UEF//eRepository
DSpace https://erepo.uef.fi
Rinnakkaistallenteet Terveystieteiden tiedekunta
2017
Genetic predisposition to adiposity is associated with increased objectively
assessed sedentary time in young children
Schnurr TM
Springer Nature
info:eu-repo/semantics/article
info:eu-repo/semantics/acceptedVersion
© Macmillan Publishers Limited All rights reserved
http://dx.doi.org/10.1038/ijo.2017.235
https://erepo.uef.fi/handle/123456789/4478
Downloaded from University of Eastern Finland's eRepository
Genetic predisposition to adiposity is associated with increased
1
objectively assessed sedentary time in young children
2 3
Theresia M. Schnurr1, Anna Viitasalo2, Aino-Maija Eloranta2, Camilla T. Damsgaard3, Yuvaraj Mahendran1, 4
Christian T. Have1, Juuso Väistö2, Mads F. Hjorth3, Line B. Christensen3, Soren Brage4, Mustafa Atalay2, 5
Leo-Pekka Lyytikäinen2,5, Virpi Lindi2, Timo Lakka2,6,7, Kim F. Michaelsen3, Tuomas O. Kilpeläinen1, Torben 6
Hansen1 7
8
1 Novo Nordisk Foundation Center for Basic Metabolic Research, Section of Metabolic Genetics, Faculty of 9
Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark 10
2 Institute of Biomedicine Physiology, School of Medicine, University of Eastern Finland, Kuopio, Finland 11
3 Department of Nutrition, Exercise and Sports, Faculty of Science, University of Copenhagen, Copenhagen, 12
Denmark 13
4 Medical Research Council Epidemiology Unit, University of Cambridge, Cambridge, United Kingdom 14
5 Department of Clinical Chemistry, Fimlab Laboratories and Faculty of Medicine and Life Sciences, 15
University of Tampere, Tampere, Finland 16
6 Department of Clinical Physiology and Nuclear Medicine, School of Medicine, University of Eastern Finland 17
and Kuopio University Hospital, Kuopio, Finland 18
7 Kuopio Research Institute of Exercise Medicine, Kuopio, Finland 19
20
Corresponding author:
21
Theresia M. Schnurr 22
NNF Center for Basic Metabolic Research 23
Faculty of Health and Medical Sciences 24
University of Copenhagen 25
Universitetsparken 1, DIKU 26
DK-2100 Copenhagen Ø 27
Tel: +45 53632729 28
E-mail: tmschnurr@sund.ku.dk 29
30
Total word count (excl. abstract): 2004 words 31
Abstract word count: 205 words 32
Figures: 2 33
Supplementary tables: 2 34
Supplementary Material: 1019 words 35
Running title: Childhood BMI and sedentary behaviour 36
Abbreviations: GRS: Genetic Risk Score, BMI: Body mass index, MVPA: moderate-to-vigorous physical 37
activity, GWAS: Genome-wide association study, 38
Key words Genetic risk score, childhood BMI, adiposity, sedentary time, mendelian randomization 39
40
Abstract
41
Increased sedentariness has been linked to the growing prevalence of obesity in children, but some 42
longitudinal studies suggest that sedentariness may be a consequence rather than a cause of increased 43
adiposity. We used Mendelian randomization to examine the causal relations between body mass index 44
(BMI) and objectively assessed sedentary time and physical activity in 3-8 year-old children from one 45
Finnish and two Danish cohorts [NTOTAL=679]. A genetic risk score (GRS) comprised of 15 independent 46
genetic variants associated with childhood BMI was used as the instrumental variable to test causal effects 47
of BMI on sedentary time, total physical activity, and moderate-to-vigorous physical activity (MVPA). In 48
fixed effects meta-analyses, the GRS was associated with 0.05 SD/allele increase in sedentary time 49
(P=0.019), but there was no significant association with total physical activity (beta=0.011 SD/allele, P=0.58) 50
or MVPA (beta=0.001 SD/allele, P=0.96), adjusting for age, sex, monitor wear-time and first three genome- 51
wide principal components. In two-stage least squares regression analyses, each genetically instrumented 52
one unit increase in BMI z-score increased sedentary time by 0.47 SD (P=0.072). Childhood BMI may have a 53
causal influence on sedentary time but not on total physical activity or MVPA in young children. Our results 54
provide important insights into the regulation of movement behaviour in childhood.
55
56
Introduction
57
Increased sedentary time and decreased physical activity have been linked to the recent increase in the 58
prevalence of overweight and obesity among children (1, 2). However, evidence from longitudinal studies 59
suggests that decreased physical activity and increased sedentary time may be an outcome rather than a 60
cause of increased adiposity in children (3, 4).
