfamilies Appetite 54 369ndash377

Hingle M D OrsquoConnor T M Dave J M amp Baranowski T (2010) Parentalinvolvement in interventions to improve child dietary intake A systematic reviewPreventive Medicine 51 103ndash111

Hu F B (2013) Resolved There is sufficient scientific evidence that decreasingsugar-sweetened beverage consumption will reduce the prevalence of obesityand obesity-related diseases Obesity Reviews 14 606ndash619

Hubbs-Tait L Kennedy T S Page M C Topham G L amp Harrist A W(2008) Parental feeding practices predict authoritative authoritarian andpermissive parenting styles Journal of the American Dietetic Association 1081154ndash1161

49EFC Sleddens et alAppetite 79 (2014) 42ndash50

nsns

nsns

ns ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

ASnacking Sugar sweetened

beverageFruit

Behavioral control OverprotectionBehavioral control

6-8 yr 6-8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 21 Food parenting practice lsquoencouragementrsquo

nsns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Snacking Water

OverprotectionOverprotection

8 yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 22 Food parenting practice lsquocovert controlrsquo

ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Fruit Water

Behavioral control Coercive control

8 yr 6-8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 23 Food parenting practice lsquorestrictionrsquo

ns

ns

ns ns

ns

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidrad nats(no itaicoss

A

Sugar sweetened beverage

Fruit WaterFruit

Behavioral control Behavioral control Overprotection Nurturance

6-8 yr 6-8 yr 8-yr 8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 24 Food parenting practice lsquoinstrumental feedingrsquo

ns

-025

-02

-015

-01

-005

0

005

01

015

02

025

)tneiciff eocnoi sserger

dezidr ad nats(no itaicoss

A

Sugar sweetened beverage

Behavioral control

8 yr

Low (left bar) Intermediate (middle bar) High (right bar)

Fig 25 Food parenting practice lsquopressure to eatrsquo

Fig 2 Stratified analyses the food parentingndashchild dietary behavior relationship moderated by general parenting Note Red bars represent significant associations betweenfood parenting and undesirable behavior (ie increased unhealthy intakedecreased healthy intake) green bars represent significant associations between food parentingand desirable behavior (ie increased healthy intakedecreased unhealthy intake) white represent nonsignificant associations between food parenting and child dietaryintake bars within each cluster represent the stratified levels of the moderator (iegeneral parenting constructs) p le 05 p le 01 p le 001

48 EFC Sleddens et alAppetite 79 (2014) 42ndash50

not reported) Low weight status has been proven to be related withmore food avoidant behaviors (Jansen et al 2012 Sleddens Kremersamp Thijs 2008) As a result parents could apply coercing food par-enting practices trying to increase their childrsquos calorie intake andprobably also increasing their childrsquos unhealthy dietary behavior (iesnacking and sugar-sweetened beverage intake) Our moderationresults should be interpreted with caution and more studies areneeded before drawing definite conclusions

Study strengths and limitations

This study benefitted from a longitudinal design with mea-sures of child dietary behavior repeated at age 6 and 8 The in-cluded variables (food parenting practices and general parenting)were measured with instruments validated for the Dutch context(Bogers et al 2004 Brants et al 2006 Dutman et al 2010 Sleddenset al 2010 2014) Moderation analysis was possible thanks to alarge sample size and sufficient diversity within the study Weconfirmed the operation of higher-order moderation processesimplying that parenting factors at higher more distal levels alterthe impact of food parenting practices at more proximal levels Werecommend future studies to include the theory-based examina-tion of possible moderation effects and ensure sufficient study sizeto do so or examine specific hypotheses in smaller scale studies withcareful selection of the contextual situation

