Abstract
Background: Synchrony between daily feeding-fasting signals and circadian rhythms has been shown to improve metabolic health in animals and adult humans, but the potential programming effect on fetal growth is unknown.
Objective: We examined the associations of the maternal night-fasting interval during pregnancy with offspring birth size and adiposity.
Methods: This was a cross-sectional study of mother-offspring dyads within the Growing Up in Singapore Towards healthy Outcomes (GUSTO) cohort. For 384 mothers aged 30.8 ± 4.8 y (mean ± SD), the night-fasting interval at 26–28 wk of gestation was determined from a 3-d food diary based on the average of the fasting duration at night (1900–0659). Offspring birth weight, length, and head circumference were measured and converted to weight-for-gestational age (GA), length-for-GA, and head circumference–for-GA z scores, respectively, by using local customized percentile charts. The percentage of neonatal total body fat (TBF) was derived by using a validated prediction equation. Multivariable general linear models, stratified by child sex, were performed.
Results: The mean ± SD maternal night-fasting interval was 9.9 ± 1.3 h. In infant girls, each 1-h increase in the maternal night-fasting interval was associated with a 0.22-SD (95% CI: 0.05-, 0.40-SD; P = 0.013) increase in birth head circumference–for-GA and a 0.84% (95% CI: 0.19%, 1.49%; P = 0.012) increase in TBF at birth, after adjustment for confounders. In infant boys, no associations were observed between the maternal night-fasting interval and birth size or TBF.
Conclusions: An increased maternal night-fasting interval in the late second trimester of pregnancy is associated with increased birth head circumference and TBF in girls but not boys. Our findings are in accordance with previous observations that suggest that there are sex-specific responses in fetal brain growth and adiposity, and raise the possibility of the maternal night-fasting interval as an underlying influence. This trial was registered at clinicaltrials.gov as NCT01174875.
Keywords: birth outcomes, food timing, head circumference, obesity, pregnancy diet
Introduction
Over time, living things have developed intrinsic timing systems of circadian rhythmicity that orchestrate multiple cellular functions necessary for life (1). Under the influence of light-dark cycles, these rhythms have evolved to accommodate the body’s dependence on alternate periods of feeding and fasting (2). Indeed, synchrony between circadian rhythms and feeding-fasting signals ensure that anabolic and catabolic pathways are coordinated with activity-rest cycles (3). Frequent eating and lack of a defined fasting period may disrupt the normal counterregulatory metabolic process and diminish repair activity that occurs during fasting, thus leading to a risk of metabolic diseases (3).
In human studies, the literature has mostly focused on shift work, showing that maternal shift work (representing a pattern of short nighttime fasting) during pregnancy increased the risk of having low-birth-weight offspring, although findings have been inconsistent (4–6). It is unclear if these results may also be applicable to pregnant women with late-night eating behavior (7). Given the emerging evidence that prenatal circadian rhythms can influence fetal outcomes (8, 9), we hypothesized that daily fasting periods in pregnancy may influence fetal growth and adiposity.
There is strong evidence from animal studies demonstrating that nutritional manipulation of feeding-fasting periods during pregnancy can program growth and adipose tissue development of offspring in a sex-specific manner (10). In nocturnal rodents, pregnant dams fed during the light phase delivered male offspring with lower body weights at birth, followed by greater weight gain from 4 to 11 wk of age, than that of pregnant dams fed ad libitum (11). To date, no studies to our knowledge have considered the specific influence of nighttime fasting on fetal growth and adiposity development. Little is known if the duration of nighttime fasting during pregnancy can lead to differential responses in fetal development, and in sex-specific differences in particular.
Nighttime fasting during pregnancy may be a potential dietary approach to improve neonatal outcomes and later child health. We leveraged on Singapore’s location, where sunlight exposure occurs at a fairly constant length of 12 h/d throughout the year, as an ideal environment to conduct circadian-related studies among pregnant women in the general population. In this study, we examined the associations of the maternal night-fasting interval in the late second trimester of pregnancy on birth size and adiposity among infant girls and boys.
Methods
Study design and participants.
Data were drawn from the GUSTO (Growing Up in Singapore Towards healthy Outcomes) study (clinicaltrials.gov identifier NCT01174875). Details of the GUSTO study have been provided elsewhere (12). This study was conducted according to the guidelines in the Declaration of Helsinki. Ethical approval was obtained from the Domain Specific Review Board of the Singapore National Healthcare Group (reference D/09/021) and the SingHealth Centralised Institutional Review Board (reference 2009/280/D).