61
Genetic variants associated with body mass index (BMI) can be utilized as instrumental variables in 62
Mendelian randomization to test for causal relationships between adiposity and physical activity or 63
sedentary behaviour. In 2014, Richmond et al. performed instrumental variable analyses in 4296 children 64
11 years of age from the UK using a genetic risk score (GRS) for obesity (5), derived from 32 gene variants 65
identified in a published genome-wide association study (GWAS) of adult BMI (6). Genetic predisposition to 66
higher BMI was robustly associated with longer sedentary time and lower levels of physical activity (5), 67
suggesting causality. However, these findings remain to be replicated in younger children in whom genetic 68
determinants of movement behaviour may be particularly discernible due to higher tendency for voluntary 69
and spontaneous, play-oriented activity (7, 8). Further, a recent GWAS in children identified 15 loci for 70
childhood BMI (9), making it possible to generate a more specific instrumental variable for childhood 71
adiposity than the GRS for adult BMI used by Richmond et al. (5).
72
The aim of the current study was to investigate whether a GRS of 15 loci for childhood BMI is associated 73
with objectively assessed sedentary time and physical activity in young children.
74
75
Methods
76
Participants
77
The participants of the study include 287 Danish children 3 years of age from the Småbørns Kost Og Trivsel I 78
and II (SKOT I and II) studies (10) and 400 Finnish children from the Physical Activity and Nutrition in 79
Children (PANIC) study (11). Details on the recruitment, inclusion criteria and ethical approvals in SKOT I, 80
SKOT II, and PANIC are provided in Supplementary Material 1.
81
Measurement of body size and composition
82
In the SKOT I and II studies, body weight was measured by the Tanita WB-100MA digital scale (Tanita 83
Corporation, Tokyo, Japan) and body height by the 235 Heightronic digital stadiometer (QuickMedical, 84
Issaquah, WA, USA). The age and gender-specific BMI z-score was calculated using the WHO Anthro 85
software, version 3.2.2 (12). In the PANIC study, body weight was measured using the InBody® 720 86
bioimpedance device (Biospace, Seoul, Korea) and body height using a wall-mounted stadiometer. Age and 87
gender-specific BMI z-score was calculated based on Finnish reference data (13).
88
Assessment of sedentary time, total physical activity and MVPA
89
In the SKOT I and II studies the ActiGraph GT3X accelerometer (ActiGraph LLC, Pensacola, FL, USA), and in 90
the PANIC study Actiheart (Actiheart, CamNTech Ltd., Cambridge, UK) was used to assess sedentary time 91
and physical activity. Details on the assessment of activity behaviours are provided in Supplementary 92
Material 1.
93
Genotyping, SNP selection, and genetic risk score construction
94
Children in SKOT I and II were genotyped using the Illumina Infinium HumanCoreExome Beadchip. Children 95
in the PANIC study were genotyped using the Illumina Custom Infinium Cardio-Metabochip and the Illumina 96
Infinium HumanCoreExome Beadchip (Illumina, San Diego, CA, USA) and the genotypes from the two arrays 97
were combined (see Supplementary Material 1 for information on quality control). The SNPs included in 98
the GRS were selected based on a previously published GWAS meta-analysis in children 2-10 years of age 99
(9) that identified 15 independent loci associated with BMI at genome-wide significance (p<5×10-8). We 100
constructed a weighted BMI-increasing GRS by summing the number of BMI-increasing alleles weighted by 101
the effect sizes of the variants estimated in the GWAS discovery study (Supplementary Material 1, 102
Supplementary Table 1).
103
Statistical analysis
104
All association analyses were performed using R, version 3.3.1. Only children with valid physical activity and 105
genotype data (nSKOT I=208; nSKOT II=71; nPANIC=400) were included in the present analyses. Sedentary time, 106
total physical activity, and moderate-to-vigorous intensity physical activity (MVPA) variables were rank 107
inverse normally transformed to approximate normal distribution with a mean of 0 and standard deviation 108
(SD) of 1, and the effect sizes are thus reported in SD units of the inverse normally transformed trait.
109
The associations of the BMI z-score as well as the BMI-increasing GRS with sedentary time, physical activity 110
and MVPA were analysed by linear regression adjusting for age, sex, and monitor wear-time. The 111
association of the BMI-increasing GRS with the BMI z-score was analysed by linear regression adjusting only 112
for monitor wear-time, because the BMI z-score is age and sex-specific. The BMI-increasing GRS did not 113
show an association with additional potential confounders in PANIC, the largest cohort included in the 114
meta-analysis (sleep, socioeconomic status; p > 0.05, data not shown). The causal relationships between 115
BMI and activity behaviours were tested using two-stage least squares regression analyses implemented in 116
the ‘AER’ package in R (version 3.3.3). We used the Durbin-Wu-Hausman (DWH) test for endogeneity and 117
calculated the F-statistic for the PANIC cohort (F-statisticPANIC) to compare effect estimates between the 118
instrumental and observational analyses (14). To test for potential directional pleiotropy in the genetic 119
instrument, we used Egger regression, implemented in the ‘MendelianRandomization’ package in R 120
(version 3.3.3), where the deviation of the intercept from zero provides evidence of pleiotropy (15). The 121
associations of the BMI-increasing GRS, two-stage least squares regression and Egger regression analyses 122
were additionally adjusted for the first three genome-wide principal components of the respective study.
