Pediatric Pulmonology 41:338–344 (2006)

Is Significant Cystic Fibrosis-Related Liver Disease aRisk Factor in the Development of Bone

Mineralization Abnormalities?

George Alex, PhD, FRACP,1 Anthony G. Catto-Smith, MD, FRACP,1,2,3 Michael Ditchfield, MD, FRACR,4

Robert Roseby, PhD, FRACP,5 Philip J. Robinson, MD, PhD, FRACP,1,2,3

Fergus J. Cameron, MD, FRACP,2,3,6 and Mark R. Oliver, MD, FRACP1,2,3*

Summary. In order to assess the effects of significant cystic fibrosis-related liver disease (CFLD)

on bone health, we compared the bonemineral status of older children and adolescentswithCFLD

to thosewith cystic fibrosis (CF) alone. Thirteen children (age range, 10–19 years) fromour clinical

CF services were identified with significant CFLD (9 of these 13 patients had clinical and

radiological evidence of portal hypertension). This cohort was then matched by age, gender, and

anthropometric measurements with equal numbers of patients with CF alone. All patients had a

dual-energy X-ray absorptiometry (DEXA) scan to determine bone mineral content (BMC), bone

area (BA), bonemineral density (BMD), andbonemineral apparent density (BMAD) in the regionof

the lumbar spine. Blood was drawn to determine serum vitamin A, D, E, and K status and liver

function tests. The best forced expired volume in 1 sec (FEV1) for each patient in the 12 months

around the time of the scan was also documented. Patients with CFLD had slightly worse FEV1

(82� 20% vs. 91�16%, P¼0.05) and significantly higher alanine aminotransferase (65.5�35 IU/l vs. 30� 20 IU/l,P¼ 0.01) than thosewithCFalone. Themean lumbar spineBA,BMC,BMD,

and BMAD were not different between children with CFLD and CF. In conclusion, the presence of

significant liver disease in children with CF does not appear to be an additional risk factor for the

development of abnormal bone mineralization. Pediatr Pulmonol. 2006; 41:338–344.

� 2006 Wiley-Liss, Inc.

Key words: cystic fibrosis; liver fibrosis; portal hypertension; bone mineral density;

dual-energy X-ray absorptiometry scan.

INTRODUCTION

Cystic fibrosis (CF) is an autosomal-recessive disorderthat results from the inheritance of a defect in the cftr gene.This gene encodes for the cystic fibrosis transmembraneconductance regulator (CFTR) protein, which is presenton numerous epithelial surfaces. The clinical manifesta-tions of this condition include chronic suppurative lungdisease, exocrine pancreatic insufficiency, and biliarycirrhosis.1 As the median age of survival increases, othercomplications of this disorder are becoming apparent,such as lowbonemineral density (BMD).2–6 Some studiessuggest that increased pulmonary disease activity andpoor nutritional status have an adverse effect on thebone health of children with CF.2,6 It was proposed thattheBMDabnormalities of osteoporosis and osteopenia arelikely to be secondary to a combination of several factors.These include malnutrition, malabsorption, chronic pul-monary inflammation, hypogonadism, vitamin D and Kdeficiency, and glucocorticoid use.2,3,7,8 Several popula-tion studies (independent of CF) demonstrated a sig-nificant relationship between BMD Z-score and fracturerisk. A lumbar-spine BMDZ-score of less than�2 confersa 4–6-fold increase in the risk of sustaining a fracture.9–11

1Department of Gastroenterology and Nutrition, Royal Children’s Hospital,

Melbourne, Australia.

2Department of Paediatrics, University of Melbourne, Melbourne, Australia.

3Murdoch Children’s Institute, Melbourne, Australia.

4Department of Radiology, Royal Children’s Hospital, Melbourne, Australia.

5Department of Thoracic Medicine, Royal Children’s Hospital, Melbourne,

Australia.

6Department of Endocrinology, Royal Children’s Hospital, Melbourne,

Australia.

Grant sponsor: Nutricia, Australia; Grant sponsor: Murdoch Children’s

Institute.

*Correspondence to: Mark R. Oliver, M.D., FRACP, Department of

Gastroenterology and Clinical Nutrition, Royal Children’s Hospital,

Flemington Road, Parkville, Victoria 3052, Australia.

E-mail: mark.oliver@rch.org.au

Received 22 August 2005; Accepted 24 August 2005.