There were also some limitations that need special attentionHighly educated parents and parents with an lsquoalternativersquo lifestylewere overrepresented in our sample partially due to the choice ofrecruitment methods (Gubbels et al 2009) although mean valuesfor the parenting variables (ie general parenting and food parent-ing) were comparable with previous studies (eg Sleddens et al2010 2014) We therefore adjusted all analyses for highest educa-tion level attained and recruitment channel Additionally all of themeasured variables were reported by parents This may evoke socialdesirability The associations between food parenting practices andchildrenrsquos dietary behavior should therefore be interpreted withcaution Cronbachrsquos alpha for the pressure to eat scale of the CFQwas quite low although this finding is in accordance with some pre-vious studies using this scale (eg Boles et al 2010 Gregory Paxtonamp Brozovic 2010) In spite of the longitudinal analysis causality isdifficult to establish since part of the associations (and interac-tions) may be modified by parental adaptations to unwanted be-havior Snacking and intake of sugar-sweetened drinks were studiedas unhealthy behaviors and fruit and water intake as healthy be-haviors Of course other behaviors such as breakfast and vegeta-ble consumption are important as well to determine dietary intakeFinally any choice of single food groups as healthy or unhealthy isdeemed to be debatable due to complexities such as substitution(eg between fruit natural fruit drinks sweetened fruit drinks andsoft drinks tea water and milk drinks) and the ambiguities of re-lations with specific health indicators such as nutritional imbal-ances dental health and overweight development Further studiesare needed with specific health outcomes to evaluate whether mod-eration by contextual factors as shown in our study with health be-haviors as outcome also translate to health outcomes such asoverweight development

Conclusion

Our results show that food parenting practices are important de-terminants in explaining child dietary behavior and that general par-enting behaviors moderate this association Future research effortsshould continue to focus on testing the influence of factors (eggeneral parenting but also child characteristics such as tempera-ment and eating style) impacting on the food parentingndashchild eatingandor weight relationship in order to gain insights into relevant con-

textual factors that need to be taken into account in designinginterventions

References

Birch L L Fisher J O Grimm-Thomas K Markey C N Sawyer R amp JohnsonS L (2001) Confirmatory factor analysis of the Child Feeding Questionnaire Ameasure of parental attitudes beliefs and practices about child feeding andobesity proneness Appetite 36 201ndash210

Blissett J amp Haycraft E (2008) Are parenting style and controlling feeding practicesrelated Appetite 50 477ndash485

Blissett J Meyer C amp Haycraft E (2011) The role of parenting in the relationshipbetween childhood eating problems and broader behaviour problems Child CareHealth and Development 37 642ndash648

Bogers R P Van Assema P Kester A D M Westerterp K R amp Dagnelie P C (2004)Reproducibility validity and responsiveness to change of a short questionnairefor measuring fruit and vegetable intake American Journal of Epidemiology 159900ndash909

Boles R E Nelson T D Chamberlin L A Valenzuela J M Sherman S N JohnsonS L et al (2010) Confirmatory factor analysis of the Child Feeding Questionnaireamong low-income African American families of preschool children Appetite 54402ndash405

Brants H Stafleu A Ter Doest D Hulshof K amp Thijs C (2006) Ontwikkeling vaneen voedselfrequentievragenlijst Energie-inneming van kinderen van 2 tot enmet 12 jaar [Development of a food frequency questionnaire Energy intake ofchildren 2-12 years of age] Voeding Nu 2 25ndash28

Brown K A Ogden J Voumlgele C amp Gibson E L (2008) The role of parental controlpractices in explaining childrenrsquos diet and BMI Appetite 50 252ndash259

Cislak A Safron M Pratt M Gaspar T amp Luszczynska A (2012) Family-relatedpredictors of body weight and weight-related behaviours among children andadolescents A systematic umbrella review Child Care Health and Development38 321ndash331

Clark H R Goyder E Bissell P Blank L amp Peters J (2007) How do parentsrsquochild-feeding behaviours influence child weight Implications for childhoodobesity policy Journal of Public Health 29 132ndash141

Daniels M C amp Popkin B M (2010) Impact of water intake on energy intake andweight status A systematic review Nutrition Reviews 68 505ndash521

Daniels S R (2009) Complications of obesity in children and adolescentsInternational Journal of Obesity 33(s1) 60ndash65

Darling N amp Steinberg L (1993) Parenting style as context An integrative modelPsychological Bulletin 113 487ndash496

Dutman A E Stafleu A Kruizinga A Brants H A M Westerterp K R KistemakerC et al (2010) Validation of an FFQ and options for data processing using thedoubly labelled water method in children Public Health Nutrition 14 410ndash417

Eurostat (2007) Task force on core social variables Final report Luxembourg EuropeanCommunitees

Faith M S Scanlon K S Birch L L Francis L A amp Sherry B (2004) Parent-childfeeding strategies and their relationships to child eating and weight status ObesityResearch 12 1711ndash1722

Fisher J O amp Birch L L (1999) Restricting access to foods and childrenrsquos eatingAppetite 32 405ndash419

