Association of the FTO (rs9939609) and MC4R (rs17782313) gene polymorphisms with maternal body weight during pregnancy

Accepted Manuscript Association of the FTO (rs9939609) and MC4R (rs17782313) gene polymorphisms with maternal body weight during pregnancy Maisa Cruz Martins, M.Sc., Janet Trujillo, Ph.D., Dayana Rodrigues Farias, M.Sc., Claudio Jose Struchiner, Ph.D., Gilberto Kac, Ph.D. PII:

S0899-9007(16)30070-3

DOI:

10.1016/j.nut.2016.04.009

Reference:

NUT 9768

To appear in:

Nutrition

Received Date: 26 January 2016 Revised Date:

15 April 2016

Accepted Date: 27 April 2016

Please cite this article as: Martins MC, Trujillo J, Farias DR, Struchiner CJ, Kac G, Association of the FTO (rs9939609) and MC4R (rs17782313) gene polymorphisms with maternal body weight during pregnancy, Nutrition (2016), doi: 10.1016/j.nut.2016.04.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

ACCEPTED MANUSCRIPT Association of the FTO (rs9939609) and MC4R (rs17782313) gene polymorphisms with maternal body weight during pregnancy

RI PT

Running title: FTO and MC4R polymorphisms and pregnancy

Maisa Cruz Martins, M.Sc.1,3

SC

Janet Trujillo, Ph.D.1

Claudio Jose Struchiner, Ph.D.2 *

Gilberto Kac, Ph.D.1

1

M AN U

Dayana Rodrigues Farias, M.Sc.1

Nutritional Epidemiology Observatory, Department of Social and Applied Nutrition,

Brazil.

TE D

Institute of Nutrition Josué de Castro, Rio de Janeiro Federal University, Rio de Janeiro,

National School of Public Health, Oswaldo Cruz Foundation, Rio de Janeiro, Brazil.

3

Graduate Program in Nutrition, Institute of Nutrition Josué de Castro, Rio de Janeiro

EP

2

AC C

Federal University, Rio de Janeiro, Brazil.

*Author to correspondence: Gilberto Kac, Av. Carlos Chagas Filho, 373, CCS, Bloco J2, sala 29, Cidade Universitária – Ilha do Fundão, 21941-902, Rio de Janeiro, Brazil. Phone: +55 21 39386595, e-mail: [email protected]

2

ACCEPTED MANUSCRIPT

Authorship M.C.M. designed the research, performed the genotyping, analyzed the data, wrote the paper, and had primary responsibility for the final content; J.T. performed the

RI PT

genotyping, critically revised the manuscript and conducted the laboratory work; D.R.F. performed the genotyping, analyzed the data and critically revised the manuscript; C.J.S. critically revised the manuscript and G.K. designed the research, critically revised

AC C

EP

TE D

M AN U

approved the final version of the manuscript.

SC

the manuscript and had primary responsibility for the final content. All authors read and

3

ACCEPTED MANUSCRIPT Abstract

2

Objective: Fat mass and obesity (FTO) and melanocortin 4 receptor (MC4R) genes have

3

been consistently associated with the risk of obesity, but few studies have examined the

4

association of the obesity-risk alleles with gestational outcomes. Our aim was to

5

evaluate the association between single-nucleotide polymorphisms (SNPs) of the FTO

6

(rs9939609) and MC4R (rs17782313) genes with changes in maternal body weight

7

during pregnancy.

8

Methods: A sample of 136 pregnant women were followed in a prospective cohort at 5-

9

13, 20-26 and 30-36 gestational weeks and 30-45 days postpartum. SNPs were analyzed

10

by real-time PCR. Associations between polymorphisms and the outcomes were

11

investigated through longitudinal linear mixed-effects models, multiple linear regression

12

models and Poisson regression models.

13

Results: A SNP in the FTO (rs9939609) gene but not in the MC4R (rs17782313) gene

14

was significantly associated with pre-pregnancy BMI ≥ 25 kg/m2 (RRFTO=2.1; 95%CI

15

1.4 to 3.1). Neither SNP was statistically associated with excessive gestational weight

16

gain (GWG) and postpartum weight retention (PPWR). For FTO (rs9939609) gene,

17

women with the AA genotype were heavier in the body weight trajectory of pregnancy,

18

but not when their weight had been adjusted for pre-pregnancy BMI (βFTO=0.5 kg;

19

95%CI -1.9 to 3.0). These women started pregnancy heavier but gained less weight

20

(FTO*gestational age=-0.1; 95%CI: -0.2 to 0.03) when compared to those with at least

21

one T allele.

22

Conclusions: The FTO (rs9939609) AA genotype is positively associated with pre-

23

pregnancy excessive weight. We found no evidence of a significant effect of the MC4R

24

(rs17782313) or the FTO (rs9939609) gene polymorphisms on the GWG and PPWR.

25

Keywords: pregnancy, weight gain, weight retention, polymorphism, cohort study.

26

AC C

EP

TE D

M AN U

SC

RI PT

1

4

ACCEPTED MANUSCRIPT 27

Introduction Pregnancy is a complex process in women lifecycle, in which several metabolic

29

adaptations happen to support fetal development, delivery and lactation. Excessive

30

gestational weight gain (GWG) is associated with increased risks of pregnancy-induced

31

hypertension, caesarean delivery and a large-for-gestational age infant [1, 2]. It is also

32

the primary factor contributing to increased postpartum weight retention [3, 4] and may

33

contribute to the epidemic of obesity among women of childbearing age [5, 6].

RI PT

28

An individual’s susceptibility to obesity is thought to result from a combination

35

of their genetics, behavior and environment [7, 8]. Genetic factors play an important

36

role in the development of obesity. Up to date, several single nucleotide polymorphisms

37

(SNPs) of obesogenic genes associated with the increased risk of overweight or obesity

38

and with behavioral risk factors have been published [8, 9]. In particular, the variant

39

rs9939609 T/A of the fat mass and obesity-associated protein (FTO) gene, located in the

40

intron 1 on chromosome 16q12.2 [10], and the variant rs17782313 T/C of the

41

melanocortin-4 receptor (MC4R) gene, mapped 188 kb downstream of the gene and

42

encoded by a single exon gene on chromosome 18q22 [11], have been well documented

43

as major contributors to obesity populations [10, 12-15].