DOI 10.1002/ppul.20337

Published online 13 February 2006 in Wiley InterScience

(www.interscience.wiley.com).

� 2006 Wiley-Liss, Inc.

Although this information was obtained from peri- andpostmenopausal women, it is used as a guideline in theabsence of any other pediatric data. It is imperative thatwe attempt to identify high-risk subgroups with CF andaddress the issue of low BMD in a proactive manner, asthese patients run a significant risk of developing patho-logical fractures through the course of their life.

Abnormalities in adult bone mineralization occur in arange of chronic liver diseases associated with eitherfibrosis-cirrhosis (on a clinical, radiological, and/or histo-logical basis) or severe cholestasis (defined as serumbilirubin more than three times the upper limit of normalfor more than 6 months) due to a large variety of causes.12

As stated earlier, severe fibrosis and cirrhosis (with orwithout portal hypertension) was described in CF andprobably affects 10–20% of this population.13,14 Hence,it is possible that significant liver disease may well be anadditional risk factor for the development of BMDabnormalities in the CF population. The aim of our studywas to assess whether bone mineralization of the lumbarspine in patients with significant cystic fibrosis-relatedliver disease (CFLD) was different from that in patientswith CF alone. We also wished to examine the relation-ship between measurements of bone mineralization andnutrition and lung function in our entire study cohort.

PATIENTS AND METHODS

We identified and recruited 14 patients, aged between10–18 years, attending our CF outpatient department(310 patients in total), with clinical evidence of significantand established CFLD. All children with CFLD wereknown to have significant liver disease at the commence-ment of the study. Patients with minor elevations of liverenzymes that were not elevated for a 6-month period andwho had a normal abdominal examination were excluded.One child with portal hypertension was excluded due tothe coexistence of another malabsorptive comorbidity(celiac disease). For each patient with CFLD, a controlwith CF but without clinical or biochemical evidence ofliver dysfunction was prospectively recruited. Patients in

each group were matched for gender, body mass index(BMI), height, and bone age. Matching criteria were: forBMI score, �1; for height, �10 cm; and for bone age,within 12 months.The diagnosis of CFwas confirmed in all patients on the

basis of clinical manifestations and abnormal sweationtophoresis. Either 3-day fecal fat balance or the pre-sence of fat globules on fecal microscopy, in conjunctionwith a clinical history of steatorrhea, defined pancreaticinsufficiency.All patients in both groups had an analysis oftheir genotype. The diagnosis of significant liver diseasewas based on clinical assessment, the finding of apersistently elevated alanine aminotransferase (ALT)(beyond our normal laboratory limits), and ultrasonogra-phy. All children with CFLD had a hard liver edgeextending at least 2–3 cm beyond the right costal marginby itself or with the additional finding of firm splenome-galy, suggestive of portal hypertension. Ultrasonographywas used to assess hepatic size, texture and extent of thechange(s) (focal/diffuse), and splenic size. All CFLDpatients were screened for other causes of liver disease.This meant that blood was drawn for determination ofanti-nuclear antibody, anti-smooth muscle antibody, anti-liver-kidney-microsomal antibody (autoimmune hepati-tis), alpha 1-antitrypsin level, and phenotype. Hepatitis-Bsurface antigen, hepatitis-C antibody, ceruloplasmin andserum copper (Wilson’s disease), and urinary metabolicscreenswere also performed.Children in the control grouphad been examined at routine CF outpatient visits byM.R.O. every 3months for 12months before bone densitystudies were carried out. None of these patients hadclinically documented liver disease as set down by the CFFoundation Hepatobiliary Disease Consensus Group.13

All patients in this study (n¼ 26) underwent dual-energy X-ray absorptiometry (DEXA) scanning by thesame radiology technician, using the same densitometer(Delphi W, Hologic QDR 4500, Inc., Waltham, MA). Thelumbar spine alone was scanned, as we were required tominimize the radiation dose.Anterior- posterior scans (fanarray) were obtained for L1–L4. The manufacturer’ssoftware calculated BMC in grams, projected area in cm2,and areal BMD g/cm2 for each vertebral body. Thesevalues were used to obtain the average bone mineralcontent (BMC) and BMD of L1–L4. Volumetric bonemineral density (BMAD) was estimated from DEXAmeasurements, using published geometric correctionfactors.15 All scans were done at least 6 weeks after ahospital admission.Anthropometric measurements including weight (digi-

tal scale) and height (stadiometer) for each patient werecollected on the day of their DEXA scan. Following this,height-for-ageZ-scores,BMI, andBMIZ-scoreswere cal-culated, using the epidemiological software program EpiInfo 2000 (Centers for Disease Control and Prevention,Altanta, GA).

