gotta catch’em all! pokémon go and physical activity · manuscript id bmj.2016.035580.r1 article...
TRANSCRIPT
-
Confidential: For Review O
nly
Gotta Catch’em All! Pokémon GO and Physical Activity: Difference in Differences Study
Journal: BMJ
Manuscript ID BMJ.2016.035580.R1
Article Type: Christmas
BMJ Journal: BMJ
Date Submitted by the Author: 07-Nov-2016
Complete List of Authors: Howe, Katherine; Harvard T.H. Chan School of Public Health, Epidemiology; Harvard T.H. Chan School of Public Health, Social and Behavioral Sciences Suharlim, Christian; Harvard T.H. Chan School of Public Health, Center for Health and Decision Science, Department of Health Policy and Management
Ueda, Peter; Harvard T.H. Chan School of Public Health, Department of Global Health and Population Howe, Daniel; n/a Kawachi, Ichiro; Harvard School of Public Health, Department of Society Human Development and Rimm, Eric; HSPH
Keywords: physical activity, Pokemon Go
https://mc.manuscriptcentral.com/bmj
BMJ
-
Confidential: For Review O
nly
1
Gotta Catch’em All! Pokémon GO and Physical Activity: Difference in
Differences Study
Katherine B Howe, MPH1,2*
; Christian Suharlim, MD MPH3*
; Peter Ueda, MD PhD4,5*
; Daniel
Howe, BA; Ichiro Kawachi, MBChB PhD2; Eric B. Rimm, ScD
1,6,7
1. Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA
2. Department of Social and Behavioral Sciences, Harvard T.H. Chan School of Public
Health, Boston, MA
3. Center for Health and Decision Science, Department of Health Policy and Management,
Harvard T.H. Chan School of Public Health, Boston, MA
4. Department of Global Health and Population, Harvard T.H. Chan School of Public
Health, Boston, MA
5. Clinical Epidemiology Unit, Institute of Medicine, Solna, Karolinska Institutet, Sweden
6. Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston, MA
7. Channing Division of Network Medicine, Department of Medicine, Brigham and
Women’s Hospital, Harvard Medical School, Boston, MA
*These authors contributed equally to this work.
Correspondence to: Katherine Howe [email protected]
Page 1 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
2
Abstract:
Objectives To estimate the effect of playing Pokémon GO on the number of steps taken per day
up to 6 weeks after installing the game.
Design Cohort study using online survey data.
Participants Amazon Mechanical Turk survey participants (n=1,182) residing in the US, aged
18 to 35 years and using iPhone 6 series smartphones.
Main outcome measures Number of daily steps taken each of the four weeks before and six
weeks after installation of Pokémon GO, automatically recorded in the Health application of the
iPhone 6 series smartphones, and reported by the participants. A difference-in-difference
regression model was used to estimate the change in daily steps in Pokémon players compared
with non-players.
Results 560 (47.4%) of the survey participants reported playing Pokémon Go and walked on
average 4,256 steps (SD=2,697) per day in the four weeks prior to installation of the game. The
difference-in-difference analysis showed that the daily average steps increased for Pokémon Go
players during the first week of installation by 955 additional steps (95% CI = 697-1,213), and
then this increase gradually attenuated over the subsequent five weeks. By the sixth week after
installation, daily steps had gone back to pre-installation levels. There was no significant effect
modification of Pokémon GO by sex, age, race group, weight status, or walkability of the area of
residence.
Conclusions Pokémon Go was associated with an increase in the daily number of steps after
installation of the game. However, the association was moderate and was no longer observed
after six weeks.
Page 2 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
3
Introduction
Pokémon GO is an augmented reality game in which players search real-world locations for
cartoon characters appearing on their smartphone screen. Since its launch in July 2016, the game
has been downloaded over 500 million times worldwide. Because walking is encouraged while
playing, Pokémon GO has been suggested to increase physical activity and improve public health,
but these claims have been based on anecdotal evidence.[1–3]
We used an online survey and automatically recorded step count data from iPhone devices to
estimate the change in daily steps following installation of Pokémon GO among young adults in
the US.