123
We pooled the results from the SKOT I, SKOT II and PANIC studies by fixed effects meta-analyses using the 124
‘meta’ package in R (version 4.6.0).
125
126
Results
127The characteristics of children from the SKOT I, SKOT II and PANIC studies are summarized in 128
Supplementary Table 2. The average age of the children was 3.0 years (range 2.9-3.3 years) in SKOT I; 3.0 129
years (range 2.9-3.2 years) in SKOT II; and 7.6 years (range 6.6-9.0 years) in PANIC. The GRS was normally 130
distributed in all three cohorts, with a mean (range) of 8.6 (3.8-14.7), 9.0 (5.0-17.8) and 9.3 (3.7-16.1) BMI- 131
increasing alleles in SKOT I, SKOT II and PANIC, respectively.
132
A higher BMI z-score was associated with increased sedentary time (β=0.22 SD, P=7.6x10-9) and reduced 133
MVPA (β=-0.17 SD, P=1.1x10-5), but not with total physical activity (β=0.003 SD, P=0.94) (Figure 1).
134
Heterogeneity was observed in the association of BMI z-score with sedentary time and MVPA (phet<0.05).
135
136
Figure 1.
137
[insert Figure 1]
138
Forest plots showing the associations of BMI z-score (left column), childhood BMI-increasing GRS (middle 139
column) and genetically predicted BMI z-score (right column) with A. sedentary time, B. total physical 140
activity, and C. moderate-to-vigorous physical activity (MVPA). For the GRS associations, the results are 141
aligned according to the BMI-increasing allele of the GRS. All analyses are adjusted for age, gender, monitor 142
wear-time and first three principal components. The effects were pooled using fixed effects models. The 143
estimated per-BMI z-score, per-allele and per-genetically predicted BMI z-score effect sizes are reported in 144
SD units based on inverse normally transformed outcome trait. Heterogeneity statistics include the I2 value 145
that describes the percentage of variation across the meta-analysis that is due to heterogeneity, and phet, 146
the p-value for the χ2 test of heterogeneity.
147 148
A higher BMI-increasing GRS was associated with a higher BMI z-score (β=0.056 SD/allele, P=0.003) and 149
longer sedentary time (β=0.040 SD/allele, P=0.019), suggesting a causal effect of BMI z-score on sedentary 150
behavior (Figure 2). In two-stage least squares analyses, each genetically instrumented one unit increase in 151
BMI z-score increased sedentary time by 0.47 SD (P=0.072, F-statisticPANIC=8.2), and no difference was found 152
between the observational and genetically instrumented estimates in the DWH test (P>0.05). We found no 153
evidence of directional pleiotropy in the genetic instrument using the Egger intercept test (PINTERCEPT=0.28), 154
and the causal estimate from Egger regression was directionally consistent with that derived from the two- 155
stage least squares method.
156
There was no significant association between the BMI-increasing GRS and MVPA (β=0.001, P=0.96) or total 157
physical activity (β=0.011, P=0.58), and two-stage least squares analyses were not suggestive of a causal 158
effect of BMI on MVPA (β=-0.026, P=0. 94, F-statisticPANIC=7.5) or physical activity (β=0.22, P=0. 55, F- 159
statisticPANIC=7.5) (Figure 1).
160
161
Figure 2.
162
[insert Figure 2]
163
Mendelian randomization analysis to test the causal effect of childhood BMI on sedentary time. Beta values 164
are expressed in units of standard deviation (SD) of the inverse-normally transformed traits. GRS = Genetic 165
risk score, BMI z-score = age- and sex-specific BMI standard deviation score, NTOTAL= number of individuals 166
included in meta-analysis.
167 168
Discussion
169
In the present study, a GRS for childhood BMI was nominally significantly associated with BMI and 170
sedentary time, but not with total physical activity or MVPA. Our results may suggest that higher adiposity 171
is causally associated with longer sedentary time but not with decreased physical activity in young children.
172
Consistent with our findings, Richmond et al. (5) found that a higher GRS for BMI was positively associated 173
with longer daily sedentary time in 11-year old children from the UK. However, they also reported that a 174
higher GRS was associated with lower levels of total physical activity and MVPA, whereas we found no 175
association between the GRS and total physical activity or MVPA. While the sample sizes for the present 176
analyses were smaller than in the study by Richmond et al., we observed an effect close to zero for the 177
association of the GRS with physical activity and MVPA, and with confidence intervals suggesting that little 178
or no effect is present in 3-8 year old children. Nevertheless, our findings should ideally be validated in 179
further studies including large samples of young children with objectively measured activity behaviour.
180
The age of the children and country-specific differences in the education system may partly explain the 181
observed differences in the results of the study by Richmond et al (5) and our study. In our study, we also 182
found heterogeneity in the association of the BMI z-score with sedentary time and MVPA, and visual 183
observation of the forest plots indicated that the two SKOT cohorts show consistent results which differ 184
from those seen for the PANIC cohort, which may be due to the different age range of children included in 185
these cohorts. Most 3-year-old Danish children attend kindergarten where physical activity typically 186
consists of play-oriented activities (16) and the children are free to choose whether to play passively or 187
actively. The Finnish children 6-8 years of age were first graders in primary schools when they were invited 188
to participate in the PANIC study. They were thus more likely to engage in play-oriented physical activity 189
because of their recent pre-school times than the 11-year-old children from the UK, although they also 190
spent longer periods of time in sedentary and non-sedentary activities during school hours. The tendency 191
to engage in voluntary and play-oriented activities in younger children could explain the lack of association 192
between the GRS for childhood BMI and physical activity in the present study.