ABBREVIATIONS

BA Bone area

BMAD Bone mineral apparent density

BMC Bone mineral content

BMD Bone mineral density

BMI Body mass index

CF Cystic fibrosis

CFLD Cystic fibrosis-related liver disease

CFTR Cystic fibrosis transport regulator

CI Confidence interval

DEXA Dual-energy X-ray absorptiometry

FEV1 Forced expiratory volume in 1 sec

HAZ Height-for-age Z-score

SD Standard deviation

Bone Disease With CF-Related Liver Disease 339

Spirometry measurements were obtained from allpatients, and forced expiratory volume in 1 second (FEV1)was expressed as percentage predicted after adjustmentfor weight, height, and gender.16 The best results within12 months of the DEXA scan were used for comparison.Sputum cultures at time of DEXA scan were documented,as well as the number of admissions to hospital forintravenous antibiotics and intensive physiotherapy. Allpatients who had Pseudomonas aeruginosa and Staphy-lococcus aureus were on long-term oral antibiotics.Patients had liver function tests and serum vitamin Aand E (high-performance liquid chromatography), D(radioimmunoassay), and E assays and coagulationscreens at the time of the scan. Bone age measurements(done at the time of the DEXA scan) were based on themethod by Greulich and Pyle of comparing X-rays of theleft hand, using the standards available in Greulich andPyle.17Data concerning the use of ursodeoxycholic acid inchildren with liver disease and glucocorticoid therapy inboth patient groups were collected by historical recall andreview of chart and pharmacy records. Liver function dataat time of diagnosis of significant liver disease werecompared with data obtained at the time of the DEXAscan. This study was approved by the Ethics in HumanResearch Committee at the Royal Children’s Hospital,Melbourne.

Data Analysis

Normally distributed data were presented as mean�standard deviation. A comparison of chronological andbone age, height-for-age Z-scores, BMI Z-scores, FEV1

(%) vitamin levels, and liver function tests between thetwo patient groups was made using Student’s t-test.Univariate relationships between bone mineralizationmeasurements and continuous variables (height-for-ageZ-scores, BMI Z-scores, and FEV1) in the entire cohort(n¼ 26) were assessed by testing the significance of theSpearman rank correlation coefficient. Bone measure-ments were then modeled using multiple linear regressionto determine which of the significantly associatedvariables predicted bone mineral measurements whileallowing for the effects of the others. Fisher’s exact testwas performed to compare genetic, sputum culture, andcorticosteroid data.

RESULTS

Anthropometric and Pulmonary Function

Each of the two groups, CFLD and CF, consisted of10males and 3 females. Themean age of theCFgroupwas13.5� 1.7 years, and of the CFLD group, 14.1� 1.9 years(P¼ 0.20). The CF and CFLD patients were matched forbone age (CF group, 12.9� 1.9 years, vs. 12.9� 2.3 yearsin CFLD, P¼ 0.89), which acted as a surrogate for

pubertal staging. Height-for-age Z-scores (HAZ) weresimilar in both groups (CF group, �0.73� 1, vs. �1.1�1.3 in CFLD, P¼ 0.26). There was no difference innutritional status between the two patient groups asmeasured by BMI Z-scores (CF group, �0.44� 1.4, andCFLD, �0.73� 0.83, P¼ 0.57). However, pulmonaryfunction (FEV1(%)) was more impaired in patients withCFLD than in patients with CF alone (CFLD, 82� 20 vs.91� 16 with CF, P¼ 0.05).

Inpatient Admissions and Sputum Cultures

At the time of the DEXA scan, children with CF alonehad 4.3� 6.5 admissions vs. 4� 5.26 admissions inCFLDpatients (P¼ 0.88). Pseudomonas aeruginosa was iso-lated in 22 patients, one child had Staphylococcus aureus,two had no growth, and data were unavailable for onepatient. There were no significant differences betweenthe CF and CFLD group with regard to Pseudomonasaeruginosa infection.