Methods
Study population
We conducted an online survey using the Amazon Mechanical Turk (MTurk), between August
1st and 31st, 2016. MTurk is a website for recruitment of online workers who receive a small
compensation for completing tasks including responding to surveys. It is widely used for
research purposes: it allows for fast recruitment of diverse study populations and the reliability of
the answers is similar to that of traditional survey methods.[4–6]
We recruited survey participants who were aged 18 to 35 years, resided in the US, and used an
iPhone 6 series smartphone, because these devices automatically record the number of steps
taken while carrying the device. Participants were paid 2 USD for completion of the survey. Of
the 2,225 individuals who responded to the survey, we excluded 115 participants who had
downloaded Pokémon GO but not achieved a “trainer level” of 5 or more which takes around 2 h
of walking to reach. At this level, functions integral to the game are unlocked, including joining
a team and competing for Pokemon Gyms. We further excluded 923 participants who did not
complete the full survey likely due to the complexities of multiple screen captures necessary for
participation. Our final study population consisted of 1,182 individuals.
Page 3 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
4
Covariates
Survey participants were asked about their zip code of residence, education level, household
income, marital status, and weight and height from which body mass index (BMI) in kg/m2 was
calculated. Participants who reported that they played Pokémon Go were asked to upload a
screenshot (picture of the screen content) of their Pokémon Go application which shows the
installation date. We matched the zip code of residence with a validated neighborhood
walkability index (Walk ScoreⓇ), which is categorized into four groups (Car-Dependent,
Somewhat Walkable, Very Walkable, and Walker’s Paradise).[7,8]
Participants were asked to upload screenshots showing their number of steps/day while carrying
their iPhone, for each day during the two months preceding the date of study participation. This
information is recorded by default in the ‘Health’ application of iPhone 6 series, using the
device’s built-in accelerometer.
Statistical analyses
We compared the average number of steps taken per day for each of the four weeks prior to
installing the game vs. the steps during each of the six weeks after installation in Pokémon
players and non-players respectively. For non-players, the number of steps was compared before
and after the median date of installation among Pokémon-playing participants. To examine the
causal association between Pokémon GO and daily steps, we conducted a difference-in-
difference analysis. This method attempts to mimic an experimental research design by
comparing the change in the outcome over time in the treated vs. control group, under the
assumption that the differences between the groups would have remained constant under no
treatment. We used a multivariate regression model; the estimate of change in steps after
Pokémon GO was obtained from interaction indicators for Pokémon GO players and each week
after installation of the game. The model also adjusted for time-invariant differences between
players and non-players, and fluctuations in steps across weeks before and after installation. We
calculated robust standard errors to account for the repeated measurements within the same
individuals.
Page 4 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
5
In separate models, we assessed effect modification of Pokémon GO by sex, age group (18-24
years and 25-35 years), household income (
-
Confidential: For Review O
nly
6
associated with an increase in steps of 955 (95% CI: 697 - 1,213) during the first week, the effect
was gradually attenuated over the subsequent weeks and by week six it was not significant (130
[-593 – 851]) (Figure 2 and Supplementary Table 1).
The analyses on effect modification of Pokémon GO by baseline variables showed no significant
interaction between the variables, Pokémon GO and week since installation of the game.
(Supplementary Table 2). Adjusting for differences in time trends in steps by baseline variables
did not materially affect the results (Supplementary Table 3).