193
While our results are suggestive of an effect of adiposity on sedentary behaviour, we could not investigate 194
whether a genetic predisposition to sedentary behaviour reciprocally results in higher BMI, because no 195
genetic variants associated with sedentary behaviour have yet been robustly identified (17). Similarly, we 196
could not examine whether MVPA has a causal effect on BMI in young children, and whether such an effect 197
explains the observed association between higher BMI and lower MVPA. Furthermore, we cannot fully 198
exclude the possibility of residual pleiotropy, i.e. that the selected genetic variants act not only on BMI but 199
also on other phenotypes related to sedentary time.
200
In conclusion, we showed that young children with higher genetic risk for obesity have increased 201
objectively measured sedentary time but not decreased physical activity, suggesting that obesity may be 202
causally associated with longer time spent in sedentary pursuits at this age. Reducing BMI may thus be an 203
effective strategy to reduce sedentariness in overweight children. While the mechanisms underlying the 204
potential causal relationship between BMI and sedentary time remain unclear, they are likely to involve 205
both physiological factors and factors related to the family environment (18). Our findings provide novel 206
insights into the regulation of movement behaviour in childhood and suggest that more attention should be 207
given to the sedentary-time increasing effect of obesity in young children.
208
209
Supplementary information is available at the International Journal of Obesity’s website.
210 211
Conflict of interest
212
The authors declare no conflict of interest.
213 214
Acknowledgements
215
We specially want to express our thanks to the participant children and their parents that were part of the 216
SKOT I, SKOT II and PANIC studies. This project was carried out as part of the research programme 217
"Governing Obesity" funded by the University of Copenhagen Excellence Programme for Interdisciplinary 218
Research (www.go.ku.dk) and was supported by the Danish Diabetes Academy supported by the Novo 219
Nordisk Foundation. The SKOT studies were supported by grants from The Danish Directorate for Food, 220
Fisheries and Agri Business as part of the ‘Complementary and young child feeding (CYCF) – impact on 221
short- and long-term development and health’ project. The PANIC study was funded by grants from 222
Ministry of Social Affairs and Health of Finland, Ministry of Education and Culture of Finland, Finnish 223
Innovation Fund Sitra, Social Insurance Institution of Finland, Finnish Cultural Foundation, Juho Vainio 224
Foundation, Foundation for Paediatric Research, Doctoral Programs in Public Health, Paavo Nurmi 225
Foundation, Paulo Foundation, Diabetes Research Foundation, Yrjö Jahnsson Foundation, Finnish 226
Foundation for Cardiovascular Research, Research Committee of the Kuopio University Hospital Catchment 227
Area (State Research Funding), Kuopio University Hospital (previous state research funding (EVO), funding 228
number 5031343), and the city of Kuopio. The Novo Nordisk Foundation Center for Basic Metabolic 229
Research is an independent research center at the University of Copenhagen partially funded by an 230
unrestricted donation from the Novo Nordisk Foundation (http://metabol.ku.dk). The work of Soren Brage 231
was funded by the UK Medical Research Council [MC_UU_12015/3]. Tuomas O. Kilpeläinen was supported 232
by the Danish Council for Independent Research (DFF – 1333-00124 and Sapere Aude program grant DFF – 233
1331-00730B).
234
235
Data availability
236
Relevant data for the present study are within the paper and its Supporting Information files. If you wish to 237
see additional data, the authors confirm that, for approved reasons, some access restrictions apply to the 238
data underlying the findings. Data is available from the Novo Nordisk Foundation Center for Basic 239
Metabolic Research, section of Metabolic Genetics whose authors may be contacted 240
at torben.hansen@sund.ku.dk.
241
242
References 243
244
1. Prentice-Dunn H, Prentice-Dunn S. Physical activity, sedentary behavior, and childhood 245
obesity: a review of cross-sectional studies. Psychology, health & medicine. 2012;17(3):255-73.
246
2. Must A, Tybor D. Physical activity and sedentary behavior: a review of longitudinal studies of 247
weight and adiposity in youth. International journal of obesity. 2005;29:S84-S96.
248
3. Hjorth MF, Chaput J-P, Ritz C, Dalskov S-M, Andersen R, Astrup A, et al. Fatness predicts 249
decreased physical activity and increased sedentary time, but not vice versa: support from a longitudinal 250
study in 8-to 11-year-old children. International journal of obesity. 2014;38(7):959-65.
251
4. Metcalf BS, Hosking J, Jeffery A, Voss L, Henley W, Wilkin T. Fatness leads to inactivity, but 252
inactivity does not lead to fatness: a longitudinal study in children (EarlyBird 45). Archives of disease in 253
childhood. 2010:archdischild175927.