Pregnant women who received antenatal care (<14 wk of gestation) from June 2009 to September 2010 in the 2 major public maternity units in Singapore (KK Women’s and Children’s Hospital and National University Hospital) were recruited into the GUSTO study. These pregnant women were ≥18 y old, were citizens or permanent residents of Singapore, and had homogeneous parental ethnic groups (Chinese, Malay, or Indian). Informed written consent was obtained from all women.
Of 1161 pregnant women who attended clinic visits at 26–28 wk of gestation, 645 women agreed to participate in detailed dietary assessment with the use of 3-d food diaries. A total of 392 women completed 3-d food diaries, whereas 242 and 11 women only completed 2- and 1-d food diaries, respectively. To provide accurate quantitative information on habitual dietary intake of women, we only included those with completed 3-d food diaries in this study. Of these women, 4 dropped out of the study, and 4 delivered at <34 wk of gestation. The sample size in this study was comprised of 384 mother-offspring pairs (Figure 1).
FIGURE 1.
Flow chart of study inclusion.
Data collection.
At 26–28 wk of gestation, trained personnel conducted detailed interviews and measurements at a clinic visit. They collected data on maternal sociodemographics, obstetric history, medical history, physical activity, bedtime, and sleep duration. A structured questionnaire was used to derive the amount of physical activity during pregnancy. The frequency (days per week) and duration (minutes) of physical activity were used to compute the total score of physical activity in metabolic equivalent task (minutes per week) units. Women were classified as not highly active and highly active according to the cutoffs of <3000 and ≥3000 metabolic equivalent task minutes per week, respectively (13, 14). Bedtime and sleep duration at night were examined with the use of the Pittsburgh Sleep Quality Index questionnaire (15).
Anthropometry.
Maternal height was measured to the nearest 0.1 cm with the use of a portable stadiometer (SECA 213) at 26–28 wk of gestation. Self-reported prepregnancy weight and measured weight at the first antenatal visit (≤14 wk of gestation) were collected. BMI was calculated as weight divided by squared height (kg/m2). Because maternal BMI at the first antenatal visit was strongly correlated with prepregnancy BMI (r = 0.97, P < 0.001) and had a low percentage of missing data (n = 21, 0.1%), it was used for analyses in this study. Serial measurements of maternal weight throughout pregnancy were collected from the hospital case notes. A linear mixed-effects model with the best linear unbiased predictor was used to estimate the linear trajectory of gestational weight gain (GWG) per week between 15 to 35 wk of gestation for each individual (16).
Dietary assessment.
Dietary intake was evaluated with the use of a 3-d food diary at 26–28 wk of gestation. Women were asked to record the time, type, and amount of food and beverage consumed for 2 consecutive weekdays and 1 weekend day at home. Food pictures attached in the front page of the food diaries assisted women in quantifying portion sizes of various foods and beverages. Total daily energy intake was assessed with the use of nutrient analysis software (Dietplan; Forestfield Software) based on a food composition database of locally available foods (17). For mixed dishes not found in the local database, nutrient analyses of recipes were conducted with the use of the nutrient software. For other food items not found in the database, nutrient information was obtained from food labels or the USDA national nutrient database (18).
Daily eating episodes were defined as the number of timestamps associated with food or beverage consumption that provided ≥210 kJ (∼50 kcal) with time intervals between eating episodes of ≥15 min (19). The nighttime period was defined according to the local time from sunset to sunrise (1900–0659), as in our previous studies (20, 21). The percentage of nighttime energy intake was calculated as the amount of energy intake (kilocalories) from 1900 to 0659 divided by the total energy intake (kilocalories) per day multiplied by 100. The night-fasting interval was determined based on the longest fasting interval between calorie-containing food or beverage taken from 1900 to 0659. The total daily energy intake, daily eating episodes, percentage of nighttime energy intake, and night-fasting interval from the 3-d food diaries were averaged (n = 384).
Gestational diabetes mellitus diagnosis.
At 26–28 wk of gestation, women underwent a 75-g oral-glucose-tolerance test to diagnose gestational diabetes mellitus (GDM). Plasma glucose concentrations at 0 (fasting glucose) and at 120 min (2-h glucose) following the oral glucose load were measured by colorimetry [Advia 2400 Chemistry System (Siemens Medical Solutions Diagnostics) and Beckman LX20 Pro Chemistry Analyzer (Beckman Coulter)]. GDM was diagnosed according to the 1999 WHO criteria: ≥7.0 mmol/L for fasting plasma glucose concentration and ≥7.8 mmol/L for 2-h postglucose concentration (22). Women with GDM subsequently received either diet or insulin treatment according to standard management protocols in both hospitals.