Discussion
In this study on the impact of Pokémon GO on physical activity in young US adults, we found
that the game was associated with an increase in daily steps of 955 (95% CI= 697-1,213) in the
first week after installation. The association was attenuated during the subsequent weeks and was
no longer significant by the sixth week. Assuming steps of 0.8 m at a pace of 4 km/h,[11] the
change in steps in the first week would translate into 11 minutes of additional walking per day,
which is around half of the WHO recommendation of 150 minutes/week.[12]
Pokémon GO has been suggested to improve public health by promoting physical activity.[1–3]
Our results, however, indicate that the health impact of Pokémon GO may be moderate in our
study population. Interventions designed to increase walking typically increase steps by 2500
steps/day.[13] Even if smaller amounts of physical activity may also be important for health
outcomes,[12,14] the increase in steps from Pokémon GO, as with many physical activity
interventions,[15] was not sustained over time. Pokémon GO may also entail risks such as
injuries and traffic accidents.[16,17]
Strengths of this study include the use of automatically recorded steps data from the iPhone
which has been shown to accurately track step counts,[18] validation of Pokémon GO
installation and date using screenshots, and a study population from a wide range of socio-
demographic groups and regions of the US.
Page 6 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
7
Our study has limitations. First, steps were only recorded when the iPhone was carried. This may
lead to overestimation of the effect of Pokémon Go because physical activity could occur when
not carrying the phone (e.g. football) whereas all Pokémon-related activity was recorded. Second,
although our results did not change materially when adjusting for a number of baseline variables,
the difference-in-difference analysis is susceptible to confounding from changes in daily steps
that were unique to Pokémon-players but independent of Pokémon GO. Third, our study
population was limited to iPhone-users in the US, and MTurk workers are not representative of
the general population, e.g. in terms of sex (overrepresentation of women), political views and
personality.[6] Generalizability of our results may therefore be limited.
While the association between Pokémon GO and change in steps was short-lived in our study,
there may be individuals who sustain increased physical activity through the game. The effect of
Pokémon GO on physical activity may also be different in children which were not included in
our study. There may also be other potential benefits, such as increased social connectedness and
mood improvement.
Page 7 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
8
What is already known on this topic
● Pokémon GO is an augmented reality game overlaying graphics on the real world using
the camera function of smartphones.
● The game has been downloaded over 500 million times, and it is available in over 100
countries and regions.
● The game requires walking and has been suggested to increase physical activity.
What this study adds
● We estimated the changes in physical activity associated with playing Pokémon GO.
● Pokémon GO was associated with an increase in daily number of steps by 955 (95%
confidence interval = 697-1,213) steps during the first week after installation of the game.
The increase was attenuated with time and was no longer significant by the sixth week
after installation.
● Pokémon GO was associated with a moderate increase in daily steps that was not
sustained over time. The impact of Pokémon GO on physical activity in the broad
population may be limited.
References
1 McCartney M. Margaret McCartney: Game on for Pokémon Go. BMJ 2016;354.
2 Is Pokémon Go the answer to America’s obesity problem? Guardian. 2016 Jul 13.
3 University of Leicester. Press release: Pokémon Go could ease type 2 diabetes burden. 25
Jul 2016.
4 Berinsky AJ, Huber GA, Lenz GS. Evaluating online labor markets for experimental
research: Amazon.com’s mechanical turk. Polit Anal 2012;20:351–68.
doi:10.1093/pan/mpr057
5 Paolacci G, Chandler J, Ipeirotis PG. Running experiments on amazon mechanical turk.
Judgm Decis Mak 2010;5:411–9. doi:10.2139/ssrn.1626226
6 Paolacci G, Chandler J. Inside the Turk: Understanding Mechanical Turk as a participant
pool. Curr Dir Psychol Sci 2014;23:184–8. doi:10.1177/0963721414531598
7 Carr LJ, Dunsiger SI, Marcus BH. Validation of Walk Score for estimating access to
walkable amenities. Br J Sports Med 2011;45:1144–8. doi:10.1136/bjsm.2009.069609
8 Duncan DT, Aldstadt J, Whalen J, et al. Validation of walk score for estimating
neighborhood walkability: an analysis of four US metropolitan areas. Int J Environ Res
Public Health 2011;8:4160–79. doi:10.3390/ijerph8114160
Page 8 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
9
9 Wooldridge J. Econometric analysis of cross section and panel data. MIT Press 2002.
10 StataCorp. Stata statistical software: release 12. StataCorp LP, 2012. College Town, TX.
11 Knoblauch R, Pietrucha M, Nitzburg M. Field Studies of Pedestrian Walking Speed and
Start-Up Time. Transp Res Rec J Transp Res Board 1996;1538:27–38. doi:10.3141/1538-
04
12 WHO: Global recommendations on physical activity for health. World Health
Organisation. 2010.