EP

TE D

M AN U

SC

34

Given the relevance of maternal and fetal fat to overall GWG, it is conceivable

45

that genetic variants that are known to be associated with adiposity might be associated

46

with GWG [16]. To the best of our knowledge, few studies were able to investigate the

47

longitudinal association between polymorphism (FTO and MC4R) and the body weight

48

throughout pregnancy. The results are still inconclusive [16-18], requiring more

49

research for elucidating the role of these genetic markers on pregnancy body weight

50

change. Thus, the aim of the present study was to examine the association of the FTO

AC C

44

5

ACCEPTED MANUSCRIPT 51

(rs9939609) and MC4R (rs17782313) gene polymorphisms with changes in maternal

52

body weight during pregnancy.

54

Methods

55

Study design

RI PT

53

We analyzed data of a prospective cohort of pregnant women who received

57

prenatal care at public health center in the city of Rio de Janeiro, Brazil, from

58

November 2009 to October 2011. Data were collected at three time points during the

59

pregnancy: 5th to 13th (first trimester), 20th to 26th (second trimester) and 30th to 36th

60

(third trimester) gestational weeks and in early postpartum (30 to 45 days).

M AN U

SC

56

Pregnant women were enrolled in the cohort if they were between their 5th and

62

13th weeks of gestation, between 20 and 40 years of age, without any history of

63

infectious or chronic diseases (except obesity) and had the intention to attend the

64

prenatal care in the selected public health center.

TE D

61

Figure 1 shows the flow the recruitment and selection of the participants from

66

the current study. At enrollment, 299 women were interviewed, and only those meeting

67

the eligibility criteria were included in the baseline sample and follow-up. Women were

68

excluded for the following reasons: confirmed pre-gestational infection diagnosis or

69

non-communicable diseases (n=14), > 13 gestational weeks at enrollment (n=16), or

70

abandoned prenatal care at the public health center (n=10). We further excluded from

71

the analysis pregnant women who suffered stillbirths (n=5) or miscarriages (n=25); had

72

twin pregnancies (n=4); did not attend the baseline visit (n=5); were missing self-

73

reported pre-pregnancy weight (n=8); and were lost to follow-up (n=9). We also

74

excluded data from 67 participants because we did not have genotype SNPs. After these

75

exclusions data from 136 participants were available for analysis.

AC C

EP

65

6

ACCEPTED MANUSCRIPT In general, no differences were observed in relation to the anthropometric and

77

socioeconomic variables when we compared women who were lost to genotyping with

78

those who had data; although, the women included in this study were younger and

79

presented a lower frequency of pre-pregnancy LTPA than those without genetic

80

information (Supplementary Table 1). However, the analyses of outcomes of interest

81

were adjusted for these variables.

RI PT

76

82

Anthropometric measurements

SC

83

At enrollment, the women were asked to record their pre-pregnancy weight.

85

Their weights (kg) were measured during pregnancy and postpartum using a digital

86

scale (Filizzola PL 150, Filizzola Ltda, Brazil). We also obtained the weight recorded in

87

the last visit of prenatal care (36th to 42th week). Height (cm) was measured only in the

88

first trimester using a portable stadiometer (Seca Ltda, Hamburg, Germany).

M AN U

84

Pre-pregnancy body mass index (BMI) [weight (kg)/height2 (m)] was calculated

90

using the self-reported pre-pregnancy weight, and the postpartum BMI, by using the

91

weight measured in the early postpartum.

TE D

89

Total GWG (kg) was estimated using the difference between the last weight

93

measured prior to delivery and the self-reported pre-pregnancy weight. Maternal weight

94

gains up to the first, second and third trimesters were calculated and divided by the

95

weeks of gestation. In addition, women were classified according to categories of GWG

96

(insufficient, normal and excessive) based on the Institute of Medicine (IOM)

97

recommendations [19].

98 99

AC C

EP

92

Postpartum weight retention (PPWR) was calculated by the difference between the early postpartum weight and the self-reported pre-pregnancy weight.

7

ACCEPTED MANUSCRIPT 100

The gestational age (weeks) was estimated based on the first ultrasound

101

performed prior to the 24th week of gestation. For those without an ultrasound or when

102

the first ultrasound was performed after the 24th week of gestation, the gestational age

103

was calculated based on the reported date of the last menstrual period (n=2).

105

RI PT

104

Covariates

Structured interviews were conducted to obtain the following maternal data:

107

maternal age at baseline (years), self-reported skin color (white/black or mixed), living

108

with partner (yes/no), education (years of schooling), self-reported pre-pregnancy

109

leisure-time physical activity (LTPA) (no/yes), smoking habit (non-smoking, former

110

smoker and current smokes), parity (number of deliveries) and the sex of the child. The

111

total energy intake (kcal/d) was assessed with a validated food frequency questionnaire

112

(FFQ) [20], which was administered in the first and third trimesters of the pregnancy.

M AN U

SC

106

114

Genotyping

TE D

113

At baseline visit, blood samples (5 ml) were collected, processed and stored at -

116

80º C until polymorphisms analysis. DNA was extracted by phenol-chloroform method.

117

SNP rs17782313 and rs9939609 of the MC4R and FTO genes, respectively, were

118

analyzed by Real Time PCR amplification (StepOnePlus™, Life Technologies) using

119

an allelic discrimination assay (TaqMan® Genotyping Master Mix assay, Life

120

Technologies). Duplicates were performed in 10% of sample with ≥ 99% agreement

121

rates.

122 123 124

AC C

EP

115

8

ACCEPTED MANUSCRIPT 125

Statistical analyses The continuous variables in asymmetric distributions were expressed as the

127

medians and interquartile range (IQR), whereas the measurement data in symmetric

128

distributions were presented as the means and standard deviations (SD) and absolute and

129

relative frequencies to describe categorical ones.

RI PT

126

We employed additive and recessive models for the FTO (rs9939609) gene

131

because previous studies had demonstrated that one risk allele has little or no effect and

132

two are required to cross the threshold [10] [12]. We employed only the dominant

133

model for the MC4R (rs17782313) gene because the number of minor allele

134

homozygotes was small in our sample (n=4). The participants were further classified

135

into two groups to investigate the combined effect of MC4R (rs17782313) and FTO

136

(rs9939609): (1) those with at least one risk allele of each polymorphism or (2) those

137

carrying only one risk allele from one of the polymorphisms or none of the risk alleles.

M AN U

SC

130

Multiple linear regression models were performed to examine the associations

139

between the changes in maternal body weight, assessed as GWG and PPWR and the

140

maternal genotypes. In addition, Poisson’s regression with robust variance was used to

141

address the association between genotypes and GWG categories according to IOM [19]

142

and also with pre-pregnancy BMI classified into two categories (<25 or ≥25 kg/m²),

143

allowing an estimation of the relative risks (RR) and it’s 95% confidence interval (95%

144

CI).