Francis L A Hofer S M amp Birch L L (2001) Predictors of maternal child-feedingstyle Maternal and child characteristics Appetite 37 231ndash243

Fuemmeler B F Yang C Costanzo P Hoyle R H Siegler I C Williams R B et al(2012) Parenting styles and body mass index trajectories from adolescence toadulthood Health Psychology 31 441ndash449

Gerards S M P L Sleddens E F C Dagnelie P C De Vries N K amp Kremers S PJ (2011) Interventions addressing general parenting to prevent or treat childhoodobesity International Journal of Pediatric Obesity 6 e28ndashe45

Golley R K Hendrie G A Slater A amp Corsini N (2011) Interventions thatinvolve parents to improve childrenrsquos weight-related nutrition intake andactivity patterns What nutrition and activity targets and behaviour changetechniques are associated with intervention effectiveness Obesity Reviews 12114ndash130

Gregory J E Paxton S J amp Brozovic A M (2010) Maternal feeding practices childeating behaviour and body mass index in preschool-aged children A prospectiveanalysis International Journal of Behavioral Nutrition and Physical Activity 7 55

Gubbels J S Kremers S P J Stafleu A Dagnelie P C Goldbohm R A De VriesN K et al (2009) Diet-related restrictive parenting practices Impact on dietaryintake of 2-year-old children and interactions with child characteristics Appetite52 423ndash429

Hennessy E Hughes S O Goldberg J P Hyatt R R amp Economos C D (2010) Parentbehavior and child weight status among a diverse group of underserved ruralfamilies Appetite 54 369ndash377

Hingle M D OrsquoConnor T M Dave J M amp Baranowski T (2010) Parentalinvolvement in interventions to improve child dietary intake A systematic reviewPreventive Medicine 51 103ndash111

Hu F B (2013) Resolved There is sufficient scientific evidence that decreasingsugar-sweetened beverage consumption will reduce the prevalence of obesityand obesity-related diseases Obesity Reviews 14 606ndash619

Hubbs-Tait L Kennedy T S Page M C Topham G L amp Harrist A W(2008) Parental feeding practices predict authoritative authoritarian andpermissive parenting styles Journal of the American Dietetic Association 1081154ndash1161

49EFC Sleddens et alAppetite 79 (2014) 42ndash50

not reported) Low weight status has been proven to be related withmore food avoidant behaviors (Jansen et al 2012 Sleddens Kremersamp Thijs 2008) As a result parents could apply coercing food par-enting practices trying to increase their childrsquos calorie intake andprobably also increasing their childrsquos unhealthy dietary behavior (iesnacking and sugar-sweetened beverage intake) Our moderationresults should be interpreted with caution and more studies areneeded before drawing definite conclusions

Study strengths and limitations

This study benefitted from a longitudinal design with mea-sures of child dietary behavior repeated at age 6 and 8 The in-cluded variables (food parenting practices and general parenting)were measured with instruments validated for the Dutch context(Bogers et al 2004 Brants et al 2006 Dutman et al 2010 Sleddenset al 2010 2014) Moderation analysis was possible thanks to alarge sample size and sufficient diversity within the study Weconfirmed the operation of higher-order moderation processesimplying that parenting factors at higher more distal levels alterthe impact of food parenting practices at more proximal levels Werecommend future studies to include the theory-based examina-tion of possible moderation effects and ensure sufficient study sizeto do so or examine specific hypotheses in smaller scale studies withcareful selection of the contextual situation

There were also some limitations that need special attentionHighly educated parents and parents with an lsquoalternativersquo lifestylewere overrepresented in our sample partially due to the choice ofrecruitment methods (Gubbels et al 2009) although mean valuesfor the parenting variables (ie general parenting and food parent-ing) were comparable with previous studies (eg Sleddens et al2010 2014) We therefore adjusted all analyses for highest educa-tion level attained and recruitment channel Additionally all of themeasured variables were reported by parents This may evoke socialdesirability The associations between food parenting practices andchildrenrsquos dietary behavior should therefore be interpreted withcaution Cronbachrsquos alpha for the pressure to eat scale of the CFQwas quite low although this finding is in accordance with some pre-vious studies using this scale (eg Boles et al 2010 Gregory Paxtonamp Brozovic 2010) In spite of the longitudinal analysis causality isdifficult to establish since part of the associations (and interac-tions) may be modified by parental adaptations to unwanted be-havior Snacking and intake of sugar-sweetened drinks were studiedas unhealthy behaviors and fruit and water intake as healthy be-haviors Of course other behaviors such as breakfast and vegeta-ble consumption are important as well to determine dietary intakeFinally any choice of single food groups as healthy or unhealthy isdeemed to be debatable due to complexities such as substitution(eg between fruit natural fruit drinks sweetened fruit drinks andsoft drinks tea water and milk drinks) and the ambiguities of re-lations with specific health indicators such as nutritional imbal-ances dental health and overweight development Further studiesare needed with specific health outcomes to evaluate whether mod-eration by contextual factors as shown in our study with health be-haviors as outcome also translate to health outcomes such asoverweight development