13 Bravata DM, Smith-Spangler C, Sundaram V, et al. Using Pedometers to Increase
Physical Activity and Improve Health. JAMA 2007;298:2296.
doi:10.1001/jama.298.19.2296
14 Tanasescu M. Exercise Type and Intensity in Relation to Coronary Heart Disease in Men.
JAMA 2002;288:1994. doi:10.1001/jama.288.16.1994
15 Teixeira PJ, Carraça E V, Markland D, et al. Exercise, physical activity, and self-
determination theory: A systematic review. Int J Behav Nutr Phys Act 2012;9:78.
doi:10.1186/1479-5868-9-78
16 Ayers JW, Leas EC, Dredze M, et al. Pokémon GO—A New Distraction for Drivers and
Pedestrians. JAMA Intern Med 2016;11:e0159885. doi:10.1001/jamainternmed.2016.6274
17 Serino M, Cordrey K, McLaughlin L, et al. Pokémon Go and augmented virtual reality
games: a cautionary commentary for parents and pediatricians. Curr Opin Pediatr
2016;28:673–7. doi:10.1097/MOP.0000000000000409
18 Case MA, Burwick HA, Volpp KG, et al. Accuracy of Smartphone Applications and
Wearable Devices for Tracking Physical Activity Data. JAMA 2015;313:625.
doi:10.1001/jama.2014.17841
Page 9 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
10
Contributions: KBH, CS, and PU contributed equally and are co-first authors. KBH, CS, and DH
conceived and designed the project and collected data. CS, PU, and KBH analyzed the data.
KBH, CS, PU, and DH wrote the first draft of the manuscript. All authors contributed to the
manuscript. IK and EBR provided methodological guidance and financial support for data
collection. EBR is the guarantor.
Funding: This research received no specific grant from any funding agency in the public,
commercial, or not-for-profit sectors.
Competing interests: All authors declare: no financial relationships with any organizations that
might have an interest in the submitted work and no other relationships or activities that could
appear to have influenced the submitted work.
Ethical approval: This study received institutional review board exemption approval on August
1st 2016 from the Harvard T.H. Chan School of Public Health (Protocol # IRB 16-1243).
Transparency: The manuscript’s guarantor affirms that the manuscript is an honest, accurate, and
transparent account of the study being reported; that no important aspects of the study have been
omitted; and that any discrepancies from the study as planned have been explained
Data sharing: Statistical code and dataset are available from the corresponding author.
Page 10 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
Table 1 Characteristics of Pokemon Go-playing and non-playing survey participants. Data are shown in number (%) unless
otherwise indicated.
Not Players (n=622) Pokemon GO Players (n=560)
Date for Pokemon Go installation,
median (IQR) . July 8th (July 7 - 11th)
Mean age in years (SD) 27.3 (4.5) 25.6 (4.3)
Age group
18-24 years 188 (30.2) 242 (43.2)
25-35years 434 (69.8) 318 (56.8)
Sex
Female 453 (72.8) 390 (69.6)
Race
Asian 32 (5.1) 38 (6.8)
White 491 (78.9) 430 (76.8)
Black 68 (10.9) 36 (6.4)
Other 31 (5.0) 56 (10.0)
Education
High school or less 29 (4.7) 51 (9.1)
Some university 208 (33.5) 237 (42.5)
College degree 267 (43) 207 (37.1)
Graduate/professional degree 117 (18.8) 63 (11.3)
Marital status
Single 336 (54) 365 (65.2)
Married/cohabitating 256 (41.2) 174 (31.1)
Divorced/separated/widowed 30 (4.8) 21 (3.8)
Household income
-
Confidential: For Review O
nly
Page 12 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
Figure 1 Average number of daily steps and 95% confidence intervals by week before and after installation of Pokemon
GO.