EP

AC C

145

TE D

138

Longitudinal linear mixed-effects (LME) models were performed to test the

146

association between the polymorphisms and weight trajectories during pregnancy.

147

These models do not assume linear weight gain over the period of gestation but allow

148

for the gradient of weight across pregnancy. LME models account for random variation

149

among individuals and between individuals [21], which allows for the estimation of

9

ACCEPTED MANUSCRIPT subject-specific means. These models also account for unbalanced data due to

151

measurements at irregular time points of observation. Gestational age (weeks) was

152

included in all models both as random and fixed effects to adjust for variations in

153

weight over time. The FTO/MC4R polymorphisms and all other covariates were

154

analyzed as fixed-effects. The quadratic gestational age term was included in all models

155

to adjust for the longitudinal changes in weight during pregnancy that resemble a

156

parabola. We also tested the interaction term between genotypes and gestational age to

157

evaluate whether the association between weight and genotypes differed between the

158

gestational weeks. The dependencies in the data were handled with an unstructured

159

covariance matrix.

M AN U

SC

RI PT

150

The multivariable analyses were adjusted for potential confounders. Covariates

161

were chosen as potential confounding factors based on the biological plausibility of the

162

association and their p-values in the bivariate analysis with each of the outcome

163

variables. All covariates that presented a p-value ≤0.20 in the bivariate analysis were

164

selected to compose the final model of each outcome variable in the regression analyses.

165

Because the association between pre-pregnancy BMI and gestational weight gain is

166

nonlinear, both linear and quadratic terms were included in the models.

EP

TE D

160

Different models were constructed for each outcome and we used the restricted

168

log-likelihood and Akaike’s information criterion as global fit criteria to select the best

169

model. A significance level at 5% was considered in all the analysis. Statistical analyses

170

were performed using STATA software (version 12.0, College Station, Texas, USA).

AC C

167

171 172

Ethical procedures

173

The study protocol was approved by the Ethics Committee of the Municipality

174

Secretary of Health of Rio de Janeiro City (Protocol number: 0139.0.314.000-09). All

10

ACCEPTED MANUSCRIPT 175

subjects enrolled in this study gave written informed consent for their participation after

176

an explanation of the study.

177

179 180

Results The median age of participants was 27 (IQR: 22 to 31) years old, and 73.5%

RI PT

178

women reported as having black or mixed skin color (Table 1).

The genotype frequencies were TT=33.8%, AT=49.3% and AA=16.9% for the

182

FTO (rs9939609) and TT=64.0%, CT=33.1% and CC=2.9% for the MC4R

183

(rs17782313). These frequencies were in Hardy–Weinberg equilibrium for both SNPs.

SC

181

The AA genotype of the FTO (rs9939609) gene, for the additive and recessive

185

models, was significantly associated with higher pre-pregnancy mean weights (p=0.01;

186

p<0.01, respectively) and BMI (p<0.01) and with higher gestational body weight at the

187

first (p=0.02; p=0.01, respectively) and the second (p=0.05; p=0.02, respectively)

188

trimesters in comparison to the other genotypes, but not with maternal height. Women

189

having the polymorphic MC4R (rs17782313) genotype did not show any statistically

190

significant difference for pre-pregnancy weight, BMI or maternal height (Table 1).

TE D

M AN U

184

Women with the AA genotype of the FTO (rs9939609) gene presented a higher

192

risk of pre-pregnancy excessive weight compared to those with AT/TT genotypes

193

(RR=2.1; 95%CI 1.4 to 3.1), even after adjustment for maternal pre-pregnancy LTPA,

194

age, skin color, education, parity, smoking and height (Table 2).

AC C

195

EP

191

The risk alleles of both polymorphisms were no longer statistically associated

196

with GWG after adjusting for pre-pregnancy BMI. In our adjusted models, neither the

197

FTO (rs9939609) (RR=0.9; 95%CI 0.5 to 1.6) nor the MC4R (rs17782313) (RR=1.1;

198

95%CI 0.7 to 1.7) risk alleles were associated with a higher risk of excessive GWG

11

ACCEPTED MANUSCRIPT 199

according to IOM guidelines (Table 2). Similar results were found for the combined

200

effect of the MC4R (rs17782313) and FTO (rs9939609) (Supplementary Table 2). We did not find a significant association between these SNPs and the PPWR

202

(Tables 1 and 2). Women with the AA genotype were at a higher risk of postpartum

203

overweight than the TT/AT carriers (RR=1.7; 95%CI 1.3 to 2.3); however, this result

204

lost significance after adjustment for the pre-pregnancy BMI (RR=1.0; 95%CI 0.8 to

205

1.2) (Table 2).

RI PT

201

Homozygous women for the A allele of the FTO (rs9939609) gene were

207

approximately 9 kg heavier during pregnancy, compared to those with the TT/AT

208

alleles in our adjusted models for gestational age, education, smoking, skin color, sex of

209

the child, age, pre-pregnancy LTPA, parity and height. The result lost significance after

210

adjustment for the pre-pregnancy BMI (βFTO=0.5 kg; 95%CI -1.9 to 3.0) (Table 3).

211

Women with the AA genotype of the FTO (rs9939609) gene presented a lower rate of

212

change of weight trajectories during pregnancy (βFTO*time =-0.1; 95%CI -0.2 to -0.03),

213

compared to those with the TT/AT (Supplementary Figure 1 and Table 3). We found

214

no association between the MC4R (rs17782313) polymorphism and weight trajectories

215

during pregnancy in our adjusted models for gestational age, education, smoking, skin

216

color, sex of the child, age, pre-pregnancy LTPA, parity and pre-pregnancy BMI (Table

217

3).

219

M AN U

TE D

EP

AC C

218

SC

206

Discussion

220

This study has three main findings. First, we observed that the FTO (rs9939609)

221

polymorphism but not the MC4R (rs17782313) polymorphism was positively associated

222

with pre-pregnancy excessive weight. Secondly, we did not find a significant

223

association between any of the polymorphisms and GWG or PPWR. Finally, we

12

ACCEPTED MANUSCRIPT 224

observed that women who were homozygous for the FTO (rs9939609) risk allele were

225

heavier when beginning pregnancy but gained less weight throughout gestation than

226

those with the AT or TT genotypes. We investigated the FTO SNP rs9939609 and the MC4R SNP rs17782313 in our

228

study because they have shown to have a strong association with body weight in

229

previous studies [10, 12-15]. The association of these genes polymorphisms with the

230

obesity phenotype in a multiethnic group such as the Brazilian population has not been

231

previously reported. The minor allele frequency observed in this study (0.42 and 0.20

232

for rs9939609 A-allele and rs17782313 C-allele, respectively) were quite similar to

233

those in the international HapMap project CEU data (0.45 and 0.26, respectively) and

234

within the range of reported values in other studies (0.44 and 0.24, respectively) [10,

235

14].