254
5. Richmond RC, Smith GD, Ness AR, den Hoed M, McMahon G, Timpson NJ. Assessing Causality 255
in the Association between Child Adiposity and Physical Activity Levels: A Mendelian Randomization 256
Analysis. PLoS medicine. 2014;11(3):e1001618.
257
6. Speliotes EK, Willer CJ, Berndt SI, Monda KL, Thorleifsson G, Jackson AU, et al. Association 258
analyses of 249,796 individuals reveal 18 new loci associated with body mass index. Nature genetics.
259
2010;42(11):937-48.
260
7. Franks PW, Ravussin E, Hanson RL, Harper IT, Allison DB, Knowler WC, et al. Habitual physical 261
activity in children: the role of genes and the environment. The American journal of clinical nutrition.
262
2005;82(4):901-8.
263
8. Väistö J, Eloranta A-M, Viitasalo A, Tompuri T, Lintu N, Karjalainen P, et al. Physical activity 264
and sedentary behaviour in relation to cardiometabolic risk in children: cross-sectional findings from the 265
Physical Activity and Nutrition in Children (PANIC) Study. International Journal of Behavioral Nutrition and 266
Physical Activity. 2014;11(1):55.
267
9. Felix JF, Bradfield JP, Monnereau C, van der Valk RJ, Stergiakouli E, Chesi A, et al. Genome- 268
wide association analysis identifies three new susceptibility loci for childhood body mass index. Human 269
molecular genetics. 2015:ddv472.
270
10. Andersen LBB, Pipper CB, Trolle E, Bro R, Larnkjær A, Carlsen E, et al. Maternal obesity and 271
offspring dietary patterns at 9 months of age. European journal of clinical nutrition. 2014.
272
11. Eloranta A, Lindi V, Schwab U, Kiiskinen S, Venäläinen T, Lakka H, et al. Dietary factors 273
associated with metabolic risk score in Finnish children aged 6–8 years: the PANIC study. European journal 274
of nutrition. 2014;53(6):1431-9.
275
12. Organization WH. WHO Anthro for personal computers, version 3.2. 2, 2011: software for 276
assessing growth and development of the world’s children. Geneva: WHO. 2010.
277
13. Saari A, Sankilampi U, Hannila M-L, Kiviniemi V, Kesseli K, Dunkel L. New Finnish growth 278
references for children and adolescents aged 0 to 20 years: length/height-for-age, weight-for- 279
length/height, and body mass index-for-age. Annals of medicine. 2011;43(3):235-48.
280
14. Baiocchi M, Cheng J, Small DS. Instrumental variable methods for causal inference. Statistics 281
in medicine. 2014;33(13):2297-340.
282
15. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments:
283
effect estimation and bias detection through Egger regression. International journal of epidemiology.
284
2015;44(2):512-25.
285
16. Timmons BW, Naylor P-J, Pfeiffer KA. Physical activity for preschool children-how much and 286
how? This article is part of a supplement entitled Advancing physical activity measurement and guidelines 287
in Canada: a scientific review and evidence-based foundation for the future of Canadian physical activity 288
guidelines co-published by Applied Physiology, Nutrition, and Metabolism and the Canadian Journal of 289
Public Health. It may be cited as Appl. Physiol. Nutr. Metab. 32 (Suppl. 2E) or as Can. J. Public Health 98 290
(Suppl. 2). Applied Physiology, Nutrition, and Metabolism. 2007;32(S2E):S122-S34.
291
17. De Geus EJ, Bartels M, Kaprio J, Lightfoot JT, Thomis M. Genetics of regular exercise and 292
sedentary behaviors. Twin Research and Human Genetics. 2014;17(04):262-71.
293
18. Hinkley T, Salmon J, Okely AD, Trost SG. Correlates of sedentary behaviours in preschool 294
children: a review. International Journal of Behavioral Nutrition and Physical Activity. 2010;7(1):66.
295 296 297
BMI z-score association P=0.938 I2=0%, phet=0.92
A. Sedentary time
B. Total physical activity
C. MVPA time
BMI z-score association P=1.12×10-5 I2=82%, phet=0.0045 BMI z-score association P=7.63×10-9 I2=86%, phet=0.001
Genetically predicted BMI z-score association P=0.938 I2=0%, phet=0.61 Genetically predicted BMI z-score association P=0.072 I2=0%, phet=0.58
Genetically predicted BMI z-score association P=0.549 I2=0%, phet=0.30 GRS association
P=0.019 I2=0%, phet=0.83
GRS association P=0.577 I2=63%, phet=0.06
GRS association P=0.957 I2=0%, phet=0.43
Figure 1. Forest plots showing the associations of BMI z-score (left column), childhood BMI-increasing GRS (middle column) and genetically predicted BMI z-score (right column)
with A. sedentary time, B. total physical activity, and C. moderate-to-vigorous physical activity (MVPA). For the GRS associations, the
results are aligned according to the BMI-increasing allele of the GRS. All analyses are adjusted for age, gender, monitor wear-time and first three principal
components. The effects were pooled using fixed effects models. The estimated per-BMI z-score, per-allele and per-genetically predicted BMI z-score effect
sizes are reported in SD units based on inverse normally transformed outcome trait. Heterogeneity statistics include the I
2value that describes the percentage
of variation across the meta-analysis that is due to heterogeneity, and p
het, the p-value for the χ
2test of heterogeneity.