Conclusion

Our results show that food parenting practices are important de-terminants in explaining child dietary behavior and that general par-enting behaviors moderate this association Future research effortsshould continue to focus on testing the influence of factors (eggeneral parenting but also child characteristics such as tempera-ment and eating style) impacting on the food parentingndashchild eatingandor weight relationship in order to gain insights into relevant con-

textual factors that need to be taken into account in designinginterventions

References

Birch L L Fisher J O Grimm-Thomas K Markey C N Sawyer R amp JohnsonS L (2001) Confirmatory factor analysis of the Child Feeding Questionnaire Ameasure of parental attitudes beliefs and practices about child feeding andobesity proneness Appetite 36 201ndash210

Blissett J amp Haycraft E (2008) Are parenting style and controlling feeding practicesrelated Appetite 50 477ndash485

Blissett J Meyer C amp Haycraft E (2011) The role of parenting in the relationshipbetween childhood eating problems and broader behaviour problems Child CareHealth and Development 37 642ndash648

Bogers R P Van Assema P Kester A D M Westerterp K R amp Dagnelie P C (2004)Reproducibility validity and responsiveness to change of a short questionnairefor measuring fruit and vegetable intake American Journal of Epidemiology 159900ndash909

Boles R E Nelson T D Chamberlin L A Valenzuela J M Sherman S N JohnsonS L et al (2010) Confirmatory factor analysis of the Child Feeding Questionnaireamong low-income African American families of preschool children Appetite 54402ndash405

Brants H Stafleu A Ter Doest D Hulshof K amp Thijs C (2006) Ontwikkeling vaneen voedselfrequentievragenlijst Energie-inneming van kinderen van 2 tot enmet 12 jaar [Development of a food frequency questionnaire Energy intake ofchildren 2-12 years of age] Voeding Nu 2 25ndash28

Brown K A Ogden J Voumlgele C amp Gibson E L (2008) The role of parental controlpractices in explaining childrenrsquos diet and BMI Appetite 50 252ndash259

Cislak A Safron M Pratt M Gaspar T amp Luszczynska A (2012) Family-relatedpredictors of body weight and weight-related behaviours among children andadolescents A systematic umbrella review Child Care Health and Development38 321ndash331

Clark H R Goyder E Bissell P Blank L amp Peters J (2007) How do parentsrsquochild-feeding behaviours influence child weight Implications for childhoodobesity policy Journal of Public Health 29 132ndash141

Daniels M C amp Popkin B M (2010) Impact of water intake on energy intake andweight status A systematic review Nutrition Reviews 68 505ndash521

Daniels S R (2009) Complications of obesity in children and adolescentsInternational Journal of Obesity 33(s1) 60ndash65

Darling N amp Steinberg L (1993) Parenting style as context An integrative modelPsychological Bulletin 113 487ndash496

Dutman A E Stafleu A Kruizinga A Brants H A M Westerterp K R KistemakerC et al (2010) Validation of an FFQ and options for data processing using thedoubly labelled water method in children Public Health Nutrition 14 410ndash417

Eurostat (2007) Task force on core social variables Final report Luxembourg EuropeanCommunitees

Faith M S Scanlon K S Birch L L Francis L A amp Sherry B (2004) Parent-childfeeding strategies and their relationships to child eating and weight status ObesityResearch 12 1711ndash1722

Fisher J O amp Birch L L (1999) Restricting access to foods and childrenrsquos eatingAppetite 32 405ndash419