Exocrine Pancreatic Function,Corticosteroid Use, and Genetic Status

All patients included in this study had pancreaticinsufficiency and were on both pancreatic enzymereplacement therapy and fat-soluble vitamin treatment,according to the guidelines set down by the CF NutritionConsensus Committee.18 Four patients received inhaledsteroids for a period of 3–6 months for the 2-year periodthat preceded the DEXA scan. There were two patientseach in the CFLD and CF group (no significance, NS).Eight of those with CFLD and 6 of 13 CF patients werehomozygous for D508 (NS).

Liver Function, Vitamin Assays,and Ultrasonography

An alternative etiology for liver disease was notidentified in any of the CFLD patients. Four patients withCFLD had a firm, enlarged liver without splenomegaly,and the remaining patients in this group had both hepaticenlargement and accompanying signs of portal hyperten-sion (splenomegaly) on clinical examination. None of thecontrol group had clinical findings consistent with thepresence of significant liver disease. Ultrasound exami-nations were performed only in children with clinicalevidence of hepatic involvement; the results are summar-ized in Table 1. Liver function tests, including themeasurement of conjugated bilirubin, gamma-glutamyl-transferase (GGT), alkaline phosphatase (ALP), albumin,and ALT, were performed in all 26 patients at the time ofthe DEXA studies, and are shown in Table 2. Total andconjugated bilirubin was in the normal range for bothgroups, and although both GGT and ALT were higher inchildren with liver disease, it was only the latter that was

340 Alex et al.

found to be significantly different. Serum 25-OH vitaminD levelswere similar in both groups andwithin our normallaboratory range of 50–160 nmol/l, as were vitamin E/lipid ratios, vitamin A levels, and coagulation profiles.

Effects of Ursodeoxycholic Acid onLiver Function Tests

CFLD children received ursodeoxycholic acid at a doseof 15–20 mg/kg per day, as per published guidelines.13

This was administered for a median period of 19 months(range, 0–48months) at the time of theDEXAscan. Therewere improvements in ALT (93� 50 IU/l at diagnosis ofliver disease vs. 63� 35 IU/l at DEXA scan,P¼ 0.02) andGGT (129.5� 130 IU/l at diagnosis of liver disease vs.75� 51 IU/L at DEXA scan, P¼ 0.04). There were nosignificant changes in conjugated bilirubin or albumin.

Bone Mineral Status

Table 3 summarizes DEXA-measured lumbar spinedata, including bone area, bone mineral content, bonemineral density of the lumbar spine expressed as a Z-score (corrected for age, gender, and bone age), and bone

mineral apparent density in the CF and CFLD groups.There were no significant differences between the twogroups when a paired analysis of the data was made. Asprevious studies2,3 showed a correlation between bonedensity and both disease activity and nutritional status,we assessed the data of all 26 patients to examine therelationship between bone density (BMC, BMD Z-scorecorrected for bone age, and BMAD) with lung function(FEV1) and anthropometric data (BMI Z-score and HAZ),using both univariate and multivariate analysis. Onunivariate studies, we found a correlation between BMCand height-for-age Z-score (r¼ 0.429, P¼ 0.029). Allother comparisons were found not to be significant.Following this, multiple regression analysis was per-formed, which confirmed these findings. The data aresummarized in Table 4.

DISCUSSION

The aim of our study was to determine whether thepresence of significant liver disease in children with CFwas associated with abnormalities in bonemineralization.We focused on patients who had clinical and sonographic

TABLE 1— Biochemical Liver Function Tests and Vitamin Levels

Liver function tests and vitamin levels CF (n¼ 13) CFLD (n¼ 13) Reference range P value

ALT (IU/l) 30 (� 20) 63.5 (� 35) <55 0.01*

GGT (IU/l) 41.8 (� 50) 75.0 (� 51) 0–40 0.16

ALP (IU/l) 275 (� 108) 301.0 (� 115) 100–350 0.75

Albumin (g/l) 41 (� 2.4) 38.6 (� 4) 33–47 0.26

PT (sec) 14.26 (� 1.23) 15.37 (� 2.29) 12–16 0.18

APTT (sec) 35.3 (� 6.2) 34.0 (� 4.66) 27–38 0.6

Vitamin A (mmol/l) 1.1 (� 0.5) 1.19 (� 0.56) 0.9–1.7 0.48

Vitamin E/lipid (mmol/mmol) 4.12 (� 2) 3.87 (� 1.66) 0.9–7.1 0.73

25-OH vitamin D (nmol/l) 53 (� 19) 59 (� 23) 22.5–93.8 0.49

*Statistically significant.