M AN U

SC

RI PT

227

The mechanisms underlying the physiological relationship between FTO

237

(rs9939609) and MC4R (rs17782313) and alterations of body weight are yet to be

238

elucidated. It is known that FTO and MC4R are highly expressed in the hypothalamic

239

region [22, 23], an area that is known to be involved in the regulation of appetite.

240

Habitual diet is one of the many environmental factors that potentially contribute to

241

inter-individual differences in body fat mass. In this study, we did not find significant

242

association between FTO SNP rs9939609 and MC4R SNP rs17782313 with energy

243

intake, which is in line with the findings of Hasselbalch et al. [24], but inconsistent with

244

the findings of other studies [25, 26]. The dietary information collected on our study is

245

based on an extensive self-reported FFQ and several components of dietary intake were

246

studied.

AC C

EP

TE D

236

247

We identified that pregnant women with the risk alleles (AA) of FTO

248

(rs9939609) had a higher risk to have pre-pregnancy overweight BMI (RR=2.1 (95%CI

13

ACCEPTED MANUSCRIPT 1.4-3.1, p<0.01), which is consistent with the findings of Lawlor et al. [16] study, where

250

FTO (rs9939609) was associated with pre-pregnancy BMI. These British pregnant

251

women presented mean difference per risk allele of 0.40 kg/m2 (95%CI 0.25-0.54). In

252

contrast, Groth et al. [18] reported that the FTO (rs9939609) risk alleles were not

253

associated with pre-pregnancy BMI in a study of low-income black pregnant women.

254

Our sample is composed of low-income women, but the degree of miscegenation in

255

Brazil is very high and the racial/ethnic composition of this study was based only in

256

self-reported skin color (73.5% black or mixed). Different populations are exposed to

257

the different environmental and genetic influences that may interact with genetic

258

variants. On the other hand, our results are consistent with investigations that have

259

indicated that FTO plays a key role in changes in adiposity-related phenotypes in

260

populations around the world [12, 13, 27].

M AN U

SC

RI PT

249

In our study, carriers of two copies of the risk allele of FTO (rs9939609) had

262

significantly higher body weight than the homozygous subjects showing the major

263

allele, which is in agreement with the results of previous studies [10, 28].

TE D

261

Common variants near MC4R have been reproducibly associated with the fat

265

mass, weight and risk of obesity [9, 14, 29]. Lawler et al. [16] found no association

266

between MC4R (rs17782313) and pre-pregnancy BMI but found a positive association

267

with the pre-pregnancy weight (p=0.001). Such association was not observed in our

268

study. Other studies have also failed to find a significant association [30, 15].

AC C

269

EP

264

A longitudinal study investigating the life-course effects of variants in the FTO

270

gene (rs9939609) and near the MC4R gene (rs17782313) demonstrated that the effects

271

strengthen throughout childhood and peak at age 20 before weakening during adulthood

272

[31]. In our study, at baseline, the median maternal age was 27 (IQR: 22 to 31) years

273

old. The effects of the FTO and MC4R genes on pre-pregnancy weight may have

14

ACCEPTED MANUSCRIPT 274

occurred due to the effects of the genes on promoting weight gain during the youngest

275

age and may continue at the same level throughout life, depending on the effect size of

276

the polymorphism and exposure to an obesogenic environment [30]. In this study, no association was found between the risk alleles of the

278

FTO/MC4R genes with GWG by trimester, total GWG and risk of excessive GWG,

279

according to the IOM categories in our adjusted models. Lawlor et al. [16] indicated that

280

the FTO (rs9939609) and MC4R (rs17782313) genes were not statistically associated

281

with GWG by the period of pregnancy and by IOM categories. Stuebe et al. [17] found

282

higher GWG among African American participants with two MC4R (rs17782313) risk

283

alleles compared with women with no MC4R (rs17782313) risk alleles but among

284

Caucasian women, MC4R carriage was inversely associated with weight gain.

285

For FTO (rs9939609) gene, Stuebe et al. [17] reported that Caucasian women

286

homozygous for the risk allele, so in thin as in obese, gained more weight than low-risk

287

allele carriers, but among women of average pre-pregnancy BMI, weight gain was

288

similar in spite of allele carriage; although, they found no association between this SNP

289

and the greater GWG. GWG includes several other components (i.e. the fetus, amniotic

290

fluid, and placenta). In addition, during pregnancy, women go through many biological,

291

hormonal and behavioral changes, which could have the potential to mask smaller

292

genetic associations and may interact and modify the susceptibility to obesity by the

293

FTO and MC4R variants, influencing the genetic contributions on GWG [19].

SC

M AN U

TE D

EP

AC C

294

RI PT

277

We observed that women with the AA genotype of the FTO (rs9939609) gene

295

had higher body weight during pregnancy; however, the association was no longer

296

significant after adjustment for the pre-pregnancy BMI. These women gained less

297

weight compared to those with the AT/TT genotypes based on the LME models. This

298

discrepancy might reflect the fact that obese women tend to gain less weight during

15

ACCEPTED MANUSCRIPT 299

pregnancy [32]. We also did not find an association between the MC4R (rs17782313)

300

polymorphism and weight trajectories during pregnancy. We did not observe an association between the FTO (rs9939609) and MC4R

302

(rs17782313) polymorphisms on the PPWR at <45 days after adjustment for the pre-

303

pregnancy BMI, which is in agreement with the study by Lawlor et al. [16], where the

304

PPWR was calculated at approximately 8 weeks after delivery. The PPWR is

305

presumably due to a combination of several factors, such as dietary intake, lack of

306

physical activity, lactation, smoking status, pre-pregnancy BMI, GWG and parity [33]

307

and also genetics [4].

SC

RI PT

301

Our study has strengths and limitations. The main strength is the availability of

309

standardized longitudinal weight measurements across pregnancy, which allowed

310

analysis of the genetic associations with the changes of body weight over time. In

311

addition, we controlled our results for important confounding variables, such as

312

gestational age, education, smoking, skin color, pre-pregnancy LTPA, age and parity.