NTOTAL= 679
β = 0.045 SD/allele P = 0.019
BMI z-score
Sedentary time Childhood specific BMI GRS
NTOTAL= 679
β = 0.056 SD/allele P = 0.003
NTOTAL= 679
β = 0.22 SD/BMI z-score unit P = 7.6x10-9
Instrumental analysis NTOTAL = 679 β = 0.47 SD/genetically predicted BMI z-score unit
P = 0.072
Figure 2. Mendelian randomization analysis to test the causal effect of childhood BMI on sedentary time. Beta
values are expressed in units of standard deviation (SD) of the inverse-normally transformed traits. GRS = Genetic
risk score, BMI z-score = age- and sex-specific BMI standard deviation score, N
TOTAL= number of individuals included
in meta-analysis.
Supplementary Material
1
Methods
2
Ethics statement
3
Prior to participation, written informed consent was obtained from all parents of the children included in 4
SKOT I and SKOT II; and from all children and parents participating in the PANIC study. The Committees on 5
Biomedical Research Ethics for the Capital Region of Denmark approved the study protocol of SKOT I (H-KF- 6
2007-0003) and SKOT II (H-3-2010-122). The Research Ethics Committee of the Hospital District of Northern 7
Savo, Finland approved the study protocol of the PANIC study. The PANIC study is registered under 8
ClinicalTrials.gov with registration number NCT01803776. All studies were conducted in accordance with 9
the principles of the Declaration of Helsinki.
10
Study population
11
SKOT I and SKOT II
12
As opposed to SKOT I, SKOT II children were all born from overweight mothers (with a pre-pregnancy BMI 13
above 30kg/m2). Recruitment and inclusion criteria have been described in detail previously (1, 2). In short, 14
the 329 children included in SKOT I were healthy singletons randomly recruited from the National Civil 15
Registry and living in Copenhagen or Frederiksberg municipality, Denmark, in 2006-2007 (3). The included 16
children were born at term and had Danish-speaking parents. The 184 children included in SKOT II met all 17
above criteria with the exception that they were recruited in 2010-2012 and were offspring of women who 18
had participated in the Treatment of Obese Pregnant Women intervention study at Hvidore Hospital, 19
Hvidovre (Denmark) (4).
20
Physical activity
21
SKOT I and SKOT II
22
The children were asked to wear the ActiGraph GT3X in an elastic belt tightly at the right hip for seven days 23
and seven nights, besides when performing water-based activities (i.e. taking a bath or swimming). Only the 24
data from children who wore the ActiGraph for at least eight hours per day for four days were included in 25
the analyses. The processing of data was conducted using Actilife software, Version 6.7.3 (ActiGraph LLC, 26
Pensacola, FL, USA). Activity was recorded with a sample rate (epoch length) of 2 seconds and was 27
reintegrated into 60-second epochs. Non-wear time during the day was defined as periods of 20 minutes or 28
more of consecutive zeroes and was excluded prior to data analysis. Usual night time sleep from parent 29
report questionnaire was used to exclude night time as non-wear time prior to data analysis. Eight children 30
in SKOT I and five children in SKOT II did not have parent report questionnaire information available. For 31
these children, usual night time sleep was defined individually for each child as the average sleep time 32
based on manual inspection of the activity graphs produced by the sleep analysis module integrated into 33
the Actilife software. We applied cut-offs based on Vector Magnitude settings: <819 counts per minute 34
(cpm) to define sedentary time and ≥3908 cpm to define MVPA, based on a validation study in preschool- 35
aged children (5). For the present study, total physical activity, expressed in counts per minute (cpm) 36
averaged over the period of valid wear time recording, and time spent in sedentary and MVPA intensities, 37
expressed as minutes per day (min/day).
38
PANIC
39
The children were instructed to wear the Actiheart device continuously for a minimum of four consecutive 40
days and nights. The monitor was attached to the chest with two standard electrocardiogram electrodes 41
(Bio Protech Inc, Seoul, South Korea) and data were recorded in 60-second epochs. The cleaning and 42
calibration of these data in the PANIC study has been described in detail previously (6). For the present 43
analyses, physical activity and time spent in sedentary and MVPA intensity records were included if they 44
contained at least 48 hours (32 hours during week-days, 16 hours during weekend days) of wear data in 45
total and at least 12 hours of morning, noon, afternoon, and evening wear data (7). Sedentary time was 46
defined as time spent at intensity of at least 1.5 metabolic equivalents, excluding sleep time. MVPA was 47
defined as time spent at intensity of at least 3 metabolic equivalents. We used the acceleration data from 48
Actiheart to define total physical activity as movement intensity.