TABLE 2— Abdominal Ultrasound Features of Patients With CFLD

Age (year) Gender (M/F)

Liver

SplenomegalySize Texture Diffuse/focal

10.25 F Large Abnormal Diffuse No

13 F Normal Abnormal Focal No

14.75 F Normal Abnormal Diffuse Yes

14 M Large Abnormal Diffuse Yes

13 M Large Abnormal Focal Yes

12.5 M Normal Abnormal Diffuse Yes

13.75 M Normal Abnormal Diffuse Yes

14.5 M Normal Abnormal Diffuse Yes

14.75 M Large Abnormal Diffuse No

15 M Normal Abnormal Diffuse No

14.5 M Normal Abnormal Diffuse Yes

15.25 M Large Abnormal Diffuse Yes

18.5 M Small Abnormal Diffuse Yes

Bone Disease With CF-Related Liver Disease 341

evidence of hepatic fibrosis and/or portal hypertension.This group was chosen as it appeared logical to assumethat if liver disease did act as an independent risk factor forabnormal bone density, then it would be patients at themore severe end of the spectrum (but not end-stage) whoought to be affected.We found no evidence of a significantdifference in BMC, BMD, BMD Z-score, or BMAD bet-ween patients with CFLD and thosewith CF alone. Further-more, the use of glucocorticoid therapy, gene status, andvitamin levels did not appear to be different in either group.Although there were no differences between the two

groups, as a whole they had significant abnormalitiesin bone mineralization compared to otherwise normalchildren. Patients in both groups (CFLD and CF) wereosteopenic (BMD Z-score of �1 to �2.5). The abnorm-alities that we identified inBMDZ-score andBMADweresimilar to those obtained by other investigators.2,3 Datafrom the entire cohort (n¼ 26) suggested that BMCcorrelated well with height-for-age Z-score. This was notaltogether surprising, as BMC is directly related to thesize of the bone, and this would undoubtedly be linked toheight. Further associations between measures of bonemineralization and nutritional markers (BMI Z-score) andlung function (FEV1)were not found. These findings are atodds with the data of others,2,3,6 who found a correlationbetween BMD, and nutritional status and pulmonarydisease activity. The relatively small number of patients inour study might explain this.

As in many previous studies,2–8 bone mass and densitymeasurements were estimated using DEXA. This pro-vides a measure of areal BMD that represents the integral(cortical and trabecular) amount of bone mineral withinthe bone envelope of the respective region of the skeletonscanned, divided by its projected bone area. As the depth(and therefore the volume) of the scanned bone is notmeasured, areal BMD fails to distinguish betweenchanges in mineral density and bone size in growingchildren. This is an important consideration, as chronicchildhood illnesses such as CF and liver disease mayimpair growth and mineralization of the skeleton. In orderto minimize the influence of bone size on the DEXA-measured BMC and BMD and BMAD, we recruitedgender-, bone age-, and height-matched patients in ourstudy. The parameter of bone age rather than chronolo-gical age was chosen as it acts as a surrogate marker forpuberty, which on its own can influence bone mineralstatus. The close matching of our patients would decreasethe possibility of confounding factors contributing to theresults.Clinical examination remains the gold standard against

which other modalities for the diagnosis of CF-relatedliver disease are judged. We classified patients as havingsignificant liver disease on the basis of finding a palpableenlarged liver with a hard edge and an enlarged spleenindicative of portal hypertension.13,14 We then confirmedour finding using ultrasound and serum markers of

TABLE 3— Bone Density Measurement for Lumbar Spine in CF and CFLD Groups

Bone density measurement CF (n¼ 13) CFLD (n¼ 13)

Mean of paired

difference CI of difference P value

BA (cm2) 50.2 (� 7.84) 48.35 (� 6.42) 1.9 �1.855 to 5.63 0.29

BMC (g) 34.57 (� 10.8) 31.36 (� 8.1) 3.21 �0.352 to 6.733 0.073

BMD (g/cm2) 0.68 (� 0.14) 0.64 (� 0.11) 0.038 �0.037 to 0.11 0.29

BMAD (g/cm3) 0.096 (� 0.019) 0.093 (� 0.015) 0.0035 �0.009 to 0.016 0.56

BMD Z-score corrected for bone age �1.515 (� 1.18) �1.946 (� 1.036) 0.4308 �0.354 to 1.22 0.26

TABLE 4— Multiple Regression Analysis for Lumbar-Spine BMC, BMD Z-Score Correctedfor Bone Age, and BMAD