313

On the other hand, some limitations from our study must be highlighted. We relied on

314

self-reported pre-pregnancy weight as is the case in most studies of GWG, and

315

misclassification of GWG may not be ruled out as for women who inaccurately reported

316

their pre-pregnancy weights. However, previous studies have shown that self-reported

317

pre-pregnancy weight is a good approximation of the true weight [34, 35]. Another

318

limitation of our study is the small sample size, which resulted in a lower power for

319

detecting a statistically significant effect of the polymorphism in the weight changes

320

during pregnancy. Besides that, we were able to find that FTO (rs9939609)

321

polymorphism showed some association with pre-pregnancy BMI, but not with MC4R

322

(rs17782313) polymorphism. This could be due to the lower power of the study, or

323

indeed, to a real small effect size in this relationship. Our results agreed in the direction

AC C

EP

TE D

M AN U

308

16

ACCEPTED MANUSCRIPT 324

and significance of the association when compared to the other two studies on the same

325

topic (Lawlor and Stuebe), that definitely presented a bigger sample size to detect an

326

effect size for the studied polymorphisms [16, 17]. In conclusion, our study found that the SNP in the FTO gene (rs9939609) but

328

not in the MC4R gene (rs17782313) was significantly associated with pre-pregnancy

329

excessive weight. However, these women gained less weight throughout gestation than

330

women with the AT and TT genotypes, and neither of the polymorphisms were

331

statistically associated with excessive GWG and PPWR. The characterization of effects

332

of the genetic factor implicated in weight gain during pregnancy and postpartum may

333

potentially guide targeted intervention for preventing obesity and the avoidance of

334

adverse pregnancy complications. However, further replications of association studies

335

with different ethnicities and larger cohorts or cohort consortiums and a sampling

336

strategy that collects additional time points throughout the pregnancy are necessary to

337

improve our understanding of the specific roles of FTO and MC4R in gestational

338

weight.

339

Acknowledgements

EP

340

TE D

M AN U

SC

RI PT

327

We would like to acknowledge Prof. Rosane Silva for her technical support in

342

the genotyping analysis and allowing us to access to the Laboratory of Macromolecular

343

Metabolism Firmino Torres de Castro. We also thank Prof. Maria das Graças Tavares

344

do Carmo for allowing us to work at the Laboratory of Nutritional Biochemistry.

345 346 347 348

AC C

341

17

ACCEPTED MANUSCRIPT 349

Financial support

350

This study was supported by the National Council for Scientific and

351

Technological Development (CNPq) and the Carlos Chagas Filho Foundation for

352

Research Support of Rio de Janeiro State (FAPERJ).

Conflict of interest

355

The authors declare that they have no conflict of interest

AC C

EP

TE D

M AN U

SC

354

RI PT

353

18

ACCEPTED MANUSCRIPT References 356

[1] Li N, Liu E, Guo J, Pan L, Li B, Wang P, et al. Maternal prepregnancy body mass

357

index and gestational weight gain on pregnancy outcomes. PloS One

358

2013;8:e82310. doi:10.1371/journal.pone.0082310. [2] Gaillard R, Durmuş B, Hofman A, Mackenbach JP, Steegers EA, Jaddoe VW.

360

Risk factors and outcomes of maternal obesity and excessive weight gain during

361

pregnancy. Obesity 2013;21:1046–55.

RI PT

359

[3] Mannan M, Doi SA, Mamun AA. Association between weight gain during

363

pregnancy and postpartum weight retention and obesity: a bias-adjusted meta-

364

analysis. Nutr Rev 2013;71:343–52.

M AN U

SC

362

365

[4] Rong K, Yu K, Han X, Szeto IM, Qin X, Wang J, et al. Pre-pregnancy BMI,

366

gestational weight gain and postpartum weight retention: a meta-analysis of

367

observational studies. Public Health Nutr 2015;18:2172–82. [5] Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, et al. Global,

369

regional, and national prevalence of overweight and obesity in children and adults

370

during 1980–2013: a systematic analysis for the Global Burden of Disease Study

371

2013. The Lancet 2014;384:766–81.

EP

TE D

368

[6] Raatikainen K, Heiskanen N, Heinonen S. Transition from Overweight to Obesity

373

Worsens Pregnancy Outcome in a BMI-dependent Manner. Obesity 2006;14:165–

374

AC C

372

71.

375

[7] Warrington NM, Wu YY, Pennell CE, Marsh JA, Beilin LJ, Palmer LJ, et al.

376

Modelling BMI trajectories in children for genetic association studies. PloS One

377

2013;8:e53897.

19

ACCEPTED MANUSCRIPT 378

[8] Albuquerque D, Stice E, Rodríguez-López R, Manco L, Nóbrega C. Current

379

review of genetics of human obesity: from molecular mechanisms to an

380

evolutionary perspective. Mol Genet Genomics 2015:1–31. [9] Speliotes EK, Willer CJ, Berndt SI, Monda KL, Thorleifsson G, Jackson AU, et al.

382

Association analyses of 249,796 individuals reveal 18 new loci associated with

383

body mass index. Nat Genet 2010;42:937–48.

RI PT

381

[10] Frayling TM, Timpson NJ, Weedon MN, Zeggini E, Freathy RM, Lindgren CM, et

385

al. A common variant in the FTO gene is associated with body mass index and

386

predisposes to childhood and adult obesity. Science 2007;316:889–94.

SC

384

[11] Farooqi IS, Keogh JM, Yeo GSH, Lank EJ, Cheetham T, O’Rahilly S. Clinical

388

spectrum of obesity and mutations in the melanocortin 4 receptor gene. N Engl J

389

Med 2003;348:1085–95. doi:10.1056/NEJMoa022050.

M AN U

387

[12] Dina C, Meyre D, Gallina S, Durand E, Körner A, Jacobson P, et al. Variation in

391

FTO contributes to childhood obesity and severe adult obesity. Nat Genet

392

2007;39:724–6.

394

[13] Loos RJ, Yeo GS. The bigger picture of FTO [mdash] the first GWAS-identified obesity gene. Nat Rev Endocrinol 2014;10:51–61.

EP

393

TE D

390

[14] Loos RJ, Lindgren CM, Li S, Wheeler E, Zhao JH, Prokopenko I, et al. Common

396

variants near MC4R are associated with fat mass, weight and risk of obesity. Nat

397

AC C

395

Genet 2008;40:768–75.

398

[15] Liu G, Zhu H, Dong Y, Podolsky RH, Treiber FA, Snieder H. Influence of

399

common variants in FTO and near INSIG2 and MC4R on growth curves for

400

adiposity in African– and European–American youth. Eur J Epidemiol

401

2011;26:463–73. doi:10.1007/s10654-011-9583-4.