Francis L A Hofer S M amp Birch L L (2001) Predictors of maternal child-feedingstyle Maternal and child characteristics Appetite 37 231ndash243

Fuemmeler B F Yang C Costanzo P Hoyle R H Siegler I C Williams R B et al(2012) Parenting styles and body mass index trajectories from adolescence toadulthood Health Psychology 31 441ndash449

Gerards S M P L Sleddens E F C Dagnelie P C De Vries N K amp Kremers S PJ (2011) Interventions addressing general parenting to prevent or treat childhoodobesity International Journal of Pediatric Obesity 6 e28ndashe45

Golley R K Hendrie G A Slater A amp Corsini N (2011) Interventions thatinvolve parents to improve childrenrsquos weight-related nutrition intake andactivity patterns What nutrition and activity targets and behaviour changetechniques are associated with intervention effectiveness Obesity Reviews 12114ndash130

Gregory J E Paxton S J amp Brozovic A M (2010) Maternal feeding practices childeating behaviour and body mass index in preschool-aged children A prospectiveanalysis International Journal of Behavioral Nutrition and Physical Activity 7 55

Gubbels J S Kremers S P J Stafleu A Dagnelie P C Goldbohm R A De VriesN K et al (2009) Diet-related restrictive parenting practices Impact on dietaryintake of 2-year-old children and interactions with child characteristics Appetite52 423ndash429

Hennessy E Hughes S O Goldberg J P Hyatt R R amp Economos C D (2010) Parentbehavior and child weight status among a diverse group of underserved ruralfamilies Appetite 54 369ndash377

Hingle M D OrsquoConnor T M Dave J M amp Baranowski T (2010) Parentalinvolvement in interventions to improve child dietary intake A systematic reviewPreventive Medicine 51 103ndash111

Hu F B (2013) Resolved There is sufficient scientific evidence that decreasingsugar-sweetened beverage consumption will reduce the prevalence of obesityand obesity-related diseases Obesity Reviews 14 606ndash619

Hubbs-Tait L Kennedy T S Page M C Topham G L amp Harrist A W(2008) Parental feeding practices predict authoritative authoritarian andpermissive parenting styles Journal of the American Dietetic Association 1081154ndash1161

49EFC Sleddens et alAppetite 79 (2014) 42ndash50

Jansen E Mulkens S amp Jansen A (2007) Do not eat the red food Prohibition ofsnacks leads to their relatively higher consumption in children Appetite 49572ndash577

Jansen P W Roza S J Jaddoe V W V Mackenbach J D Raat H Hofman Aet al (2012) Childrenrsquos eating behavior feeding practices of parents andweight problems in early childhood Results from the population-basedGeneration R Study International Journal of Behavioral Nutrition and PhysicalActivity 9 130

Johnson R Welk G Saint-Maurice P F amp Ihmels M (2012) Parenting styles andhome obesogenic environments International Journal of Environmental Researchand Public Health 9 1411ndash1426

Kremers S P J Sleddens E F C Gerards S M P L Gubbels J S Rodenburg GGevers G et al (2013) General and food-specific parenting Measures andinterplay Childhood Obesity 9(s1) 22ndash31

Kroumlller K amp Warschburger P (2009) Maternal feeding strategies and childrsquos foodintake Considering weight and demographic influences using structural equationmodeling International Journal of Behavioral Nutrition and Physical Activity 678

Kummeling I Thijs C Penders J Snijders B E Stelma F Reimerink J et al (2005)Etiology of atopy in infancy The KOALA Birth Cohort Study Pediatric Allergy andImmunology 16 679ndash684

Ledoux T A Hingle M D amp Baranowski T (2011) Relationship of fruit and vegetableintake with adiposity A systematic review Obesity Reviews 12 e143ndashe150

Murashima M Hoerr S L Hughes S O amp Kaplowitz S A (2012) Feeding behaviorsof low-income mothers Directive control relates to a lower BMI in children anda nondirective control relates to a healthier diet in preschoolers American Journalof Clinical Nutrition 95 1031ndash1037

Musher-Eizenman D R amp Holub S C (2006) Childrenrsquos eating in the absence ofhunger The role of restrictive feeding practices In R Flamenbaum (Ed) Childhoodobesity and health research (pp 135ndash156) Hauppauge NY Nova

OrsquoConnor T M Jago R amp Baranowski T (2009) Engaging parents to increase youthphysical activity A systematic review American Journal of Preventive Medicine37 141ndash149