Variable Coefficient SE P value Lower 95% CI Upper 95% CI

BMC (n¼ 26)

BMI Z-score �0.930 2.00 0.65 �5.108 3.254

FEV1 0.016 0.119 0.38 �0.142 0.355

HAZ 4.257 1.95 0.04* 0.183 8.33

BMD Z-score corrected

for bone age (n¼ 26)

BMI Z-score 0.039 0.224 0.86 �0.43 0.508

FEV1 0.011 0.01 0.46 �0.018 0.038

HAZ 0.353 0.219 0.12 �0.104 0.81

BMAD (n¼ 26)

BMI Z-score �0.001 0.003 0.76 �0.0083 0.0062

FEV1 0.0003 0.0002 0.16 �0.0001 0.0007

HAZ 0.003 0.003 0.32 �0.003 0.01

*Statistically significant.

342 Alex et al.

disturbed liver function (elevated aminotransaminase). Itis likely that our patients were correctly categorized ashaving significant liver disease on the basis that 69%had clinical and radiological evidence of portal hyper-tension.

Others studied large cohorts of CF patients, and insubgroup analysis did not find an association betweenthe presence of liver disease and measurements of bonemineralization.2,6 Buntain et al.2 assessed 10 patientswith multilobular cirrhosis and portal hypertension in apopulation of 153 adults and children with CF, and foundno association of lumbar-spine BMD and the presence ofliver disease. Conway et al.6 studied 114 adolescents andadults with CF. They defined liver disease by abnormalultrasound findings and persistently raised alkaline phos-phatase or alanine transaminase levels. Using this methodof categorizing patients as having liver disease, theyidentified 61 patients who fulfilled their criteria. Uni-variate analysis did not demonstrate a significant associa-tion betweenBMDmeasurements and liver disease. Thesedata support our findings that significant CFLD by itself isnot associated with an increased risk of developing severebone-density abnormalities.

Our findings are at variancewith studies in adults whichsuggest that the presence of cirrhosis and/or severecholestasis are important predictors of bone density.12

Osteoporosis was reported in 16–43% of adults withcirrhosis, with many of these patients sustaining patholo-gical fractures.19,20 Most adult studies of the effect ofcholestasis on bone mineralization have drawn onpatients with either primary biliary cirrhosis or sclerosingcholangitis. Osteoporosis was identified in 13–35% ofthese groups.21,22 Although some of these studies sug-gested that cholestasis itself is a risk factor for osteo-porosis, this may just be a reflection of the coexistence ofcirrhosis.

Why have we failed to define a similar relationshipbetween the presence of significant liver disease in child-ren with CF and abnormalities in bone mineralization?There are several possible explanations. First, none of ourpatients were significantly cholestatic (i.e., serum bilir-ubin more than three times the upper limit of normal forgreater than 6 months), and perhaps this is an importantdeterminant of abnormal bone mineralization. Second,all our patients were prescribed ursodeoxycholic acid, anagent that is said to reduce cholestasis and by this actionmay have a protective effect on bone density. Our datasuggest that thismay be a confounding factor, as bothALTand GGT improved on treatment with ursodeoxycholicacid. However, the disease process (in this case, fibrosis-cirrhosis) was already well-established before treatmentwith ursodeoxycholic acid was commenced. Third, it ispossible that substantial derangement in bone mineraliza-tion may take a much longer period to develop. This canonly be determined by longitudinal studies.

Our data, coupled with those of others,2,6 suggest thatthere is no clear relationship between abnormalitiesin bone mineralization and significant CF-related liverdisease. However, most of the studies have been relativelysmall and may not be adequately powered. Under thesecircumstances, a large multicenter study would providea definitive answer. This study reaffirms the issue ofsignificant bone-density abnormalities existing in our CFpopulation. It is possible that with increasing survival dueto new therapies, osteoporosis may add to the economicand personal burden for these patients.

REFERENCES

1. Gaskin K. Cystic fibrosis. In: Walker WA, Durie PR, Hamilton JR,

Walker-Smith JA, Watkins JB, editors. Pediatric gastrointestinal

disease. Volume 2. Hamilton, Ontario: Mosby; 2004;1606–1623.

2. Buntain HM, Greer RM, Schluter PJ, Wong JCH, Batch JA, Potter

JM, Lewindon PJ, Powell E, Wainwright CE, Bell SC. Bone

mineral density in Australian children, adolescents and adult with

cystic fibrosis: a controlled cross sectional study. Thorax 2004;59:

149–155.