20

ACCEPTED MANUSCRIPT 402

[16] Lawlor DA, Fraser A, Macdonald-Wallis C, Nelson SM, Palmer TM, Smith GD, et

403

al. Maternal and offspring adiposity-related genetic variants and gestational weight

404

gain. Am J Clin Nutr 2011;94:149–55. [17] Stuebe AM, Lyon H, Herring AH, Ghosh J, Wise A, North KE, et al. Obesity and

406

diabetes genetic variants associated with gestational weight gain. Am J Obstet

407

Gynecol 2010;203:283.e1–283.e17.

[18] Groth SW, Morrison-Beedy D. GNB3 and FTO Polymorphisms and Pregnancy

409

Weight

Gain

in

Black

Women.

410

doi:10.1177/1099800414561118.

Biol

Res

Nurs

2015;17:405–12.

SC

408

RI PT

405

[19] Institute of Medicine (US) and National Research Council (US) Committee to

412

Reexamine IOM Pregnancy Weight Guidelines. Weight Gain During Pregnancy:

413

Reexamining the Guidelines. Washington (DC): National Academies Press (US);

414

2009.

417 418

TE D

416

[20] Sichieri R, Everhart JE. Validity of a Brazilian food frequency questionnaire against dietary recalls and estimated energy intake. Nutr Res 1998;18:1649–59. [21] Verbeke G, Molenberghs G. Linear mixed models for longitudinal data. Springer Science & Business Media; 2009.

EP

415

M AN U

411

[22] Gerken T, Girard CA, Tung Y-CL, Webby CJ, Saudek V, Hewitson KS, et al. The

420

Obesity-Associated FTO Gene Encodes a 2-Oxoglutarate-Dependent Nucleic Acid

421 422 423

AC C

419

Demethylase. Science 2007;318:1469–72. doi:10.1126/science.1151710.

[23] Tao Y-X. The melanocortin-4 receptor: physiology, pharmacology, and pathophysiology. Endocr Rev 2010;31:506–43. doi:10.1210/er.2009-0037.

424

[24] Hasselbalch AL, Angquist L, Christiansen L, Heitmann BL, Kyvik KO, Sørensen

425

TIA. A variant in the fat mass and obesity-associated gene (FTO) and variants near

21

ACCEPTED MANUSCRIPT 426

the melanocortin-4 receptor gene (MC4R) do not influence dietary intake. J Nutr

427

2010;140:831–4. doi:10.3945/jn.109.114439. [25] Speakman JR, Rance KA, Johnstone AM. Polymorphisms of the FTO Gene Are

429

Associated With Variation in Energy Intake, but not Energy Expenditure. Obesity

430

2008;16:1961–5. doi:10.1038/oby.2008.318.

RI PT

428

431

[26] Qi L, Kraft P, Hunter DJ, Hu FB. The common obesity variant near MC4R gene is

432

associated with higher intakes of total energy and dietary fat, weight change and

433

diabetes

434

doi:10.1093/hmg/ddn242.

in

women.

Hum

Mol

Genet

2008;17:3502–8.

SC

risk

[27] Hallman DM, Friedel VC, Eissa MA, Boerwinkle E, Huber JC, Harrist RB, et al.

436

The association of variants in the FTO gene with longitudinal body mass index

437

profiles in non-Hispanic white children and adolescents. Int J Obes 2012;36:61–8.

M AN U

435

[28] Sentinelli F, Incani M, Coccia F, Capoccia D, Cambuli VM, Romeo S, et al.

439

Association of FTO Polymorphisms with Early Age of Obesity in Obese Italian

440

Subjects. J Diabetes Res 2012;2012:e872176. doi:10.1155/2012/872176.

TE D

438

[29] Willer CJ, Speliotes EK, Loos RJ, Li S, Lindgren CM, Heid IM, et al. Six new loci

442

associated with body mass index highlight a neuronal influence on body weight

443

regulation. Nat Genet 2009;41:25–34.

EP

441

[30] Vasan SK, Fall T, Neville MJ, Antonisamy B, Fall CH, Geethanjali FS, et al.

445

Associations of Variants in FTO and Near MC4R With Obesity Traits in South

446

AC C

444

Asian Indians. Obesity 2012;20:2268–77. doi:10.1038/oby.2012.64.

447

[31] Hardy R, Wills AK, Wong A, Elks CE, Wareham NJ, Loos RJ, et al. Life course

448

variations in the associations between FTO and MC4R gene variants and body

449

size. Hum Mol Genet 2010;19:545–52.

22

ACCEPTED MANUSCRIPT 450

[32] Chu SY, Callaghan WM, Bish CL, D’Angelo D. Gestational weight gain by body

451

mass index among US women delivering live births, 2004-2005: fueling future

452

obesity. Am J Obstet Gynecol 2009;200:271.e1–217.e7.

454

[33] Gunderson EP, Abrams B. Epidemiology of Gestational Weight Changes After Pregnancy. Epidemiol Rev 2000;22:261–74.

RI PT

453

[34] Park S, Sappenfield WM, Bish C, Bensyl DM, Goodman D, Menges J. Reliability

456

and validity of birth certificate prepregnancy weight and height among women

457

enrolled in prenatal WIC program: Florida, 2005. Matern Child Health J

458

2011;15:851–9.

SC

455

[35] Shin D, Chung H, Weatherspoon L, Song WO. Validity of prepregnancy weight

460

status estimated from self-reported height and weight. Matern Child Health J

461

2014;18:1667–74.

AC C

EP

TE D

M AN U

459

ACCEPTED MANUSCRIPT

RI PT

Table 1 - Maternal characteristics according to the FTO (rs9939609) and the MC4R (rs17782313) gene polymorphisms

FTO (rs9939609) N

Recessive model

Total TT/AT

AA

p-value

TT

AA

p

TT

CT/CC

p-value

27 (22 to 31)

27 (23 to 31)

24 (22 to 32)

0.36

27 (23 to 31)

27 (22 to 31)

24 (22 to 32)

0.60

27 (22 to 32)

27 (22 to 31)

0.90

Calorie intake1, kcal/d

133

136 136

100 (73.5) 9 (6 to 11)

85 (75.2) 9 (6 to 11)

15 (65.2) 9 (7 to 11)

0.32 0.96

32 (69.6) 8.5 (5 to 11)

53 (79.1) 10 (7 to 11)

15 (65.2) 9 (7 to 11)

0.32 0.12

651(70.1) 10 (6 to 11)

39 (79.6) 9 (6 to 11)

0.23 0.73

111 (81.6)

93 (82.3)

18 (78.3)

0.77

41 (89.1)

52 (77.6)

18 (78.3)

0.26

71 (81.6)

40 (81.6)

0.99

97 (71.3)

80 (70.8)

17 (73.9)

0.76

31 (67.4)

49 (73.1)

17 (73.9)