Average daily steps by week for non-Pokemon players are compared before and after the median installation date among
players (July 8th 2016)
2000
3000
4000
5000
6000
7000
Average steps per day
-4 -3 -2 -1 1 2 3 4 5 6
Week since installation
Pokemon Go players Non-players
Page 13 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nly
Figure 2 Average daily steps before and after installation of Pokemon GO in Pokemon GO players and non-players.
Among non-players, steps were compared before and after the median date of installation of the game among Pokemon GO-
playing participants. Confidence intervals are estimated using a difference-in-difference regression model. The full model is
shown in Appendix Table 1.
Page 14 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nlySupplementary Table 1 Results from the difference-in-difference model used to estimate the
association between Pokemon GO playing and change in daily steps for each week after installation
of the game.
* Week 1 represents the first week since installation of Pokemon GO. Week -4 represents four weeks before
installation of the game.
Coefficient (95% CI)
Pokemon Go player (ref = non-player) 114 (-212 to 440)
Week (ref= week -4)*
Week -3 60 (-64 to 184)
Week -2 -88 (-224 to 48)
Week -1 -64 (-200 to 73)
Week 1 -111 (-300 to 78)
Week 2 -190 (-363 to -17)
Week 3 -130 (-314 to 54)
Week 4 -225 (-422 to -28)
Week 5 -552 (-785 to -320)
Week 6 -425 (-1065 to 216)
Interaction term between Pokemon GO and week
since installation of the game*
Week 1 955 (697 to 1213)
Week 2 906 (647 to 1164)
Week 3 544 (280 to 808)
Week 4 446 (169 to 722)
Week 5 381 (43 to 720)
Week 6 130 (-593 to 853)
Page 15 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review Only
Supplementary Table 2 Interaction terms between baseline variables, Pokemon GO and week since installation of the game. Separate regression models have been
used for each baseline variable with average daily steps per week as the dependent variable and the following independent variables: the baseline variable, week (four
weeks prior to Pokemon GO installation, and six weeks after installation), and interaction terms between week since installation of Pokemon GO and Pokemon GO,
and the interaction terms shown in the table.
Interaction term with Pokemon Go, baseline variable, and week since installation of the game
Baseline variable Week 1 Week 2 Week 3 Week 4 Week 5 Week 6
Sex (ref = men) Women 47 (-602 to 696) 5 (-643 to 653) 177 (-461 to 815) 436 (-196 to 1069) 110 (-581 to 801) 473 (-356 to 1303)
Age group (ref = 18-24 years) 25-35 years 5 (-564 to 574) 57 (-524 to 638) 89 (-466 to 643) -56 (-614 to 502) 458 (-141 to 1057) 389 (-345 to 1122)
Race (ref = Asian) White -608 (-1658 to 442) -1079 (-2271 to 112) -802 (-2013 to 409) -1038 (-2235 to 159) -1214 (-2541 to 112) -477 (-2045 to 1092)
Black -632 (-1917 to 653) -772 (-2201 to 657) -662 (-2131 to 806) 17 (-1525 to 1559) -794 (-2555 to 968) -1153 (-2935 to 629)
Other -807 (-2166 to 552) -245 (-1615 to 1125) -195 (-1654 to 1264) -318 (-1939 to 1303) 289 (-1546 to 2125) 4440 (3800 to 5081)
Household income (ref = /=100k 70 (-703 to 843) 209 (-630 to 1048) 166 (-636 to 969) 143 (-633 to 919) 122 (-807 to 1050) 586 (-481 to 1653)