3. Sood M, Hambleton M, Super M, Fraser WD, Adams JE, Mughal

MZ. Bone status in cystic fibrosis. Arch Dis Child 2001;84:516–520.

4. Henderson RC, Madsen CD. Bone density in children and

adolescents with cystic fibrosis. J Pediatr 1996;128:28–34.

5. Haworth CS, Selby PL, Webb AK, Dodd ME, Musson H, Niven

R, Economou G, Horrocks AW, Freemont AJ, Mawer EB, Adams

JE. Low bone mineral density in adults with cystic fibrosis.

Thorax 1999;54:961–967.

6. Conway SP, Morton AM, Oldroyd B, Truscott JG, White H,

Smith AH, Haigh I. Osteoporosis and osteopenia in adults and

adolescents with cystic fibrosis: prevalence and associated

factors. Thorax 2000;55:798–804.

7. Ionescu AA, Nixon LS, Evans WD, Stone MD, Lewis-Jenkins V,

Chatham K, Shale DJ. Bone density, body composition and

inflammatory status in cystic fibrosis. Am J Respir Crit Care Med

2000;162:789–794.

8. Aris R, Lester G, Ontjes D. Treatment of bone disease in cystic

fibrosis. Curr Opin Pulm Med 2004;10:524–530.

9. Kroger H, Huopio J, Honkanen R, et al. Prediction of fracture risk

using axial bone mineral density in a perimenopausal population:

a prospective study. J Bone Miner Res 1995;10:302–306.

10. Jones IE, Taylor RW, Williams SM, et al. Four year gain in bone

mineral in girls with and without past forearm fractures: a DXA

study. Dual energy x-ray absorptiometry. J Bone Miner Res 2002;

17:1065–1072.

11. Goulding A, Jones IE, Taylor RW, et al. Bone mineral density and

body composition in boys with distal forearm fractures: a dual

energy x-ray absorptiometry study. J Pediatr 2001;139:509–515.

12. Collier JD, Ninkovic M, Compston JE. Guidelines on the

management of osteoporosis associated with chronic liver

disease. Gut [Suppl] 2002;50:1–9.

13. Sokol RJ, Durie PR. Recommendations for management of liver

and biliary tract disease in cystic fibrosis. J Pediatr Gastroenterol

Nutr 1999;28:1–13.

14. Tanner MS, Taylor CJ. Liver disease in cystic fibrosis. Arch Dis

Child 1995;72:281–284.

15. Wren TAL, Liu X, Pitukcheewanont P, Gilsanz V, members of

Bone Mineral Density in Childhood Study. Bone acquistion in

healthy children and adolescents: comparisons of dual-energy

X-ray absorptiometry and computed tomography measues. J Clin

Endocrinol Metab 2005;90:1925–1928.

Bone Disease With CF-Related Liver Disease 343

16. Crapo RO, Morris AH, Gardner RM. Reference spirometric

values using techniques and equipment that meet the ATS

recommendations. Am Rev Respir Dis 1981;123:659–664.

17. Greulich WW, Pyle SI. Radiographic atlas of skeletal develop-

ment of the hand and wrist. 2nd ed. Stanford, CA: Stanford

University Press; 1959.

18. Ramsey BW, Farrell PM, Pencharz P. Nutritional assessment and

management in cystic fibrosis: a consensus report. The Consensus

Committee. Am J Clin Nutr 1992;55:108–116.

19. Monegal A, Navasa M, Guanabens N, et al. Osteoporosis and

bone mineral metabolism disorders in cirrhotic patients referred

for orthotopic liver transplantation. Calcif Tissue Int 1997;60:

148–154.

20. Angulo P, Therneau TM, Jorgensen RA, et al. Bone disease in

patients with primary sclerosing cholangitis: prevalence, severity

and prediction of progression. J Hepatol 1998;29:729–735.

21. Eastell R, Dickson ER, Hodgson SF, et al. Rates of vertebral bone

loss before and after liver transplantation in women with primary

biliary cirrhosis. Hepatology 1991;14:296–300.

22. Guanabens N, Pares A, Marinosos L, et al. Factors influencing the

development of metabolic bone disease in primary biliary

cirrhosis. Am J Gastroenterol 1990;85:1356–1362.

344 Alex et al.