0.77

61 (70.1)

36 (73.5)

0.68

108 (80.6)

88 (79.3)

20 (87.0)

0.56

36 (80.0)

52 (78.8)

20 (87.0)

0.69

67 (78.8)

41 (83.7)

0.49

55 (40.4) 45 (33.1) 36 (26.5)

45 (39.8) 40 (35.4) 28 (24.8)

10 (43.5) 5 (21.7) 8 (34.8)

0.38

16 (34.8) 17 (37.0) 13 (28.2)

29 (43.3) 23 (34.3) 15 (22.4)

10 (43.5) 5 (21.7) 8 (34.8)

0.59

39 (44.8) 24 (27.6) 24 (27.6)

16 (32.6) 21 (42.9) 12 (24.5)

0.18

64 (47.8) 2279.7 (1895.1 to 2641.7)

54 (48.2) 2251.6 (1905.7 to 2610.7)

10 (45.5) 2567.5 (1648.5 to 2995.7)

18 (40.0) 2397.2 (1897.2 to 2713.8)

36 (53.7) 2199.1 (1963.8 to 2590.7)

10 (45.4) 2567.5 (1648.5 to 2995.7)

48 (55.8) 2361.8 (1883.1 to 2713.8)

16 (33.3) 2186.4 (1907.6 to 2609.3)

M AN U

136 136

136 134

EP

TE D

136

134

Dominant model

AT

Descriptive Maternal age1, years Self-reported skin color2 Black or mixed Education1, years Living with partner2 Yes Smoking habit at baseline2 Non-smoking Pre-pregnancy LTPA2 No Parity2 0 1 ≥2 Sex of the child2 Male

MC4R (rs17782313)

Additive model

SC

Characteristic

0.81 0.51

0.35 0.70

0.01 0.68

AC C

Self-reported pre-pregnancy weight3, kg 136 62.7 + 12.9 61.2 + 12.3 70.1 + 13.4 <0.01 61.8 + 11.4 60.7 + 13.0 70.1 + 13.4 0.01 61.5 + 13.0 64.7 + 12.6 0.16 Maternal height3, cm 136 159.5 + 6.5 159.4 + 6.4 160.1 + 6.8 0.62 159.9 + 6.2 159.1 + 6.6 160.1 + 6.8 0.72 158.9 + 5.8 160.7 + 7.5 0.12 Pre-pregnancy BMI3, kg/m2 136 24.6 + 4.5 24.0 + 4.2 27.4 + 5.3 <0.01 24.1 + 4.1b 23.9 + 4.3b 27.4 + 5.3a <0.01 24.3 + 4.9 25.0 + 3.9 0.43 Pregnancy weight in 1st trimester3, kg 136 63.8 + 12.5 62.5 + 12.1 70.2 + 12.8 0.01 62.7 + 11.4b 62.4 + 12.7b 70.2 + 12.8a 0.02 62.8 + 12.5 65.7 + 12.4 0.19 Pregnancy weight in 2nd trimester3, kg 128 68.6 + 12.2 67.4 + 12.0 74.1 + 12.0 0.02 68.4 + 11.6 66.7 + 12.4 74.1 + 12.0 0.05 67.4 + 11.9 70.7 + 12.7 0.14 Pregnancy weight in 3rd trimester 3, kg 135 75.3 + 12.5 74.4 + 12.6 79.7 + 11.0 0.06 75.0 + 11.6 73.9 + 13.3 79.7 + 11.1 0.16 74.4 + 11.9 76.8 + 13.5 0.30 Postpartum weight retention3, kg 122 3.9 + 5.4 4.2 + 5.2 2.4 + 6.3 0.18 3.8 + 5.1 4.4 + 5.4 2.4 + 6.3 0.36 4.3 + 5.4 3.2 + 5.5 0.26 Abbreviations: FTO - fat mass and obesity-associated gene; MC4R - melanocortin-4 receptor gene; LTPA - leisure time physical activity; BMI - body mass index. Data are presented as: 1median (IQR), p-value refers to Mann-Whitney U test or Kruskal-Wallis; 2absolute frequency (%), p-value refers to Pearson χ2 tests or Fisher's exact test; 3mean + SD, p-value refers to Student's t test or analysis of variance (ANOVA). Different letters in the same row indicate significant differences (Tukey's test p < 0.05).

ACCEPTED MANUSCRIPT

MC4R dominant model (rs17782313) risk alleles (CT/CC) Crude Adjusted Model 1 Model 2 β (95% CI) β (95% CI) β (95% CI) 10.5 (-131.2 to 152.2) 43.9 (-105.7 to 193.5) 93.7 (-52.3 to 239.6) -2.4 (-82.9 to 78.0) 33.2 (-50.9 to 117.3) 40.2 (-41.4 to 121.7) 25.6 (-53.8 to 104.9) 36.9 (-43.4 to 117.1) 26.0 (-56.2 to 108.2) -0.8 (-2.8 to 1.3) -0.1 (-2.2 to 2.1) 0.3 (-1.7 to 2.4) -265.8 (-690.4 to 158.8) -170.4 (-611.0 to 270.2) -80.0 (-298.9 to 138.9)

SC

GWG in 1st trimester1, g/wk GWG in 2nd trimester1, g/wk GWG in 3rd trimester1, g/wk Total GWG1, kg PPWR2†, g/wk

FTO recessive model (rs9939609) risk alleles (AA) Crude Adjusted Model 1 Model 2 β (95% CI) β (95% CI) β (95% CI) -163.4 (-345.5 to 18.7) -168.5 (-358.0 to 21.0) -74.4 (-265.6 to 116.8) -94.8 (-196.1 to 6.5) -100.1 (-205.4 to 5.2) -56.8 (-162.1 to 48.5) -85.0 (-184.5 to 14.5) -77.6 (-178.5 to 23.4) -90.9 (-196.3 to 14.6) -3.6 (-6.2 to -1.1)* -3.4 (-6.1 to -0.7)* -1.9 (-4.5 to 0.7) -505.9 (-1064.4 to 52.6) -474.2 (-1057.5 to 109.0) 131.2 (-166.6 to 428.9)

M AN U

Characteristics

RI PT

Table 2 - Associations of adiposity risk alleles (FTO and MC4R) with body weight changes before, during pregnancy and early postpartum