Education (ref= no college degree) College degree or
higher
48 (-530 to 626) 54 (-532 to 640) 1 (-556 to 559) 436 (-122 to 995) 221 (-384 to 826) 94 (-635 to 823)
Weight status (ref= BMI/=25 kg/m2 -340 (-928 to 247) -307 (-904 to 291) -11 (-575 to 554) -285 (-854 to 284) 327 (-284 to 937) 461 (-268 to 1191)
Page 16 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review Only
Walkability index (ref=
Car-dependent or Somewhat
walkable)
Very walkable or
Walker's Paradise
192 (-685 to 1069) -280 (-1056 to 497) -339 (-1073 to 394) -117 (-935 to 701) -87 (-963 to 789) -334 (-1332 to 664)
Urbanity (ref=rural) Suburban 268 (-648 to 1183) 374 (-538 to 1285) 488 (-368 to 1343) 472 (-435 to 1378) 667 (-262 to 1597) 234 (-888 to 1357)
Urban -198 (-829 to 433) 152 (-487 to 790) 157 (-452 to 765) 144 (-472 to 759) 321 (-338 to 981) 365 (-451 to 1180)
Page 17 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nlySupplementary Table 3 Regression analysis using average daily steps per week as the dependent variable and
adjusted for baseline covariates and including interaction terms between baseline variables in week since
installation of Pokemon Go. Interaction terms between Pokemon Go and week since installation of the game show
the difference-in-difference in daily steps after Pokemon Go installation in players vs. non-players.
Variable Coefficient (95% CI)
Baseline variables Sex (ref = men) Women -2009 (-2392 to -1626)
Age group (ref = 18-24 years) 25-35 years 51 (-285 to 387)
Race (ref = Asian) White 398 (-262 to 1057)
Black -273 (-1066 to 521)
Other 32 (-729 to 792)
Household income (ref = /=100k 55 (-377 to 487)
Marital status (ref = single or divorced) Married/partnered -463 (-837 to -89)
Region (ref=Midwest) Northeast 47 (-440 to 533)
South -483 (-907 to -59)
West -163 (-556 to 231)
Education (ref= no college degree) College degree or higher 352 (33 to 671)
Weight status (ref= BMI/=25 kg/m2 -533 (-860 to -206)
Walkability index (ref= Car-dependent or
Somewhat walkable)
Very walkable or Walker's
Paradise
1002 (509 to 1495)
Page 18 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nlyUrbanity (ref=rural) Suburban 222 (-191 to 635)
Urban 360 (-113 to 833)
Pokemon Go player (ref = non-player) Pokemon player 66 (-244 to 376)
Week (ref =Week -4)* Week -3 66 (-58 to 190)
Week -2 -91 (-227 to 45)
Week -1 -67 (-202 to 69)
Week 1 175 (-649 to 999)
Week 2 618 (-174 to 1409)
Week 3 803 (17 to 1589)
Week 4 580 (-285 to 1444)
Week 5 176 (-932 to 1283)
Week 6 -296 (-2066 to 1473)
Interaction terms
with week since
Pokemon Go
installation
Sex (ref = men) Female sex Week 1 216 (-91 to 522)
Week 2 33 (-275 to 341)
Week 3 -95 (-412 to 222)
Week 4 137 (-198 to 471)
Page 19 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nlyWeek 5 -29 (-449 to 392)
Week 6 -231 (-1435 to 973)
Age group (ref =
18-24 years)
25-35 years Week 1 -1 (-315 to 312)
Week 2 -64 (-363 to 234)
Week 3 -149 (-457 to 159)
Week 4 -103 (-416 to 209)
Week 5 175 (-179 to 529)
Week 6 588 (-508 to 1683)
Race (ref = Asian) White Week 1 54 (-575 to 684)
Week 2 -527 (-1157 to 103)
Week 3 -504 (-1124 to 116)