49
Genotyping – quality control
50
SKOT I and SKOT II
51
Genotypes were called using the Genotyping module, Version 1.9.4 of GenomeStudio software, Version 52
2011.1 (Illumina). We excluded closely related individuals and samples with extreme inbreeding 53
coefficients, mislabelled gender or call rate < 95%, duplicates and individuals identified as ethnic outliers, 54
leaving 275 individuals of SKOT I and 116 individuals of SKOT II individuals who passed all quality control 55
criteria. We applied a >95% genotype call rate filter for the inclusion of SNPs. Additional genotypes were 56
imputed into 1000 Genomes Phase 1 (8) using Impute 2 (9). The imputation quality was high (proper_info >
57
0.95) for all imputed variants included in the current study. All variants obeyed Hardy Weinberg equilibrium 58
(p > 0.05).
59
PANIC
60
Genotypes were called using Illumina BeadStudio, Version 3.3.7 (Cardio-Metabochip) and GenomeStudio 61
(HumanCoreExome Beadchip) softwares using GenCall and zCall algorithms. The final quality control was 62
done using the PLINK software, Version 1.07. Samples successfully genotyped with both Cardio-Metabochip 63
and HumanCoreExome Beadchip were merged prior to quality control. We excluded closely related 64
individuals, ethnic outliers, samples with mislabelled gender and call rate < 95%. A 95% genotype call rate 65
criterion for inclusion of SNPs was applied and SNPs with Hardy Weinberg equilibrium p<1x10-6 or MAF <1%
66
were excluded. Additional genotypes were imputed into 1000 Genomes reference panel (Phase 1 67
integrated variant set release v3) using SHAPEIT v2 for haplotyping and Impute 2 for imputing genotype 68
dosages.
69
Genetic Risk score construction
70
In SKOT I and SKOT II, eight of the 15 SNPs for childhood BMI identified by a GWAS were directly genotyped 71
(rs7550711, rs543874, rs13130484, rs987237, rs7132908, rs12429545, rs1421085, rs11676272). The 72
remaining genotypes (rs3829849, rs4854349, rs6567160, rs8092503, rs12041852, rs13253111, rs13387838) 73
were retrieved from imputed data and the estimated risk-allele dosage was used in place of the unavailable 74
risk-allele count when calculating the GRS (Supplementary Table 1).
75
In PANIC, 14 of the 15 BMI variants were directly genotyped. The remaining rs13253111 SNP was retrieved 76
from the imputed data. For six children, rs13253111 imputations were not available and could be assumed 77
to be missing at random. We imputed these to the mean allelic dosage of rs13253111 in the PANIC cohort 78
(Supplementary Table 1).
79
References 80
81
1. Andersen LB, Mølgaard C, Michaelsen KF, Carlsen EM, Bro R, Pipper CB. Indicators of dietary 82
patterns in Danish infants at 9 months of age. Food & nutrition research. 2015;59.
83
2. Andersen LBB, Pipper CB, Trolle E, Bro R, Larnkjær A, Carlsen E, et al. Maternal obesity and 84
offspring dietary patterns at 9 months of age. European journal of clinical nutrition. 2014.
85
3. Madsen AL, Larnkjær A, Mølgaard C, Michaelsen KF. IGF-I and IGFBP-3 in healthy 9month old 86
infants from the SKOT cohort: Breastfeeding, diet, and later obesity. Growth Hormone & IGF Research.
87
2011;21(4):199-204.
88
4. Renault KM, Nørgaard K, Nilas L, Carlsen EM, Cortes D, Pryds O, et al. The Treatment of 89
Obese Pregnant Women (TOP) study: a randomized controlled trial of the effect of physical activity 90
intervention assessed by pedometer with or without dietary intervention in obese pregnant women.
91
American journal of obstetrics and gynecology. 2014;210(2):134. e1-. e9.
92
5. Butte NF, Wong WW, Lee JS, Adolph AL, Puyau MR, Zakeri IF. Prediction of energy 93
expenditure and physical activity in preschoolers. Medicine and science in sports and exercise.
94
2014;46(6):1216-26.
95
6. Haapala EA, Väistö J, Lintu N, Tompuri T, Brage S, Westgate K, et al. Adiposity, physical 96
activity and neuromuscular performance in children. Journal of sports sciences. 2016:1-8.
97
7. Brage S, Westgate K, Wijndaele K, Godinho J, Griffin S, Wareham N. Evaluation of a method 98
for minimising diurnal information bias in objective sensor data. Int Conf Amb Mon Phys Act Mov 99
(Conference Proceeding). 2013.
100
8. Consortium GP. An integrated map of genetic variation from 1,092 human genomes. Nature.
101
2012;491(7422):56-65.
102
9. Howie BN, Donnelly P, Marchini J. A flexible and accurate genotype imputation method for 103
the next generation of genome-wide association studies. PLoS Genet. 2009;5(6):e1000529.
104
10. Ridgway CL, Brage S, Sharp SJ, Corder K, Westgate KL, van Sluijs EM, et al. Does birth weight 105
influence physical activity in youth? A combined analysis of four studies using objectively measured physical 106
activity. PloS one. 2011;6(1):e16125.