Neonatal anthropometric measurements.
Neonatal weight, length, and head circumference were measured by trained clinical staff within 24 h of birth with the use of standardized techniques (23). Weight was recorded to the nearest 0.001 kg with the use of a calibrated infant weighing scale (SECA 334). Recumbent crown-heel length was measured to the nearest 0.1 cm with the use of a mobile measuring mat (SECA 210). The maximum head circumference was measured to the nearest 0.1 cm across the frontal bones of the skull and over the occipital prominence at the back of the head with the use of nonelastic measuring tape (SECA 212). The gestational age (GA; weeks) of each infant at birth was determined based on the ultrasound scan at 7–11 wk. Based on local, customized GA-specific size-at-birth percentile charts (24), neonatal weight, length, and head circumference were converted to weight-for-GA, length-for-GA, and head circumference–for-GA z scores, respectively. Neonatal total body fat (TBF) mass (kilograms) was determined with the use of a validated prediction equation [−0.022 + 0.307 × weight (kg) − 0.077 × sex (boy = 1; girl = 0) − 0.019 × GA (wk) + 0.028 × subscapular skinfold (mm)], which was derived from GUSTO infants (25) and expressed in percentage of TBF (%), where the body fat mass was divided by birth weight and multiplied by 100.
Statistical analysis.
Categorical data were presented as frequencies and percentages, whereas continuous data were presented as means ± SDs. Based on a previous review of sex-specific associations between maternal nutritional manipulation during pregnancy and fetal growth (10), data for infant girls and boys were a priori analyzed separately to examine potential sex-specific differences. Chi-square tests for categorical variables and independent samples Student’s t tests for continuous variables were used to compare the 2 groups. Multivariable general linear models were performed to examine the associations of the maternal night-fasting interval with birth size and adiposity (birth weight–for-GA, birth length–for-GA, birth head circumference–for-GA, birth TBF). Data on the night-fasting interval were normally distributed and analyzed as a continuous variable in the model. Potential confounders were selected a priori based on a literature review (2, 21, 26, 27) and included maternal age, ethnicity, education, parity, BMI, employment status, physical activity, bedtime, total energy intake, eating episodes, and percentage of nighttime energy intake (1900–0659). In an exploratory analysis, night-shift status and sleep duration were not significantly associated with outcomes, and no significant changes in the effect sizes of outcomes were found with night-shift status and sleep duration, and therefore these variables were not included in our model.
To account for the potential influence of diet composition for food consumed at night, the regression models were additionally adjusted for proportions of energy intake from macronutrients during nighttime. Given that the percentage of energy from carbohydrate consumed during nighttime plus the percentage of energy from protein consumed during nighttime plus the percentage of energy from fat consumed during nighttime equals the percentage of nighttime calories, inclusion of any 3 of the 4 variables in the regression models would give identical results because of perfect collinearity. We included the percentages of energy from carbohydrate, protein, and fat consumed during nighttime in the model simultaneously.
We performed 3 additional analyses by adding GWG per week, glucose management (diet compared with insulin treatment), and maternal fasting plasma glucose concentration separately into the main adjusted model. We adjusted for glucose management as women with hyperglycemia received treatment, and for fasting plasma glucose concentration based on our previous finding that the maternal night-fasting interval was associated with the fasting plasma glucose concentration (21). GWG per week, glucose management, and fasting plasma glucose concentration were not included in our main adjusted model because they may be in the causal pathway between the maternal night-fasting interval (measured at midgestation) and birth size or adiposity, which could result in collider bias, but the additional analyses were conducted to examine for a mediating effect. To further examine dose-response, the maternal nighttime fasting interval was modeled in tertiles and included in multivariable general linear models. Sensitivity analysis was performed by excluding women who engaged in night-shift work. Results were presented as β coefficients and 95% CIs. A 2-tailed P value <0.05 was considered statistically significant. All statistical analyses were performed with the use of IBM SPSS statistics, version 20.
Results
Characteristics of participants.
We observed no significant differences (P ≥ 0.05) in maternal characteristics between included (n = 384) and excluded women (n = 777) for maternal age, BMI, GWG per week, education, parity, physical activity, night-shift status, glucose management, fasting plasma glucose concentration, sleep duration, and bedtime, with the exception of ethnicity and employment status. Women who were included in this study were more likely to be Chinese (P < 0.001) and employed (P = 0.019) than excluded women.