Week 4 -793 (-1495 to -91)
Week 5 -908 (-1800 to -17)
Week 6 -222 (-1505 to 1061)
Black Week 1 3 (-671 to 677)
Week 2 -152 (-845 to 542)
Week 3 -358 (-1047 to 331)
Week 4 -88 (-952 to 776)
Page 20 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nlyWeek 5 -448 (-1449 to 553)
Week 6 -465 (-1849 to 918)
Other Week 1 262 (-499 to 1022)
Week 2 -432 (-1191 to 327)
Week 3 -150 (-971 to 671)
Week 4 -142 (-1030 to 747)
Week 5 -339 (-1382 to 705)
Week 6 85 (-1245 to 1415)
Household income
(ref = /=100k Week 1 -311 (-695 to 73)
Week 2 -193 (-549 to 163)
Week 3 -199 (-606 to 209)
Week 4 -321 (-708 to 66)
Page 21 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nlyWeek 5 -345 (-847 to 156)
Week 6 -494 (-1336 to 348)
Marital status (ref =
single or divorced)
Married/partnered Week 1 5 (-277 to 286)
Week 2 168 (-112 to 449)
Week 3 34 (-248 to 316)
Week 4 16 (-284 to 316)
Week 5 177 (-220 to 574)
Week 6 -399 (-1101 to 304)
Region
(ref=Midwest)
Northeast Week 1 -139 (-623 to 344)
Week 2 -75 (-506 to 357)
Week 3 -39 (-499 to 421)
Week 4 -40 (-475 to 394)
Week 5 -93 (-611 to 424)
Week 6 -222 (-959 to 515)
South Week 1 -199 (-538 to 139)
Week 2 -133 (-496 to 231)
Week 3 -113 (-474 to 248)
Week 4 -152 (-523 to 218)
Page 22 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nlyWeek 5 -5 (-452 to 443)
Week 6 155 (-536 to 847)
West Week 1 -27 (-369 to 316)
Week 2 -101 (-423 to 221)
Week 3 -126 (-464 to 212)
Week 4 -222 (-567 to 124)
Week 5 -10 (-438 to 418)
Week 6 313 (-708 to 1334)
Education (ref= no
college degree)
College degree or higher Week 1 200 (-67 to 466)
Week 2 117 (-149 to 383)
Week 3 -42 (-313 to 229)
Week 4 332 (52 to 612)
Week 5 249 (-104 to 601)
Week 6 -242 (-1115 to 631)
Weight status (ref=
BMI/=25 kg/m2 Week 1 -330 (-606 to -55)
Week 2 -300 (-570 to -30)
Week 3 -249 (-524 to 27)
Week 4 -312 (-600 to -23)
Page 23 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nlyWeek 5 -62 (-397 to 272)
Week 6 -240 (-1241 to 762)
Walkability index
(ref= Car-dependent
or Somewhat
walkable)
Very walkable or Walker's Paradise Week 1 5 (-450 to 460)
Week 2 -204 (-568 to 160)
Week 3 -309 (-726 to 108)
Week 4 -354 (-817 to 110)
Week 5 -158 (-685 to 369)
Week 6 -544 (-1289 to 201)
Urbanity (ref=rural) Suburban Week 1 -238 (-585 to 108)
Week 2 -137 (-491 to 218)
Week 3 138 (-206 to 483)
Week 4 185 (-154 to 524)
Week 5 -86 (-537 to 365)
Week 6 610 (-240 to 1460)
Urban Week 1 -170 (-559 to 218)
Week 2 -62 (-450 to 325)
Week 3 30 (-363 to 423)
Week 4 -16 (-429 to 397)
Page 24 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
-
Confidential: For Review O
nlyWeek 5 -228 (-732 to 277)
Week 6 342 (-353 to 1037)
Pokemon Go player
(ref = non-player)
Pokemon Go player Week 1 968 (705 to 1231)
Week 2 929 (667 to 1191)
Week 3 484 (221 to 746)
Week 4 429 (158 to 700)
Week 5 409 (73 to 745)
Week 6 -36 (-933 to 861)
*Week -4 represents 4 weeks prior to installation of Pokemon Go. Week 1 is the first week since installation of Pokemon Go. Date of Pokemon Go
installation among non-players was set to the median date among players (July 8th, 2016).
Page 25 of 25
https://mc.manuscriptcentral.com/bmj
BMJ
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960