AC C

EP

TE D

RR pre-pregnancy BMI3 ‡ overweight (BMI ≥25 kg/m2) 2.2 (1.4 to 3.2)** 2.1 (1.4 to 3.1)** 2.1 (1.4 to 3.1)** 1.0 (0.6 to 1.6) 1.1 (0.7 to 1.7) 1.1 (0.7 to 1.7) 4‡ RR total GWG by IOM category insufficienta 0.8 (0.5 to 1.5) 0.8 (0.5 to 1.4) 0.8 (0.4 to 1.5) 0.9 (0.6 to 1.3) 0.9 (0.5 to 1.5) 0.9 (0.5 to 1.4) excessiveb 1.0 (0.5 to 1.7) 1.0 (0.6 to 1.7) 0.9 (0.5 to 1.6) 1.0 (0.7 to 1.5) 1.1 (0.7 to 1.7) 1.1 (0.7 to 1.7) RR postpartum BMI5 ‡ overweight (BMI ≥25 kg/m2) 1.7 (1.3 to 2.2)** 1.7 (1.3 to 2.3)** 1.0 (0.8 to 1.2) 1.1 (0.8 to 1.5) 1.2 (0.8 to 1.6) 0.9 (0.7 to 1.2) Abbreviations: FTO - fat mass and obesity-associated gene; MC4R - melanocortin-4 receptor; GWG - gestational weight gain; CI - confidence interval; RR - relative risk; PPWR - postpartum weight retention. * p<0.05; **p<0.01 † Calculated as the difference between post- and pre-pregnancy weights and adjusted for weeks since birth. ‡ Note that estimates in this row is relative risk (RR); Wald tests were used to derive the p values. 1 Model 1 was adjusted for sex of the child, pre-pregnancy LTPA, maternal age, skin color self-reported, parity and gestational age; Model 2= Model 1 plus pre-pregnancy BMI (linear and quadratic terms). 2 Model 1 was adjusted for sex of the child, pre-pregnancy LTPA, maternal age, skin color self-reported and parity; Model 2= Model 1 plus total weight gain. 3 Model 1 was adjusted for pre-pregnancy LTPA, maternal age, skin color self-reported, years of education, parity and smoking maternal status; Model 2= Model 1 plus maternal height. 4 Model 1 was adjusted for sex of the child, pre-pregnancy LTPA, maternal age, skin color self-reported and parity; Model 2= Model 1 plus pre-pregnancy BMI (linear and quadratic terms). 5 Model 1 was adjusted for pre-pregnancy LTPA, maternal age, skin color self-reported, years of education, parity and smoking maternal status; Model 2= Model 1 plus pre-pregnancy BMI (linear and quadratic terms). a Pre-pregnancy BMI: 18.5-24.9 kg/m2 (GWG <11.5 kg); 25.0-29.9 kg/m2 (GWG <7.0 kg); ≥ 30.0 kg/m2 (GWG <5.0 kg). b Pre-pregnancy BMI: 18.5-24.9 kg/m2 (GWG >16.0 kg); 25.0-29.9 kg/m2 (GWG >11.5 kg); ≥ 30.0 kg/m2 (GWG >9.0 kg).

ACCEPTED MANUSCRIPT

Table 3 - Longitudinal analysis between FTO (rs9939609), MC4R (rs17782313) and trajectory of body weight during pregnancy

FTO (rs9939609) TT or AT vs. AA Interaction term FTO#gestational age Likelihood AIC

8.7 (3.2 to 14.2)

0.002

-0.1 (-0.2 to -0.03) -1885.544 3789.089

0.004

2.7 (-1.5 to 6.9) -1891.617 3799.234

Aggregate score (FTO/MC4R) TT/TT or A or C vs. A e C Likelihood AIC

2.3 (-2.2 to 6.9) -1891.924 3799.848

0.206

TE D

MC4R (rs17782313) TT vs. CT or CC Likelihood AIC

RI PT

p

*

Model 22 β (95% CI) †

8.7 (3.9 to 13.6)

0.316

‡

p

Model 33 β (95% CI) †

p‡

<0.001

0.5 (-1.9 to 3.0)

0.667

-0.1 (-0.2 to -0.03) -1791.901 3619.802

0.007

-0.1 (-0.2 to -0.03) -1703.529 3445.057

0.007

2.7 (-1.0 to 6.4) -1797.900 3629.801

0.156

1.7 (-0.2 to 3.6) -1705.626 3447.252

0.071

2.2 (-1.8 to 6.2) -1798.336 3630.673

0.285

0.4 (-1.6 to 2.5) -1707.145 3450.291

0.679

M AN U

Trajectory of maternal weight

‡

SC

Model 11 β (95% CI) †

AC C

EP

Abbreviations: BMI - body mass index; CI - confidence interval; FTO - fat mass and obesity-associated gene; MC4R - melanocortin-4 receptor; AIC - Akaike’s information criterion. † β = Linear mixed-effect regression coefficient. ‡ p-value refers to the maximum likelihood estimator. 1 Model 1 was adjusted for gestational age (weeks) and also for quadratic gestational age; 2 Model 2 = Model 1 plus years of education, smoking maternal status, self-reported skin color, sex of the child, maternal age, pre-pregnancy LTPA, parity and maternal height; 3 Model 3 = Model 2 minus maternal height and plus pre-pregnancy BMI (linear and quadratic terms). * Weight at 0 week gestation (referred to as “self-reported pre-pregnancy weight”) and weight change from 5 to 13 weeks, from 20 to 26 weeks, from 30 to 36 weeks and last weight before to delivery.

ACCEPTED MANUSCRIPT Pregnant women at enrollment: 299 40 did not meet eligibility criteria: confirmed pre-gestational infection diagnosis or noncommunicable diseases [n=14], >13 gestational weeks at enrollment [n=16] or abandoned prenatal care at the public health center [n=10].

Stillbirth [n=05] Miscarriage [n=25]

225 participants

M AN U

13 Missing data: did not attend baseline visit [05], missing selfreported pre-pregnancy weight [n=08].

SC

Twin pregnancies [n=4]

RI PT

Folow-up: 259 participants

Losses to follow-up: only information at baseline [n=09].

203 participants of the current study

TE D

64 did not have blood samples for genotyping and 03 had only genotiping for one of the polymorphisms.

AC C

EP

Final sample 136 participants included in study with FTO and MC4R data

Figure 1 - Flow chart illustrating the recruitment and selection of the study population

ACCEPTED MANUSCRIPT Highlights -

Single nucleotide polymorphism (SNP) fat mass and obesity-associated (FTO) gene (rs9939609) was significantly associated with pre-pregnancy excessive weight.

-

SNP melanocortin 4 receptor (MC4R) gene (rs17782313) was not associated with

No associations were found between of the MC4R or the FTO gene polymorphisms

EP

TE D

M AN U

SC

and GWG and PPWR.

AC C

-

RI PT

pre-pregnancy excessive weight.