107 108
Supplementary Tables
109
Supplementary Table 1.
110
Overview about the 15 BMI increasing genetic variants that were included in the GRS.
111
SNP reported in
meta-analysis* Chromosome Position
Nearest Gene
EA/Non-
EA* EAF*
Effect size on BMI*
Directly genotyped = CHIP/
imputed = IMP SKOT I, SKOT II PANIC
rs13387838 2 206989692 ADAM23 A/G 0.04 0.139 IMP IMP
rs7550711 1 109884409 GPR61 T/C 0.04 0.105 CHIP CHIP
rs4854349 2 637861 TMEM18 C/T 0.83 0.090 IMP CHIP
rs543874 1 176156103 SEC16B G/A 0.20 0.077 CHIP CHIP
rs12429545 13 53000207 OLFM4 A/G 0.13 0.076 CHIP CHIP
rs11676272 2 24995042 ADCY3 G/A 0.46 0.068 CHIP CHIP
rs13130484 4 44870448 GNPDA2 T/C 0.44 0.067 CHIP CHIP
rs7132908 12 48549415 FAIM2 A/G 0.39 0.066 CHIP CHIP
rs987237 6 50911009 TFAP2B G/A 0.19 0.062 CHIP CHIP
rs1421085 16 52358455 FTO C/T 0.41 0.059 CHIP CHIP
rs6567160 18 55980115 MC4R C/T 0.23 0.050 IMP CHIP
rs12041852 1 74776088 TNNI3K G/A 0.46 0.046 IMP CHIP
rs8092503 18 50630485 RAB27B G/A 0.27 0.045 IMP CHIP
rs13253111 8 28117893 ELP3 A/G 0.57 0.042 IMP CHIP
rs3829849 9 128430621 LMX1B T/C 0.36 0.041 IMP CHIP
112
An overview of the SNPs presented in the report by Felix et al. and the SNPs investigated in the SKOT I, SKOT II and PANIC cohorts, sorted by effect 113
size on BMI. *EA (BMI increasing allele)/Non-EA, EAFs (effect allele frequencies) and effect sizes are from the SNPs reported by Felix et al. (joint 114
analyses). EA=Effect allele, EAF=Effected allele frequency.
115 116
Supplementary Table 2.
117
Cross-sectional study characteristics of children in the SKOT I, SKOT II and PANIC cohorts.
118
Trait SKOT I SKOT II PANIC
all boys girls all boys girls all boys girls
n 208 104 104 79 47 32 400 202 198
Age (years) 3.0 (0.1) 3.0 (0.1) 3.0 (0.1) 3.0 (0.1) 3.0 (0.1) 3.0 (0.1) 7.6 (0.4) 7.7 (0.4) 7.6 (0.4) Height (m) 95.8 (3.4) 96.7 (3.4) 94.9 (3.2) 97.2 (3.7) 97.3 (3.3) 97.0 (4.2) 128.7
(5.5)
130.0 (5.5)
127.5 (5.4) Weight (kg) 14.6 (1.5) 14.9 (1.5) 14.3 (1.5) 15.5 (1.9) 15.6 (1.8) 15.4 (2.2) 26.6 (4.5) 27.1 (4.4) 26.1 (4.6) BMI (kg/m2) 15.9 (1.2) 15.9 (1.1) 15.8 (1.2) 16.4 (1.4) 16.4 (1.4) 16.3 (1.4) 16.0 (1.9) 16.0 (1.8) 16.0 (2.0) BMI z-score 0.3 (0.9) 0.3 (0.8) 0.3 (0.9) 0.8 (1.0) 0.8 (1.0) 0.8 (1.0) -0.0 (1.0) -0.2 (1.0) -0.2 (1.0) Total physical
activity (cpm)*
1321 (230)
1381 (213)
1261 (233)
1252 (273)
1320 (275)
1134
(231) 380 (103) 397 (113) 362 (88) Sedentary time
(min/day) 300 (53) 301 (47) 315 (58) 324 (74) 312 (71) 347 (76) 283 (118) 274 (110) 293 (125) MVPA time
(min/day) 36 (17) 41 (18) 30 (15) 33 (18) 38 (19) 24 (12) 115 (64) 135 (66) 94 (54) GRS (Number
of BMI increasing risk alleles)
8.6 (2.1) 8.7 (2.0) 8.5 (2.1) 9.0 (2.2) 9.0 (2.3) 9.0 (2.0) 9.3 (2.0) 9.2 (2.0) 9.4 (1.9) 119
GRS = genetic risk score, MVPA = moderate-to-vigorous physical activity, cm=centimetre, kg=kilogram, 120
BMI=body mass index, BMI z-score = age- and gender-specific BMI standard deviation score, cpm=counts 121
per minute, min/day=minutes per day. * for PANIC, we used the uniaxial acceleration data from Actiheart 122
and applied a previously derived conversion factor of 5 (Actigraph counts=Actiheart counts x 5) to express 123
total physical activity in cpm (10).
124 125