Ogden J Reynolds R amp Smith A (2006) Expanding the concept of parental controlA role for overt and covert control in childrenrsquos snacking behaviour Appetite47 100ndash106

Pryor LE Tremblay R E Boivin M Touchette E amp Dubois L (2011) Developmentaltrajectories of Body Mass Index in early childhood and their risk factors An 8-yearlongitudinal study Archives of Pediatrics and Adolescent Medicine 165 906ndash912

Reilly J J amp Kelly J (2011) Long-term impact of overweight and obesity in childhoodand adolescence on morbidity and premature mortality in adulthood Systematicreview International Journal of Obesity 35 891ndash898

Rodenburg G Kremers S P J Oenema A amp Van de Mheen D (2011) Psychologicalcontrol by parents is associated with a higher child weight International Journalof Pediatric Obesity 6 442ndash449

Rodenburg G Kremers S P J Oenema A amp Van de Mheen D (2013) Associationsof parental feeding styles with child snacking behaviour and weight in the contextof general parenting Public Health Nutrition Epub March 26

Rosenheck R (2008) Fast food consumption and increased caloric intake Asystematic review of a trajectory weight gain and obesity risk Obesity Reviews9 535ndash547

Singh A S Mulder C Twisk J W R Van Mechelen W amp Chinapaw M J M (2008)Tracking of childhood overweight into adulthood A systematic review of theliterature Obesity Reviews 9 474ndash488

Sleddens E F C Gerards S M P L Thijs C De Vries N K amp Kremers S P J (2011)General parenting childhood overweight and obesity-inducing behaviors Areview International Journal of Pediatric Obesity 6 e12ndashe27

Sleddens E F C Kremers S P J De Vries N K amp Thijs C (2010) Relationshipbetween parental feeding styles and eating behaviours of Dutch children aged6-7 Appetite 54 30ndash36

Sleddens E F C Kremers S P J amp Thijs C (2008) The Childrenrsquos Eating BehaviourQuestionnaire Factorial validity and association with Body Mass Index in Dutchchildren aged 6-7 International Journal of Behavioral Nutrition and Physical Activity5 49

Sleddens E F C OrsquoConnor T M Watson K B Hughes S O Power T G Thijs Cet al (2014) Development of the Comprehensive General Parenting Questionnairefor caregivers of 5-13 year olds International Journal of Behavioral Nutrition andPhysical Activity 11 15

Stone-Romero E F amp Liakhovitski D (2002) Strategies for detecting moderatorvariables A review of conceptual and empirical issues Research in Personnel andHuman Resources Management 21 33ndash72

Taylor A Wilson C Slater A amp Mohr P (2011) Parent- and child-reported parentingAssociations with child weight-related outcomes Appetite 57 700ndash706

Topham G L Hubbs-Tait L Rutledge J M Page M C Kennedy T S Shriver LH et al (2011) Parenting styles parental response to child emotion and familyemotional responsiveness are related to child emotional eating Appetite 56261ndash264

Van der Horst K Kremers S Ferreira I Singh A Oenema A amp Brug J (2007)Perceived parenting style and practices and the consumption of sugar-sweetenedbeverages by adolescents Health Education Research 22 295ndash304

Ventura A K amp Birch L L (2008) Does parenting affect childrenrsquos eating andweight status International Journal of Behavioral Nutrition and Physical Activity5 15

Ventura A K Loken E amp Birch L L (2009) Developmental trajectories of girlsrsquo BMIacross childhood and adolescence Obesity 17 2067ndash2074

Vereecken C Rovner A amp Maes L (2010) Associations of parenting styles parentalfeeding practices and child characteristics with young childrenrsquos fruit andvegetable consumption Appetite 55 589ndash596

Wardle J amp Carnell S (2007) Parental feeding practices and childrenrsquos weight ActaPaediatrica 96 5ndash11

Wardle J Sanderson S Guthrie C A Rapoport L amp Plomin R (2002) Parentalfeeding style and the inter-generational transmission of obesity risk ObesityResearch 10 453ndash462

Webber L Cooke L Hill C amp Wardle J (2010) Associations between childrenrsquosappetitive traits and maternal feeding practices Journal of the American DieteticAssociation 110 1718ndash1722

50 EFC Sleddens et alAppetite 79 (2014) 42ndash50