Table 1 summarizes the characteristics of women and birth sizes of neonates. Among 384 women, the mean ± SD maternal night-fasting interval was 9.9 ± 1.3 h, with a range of 4.8–12.0 h. No significant differences were observed between the characteristics of women who gave birth to girls and those who gave birth to boys, with the exception of BMI. Women with infant girls had a greater BMI during early pregnancy than women with infant boys (P = 0.034). No significant differences were observed between the standardized birth weight–for-GA, birth length–for-GA and birth head circumference–for-GA z scores for girls and boys (P ≥ 0.05). Girls had a higher TBF percentage at birth than boys (P < 0.001).
TABLE 1.
Characteristics of pregnant women and neonates in the GUSTO study1
| Characteristics | Total (N = 384) | Girls (n = 182) | Boys (n = 202) | P2 |
| Pregnant women | ||||
| Age, y | 30.8 ± 4.8 | 30.5 ± 4.9 | 31.0 ± 4.7 | 0.30 |
| BMI, kg/m2 | 23.5 ± 4.7 | 24.1 ± 5.1 | 23.0 ± 4.2 | 0.034 |
| Gestational weight gain, kg/wk | 0.5 ± 0.1 | 0.5 ± 0.1 | 0.5 ± 0.1 | 0.50 |
| Ethnicity | 0.83 | |||
| Chinese | 247 (64.3) | 119 (65.4) | 128 (63.4) | |
| Malay | 92 (24.0) | 41 (22.5) | 51 (25.2) | |
| Indian | 45 (11.7) | 22 (12.1) | 23 (11.4) | |
| Education | 0.75 | |||
| None, primary, or secondary | 237 (62.0) | 114 (63.0) | 123 (61.2) | |
| University | 145 (38.0) | 67 (37.0) | 78 (38.8) | |
| Parity | 0.41 | |||
| Nulliparous | 175 (45.6) | 87 (47.8) | 88 (43.6) | |
| Multiparous | 209 (54.4) | 95 (52.2) | 114 (56.4) | |
| Physical activity | 0.51 | |||
| <3000 MET-min/wk | 310 (81.8) | 149 (83.2) | 161 (80.5) | |
| ≥3000 MET-min/wk | 69 (18.2) | 30 (16.8) | 39 (19.5) | |
| Employment status | 0.91 | |||
| Unemployed | 104 (27.1) | 50 (27.5) | 54 (26.7) | |
| Employed | 280 (72.9) | 132 (72.5) | 148 (73.3) | |
| Night-shift status | 0.43 | |||
| No | 369 (96.1) | 173 (95.1) | 196 (97.0) | |
| Yes | 15 (3.9) | 9 (4.9) | 6 (3.0) | |
| Glucose management | 0.86 | |||
| No | 319 (83.1) | 150 (82.4) | 169 (83.7) | |
| Diet | 56 (14.6) | 27 (14.8) | 29 (14.4) | |
| Insulin | 9 (2.3) | 5 (2.7) | 4 (2.0) | |
| Fasting glucose concentration, mmol/L | 4.3 ± 0.5 | 4.4 ± 0.6 | 4.3 ± 0.4 | 0.06 |
| Sleep duration, h | 7.1 ± 1.4 | 7.0 ± 1.5 | 7.3 ± 1.4 | 0.26 |
| Bedtime | 2308 ± 0131 | 2312 ± 0124 | 2304 ± 0139 | 0.49 |
| Total energy intake, kcal/d | 1877 ± 462 | 1842 ± 485 | 1908 ± 439 | 0.16 |
| Nighttime energy intake,3 % | 30.3 ± 12.9 | 29.6 ± 12.2 | 30.6 ± 13.4 | 0.47 |
| Nighttime carbohydrate intake,4 % | 14.8 (6.5) | 14.6 (6.4) | 15.0 (6.7) | 0.51 |
| Nighttime protein intake,4 % | 5.1 (2.5) | 5.1 (2.4) | 5.1 (2.6) | 0.83 |
| Nighttime fat intake,4 % | 10.4 (5.6) | 10.3 (5.3) | 10.4 (5.8) | 0.90 |
| Eating episodes, n/d | 4.6 ± 1.2 | 4.5 ± 1.3 | 4.7 ± 1.2 | 0.35 |
| Night fasting, h | 9.9 ± 1.3 | 10.0 ± 1.4 | 9.9 ± 1.2 | 0.55 |
| Neonates | ||||
| GA, wk | 38.5 ± 1.2 | 38.5 ± 1.2 | 38.5 ± 1.1 | 0.86 |
| Birth weight, g | 3133.9 ± 425.8 | 3065.8 ± 404.7 | 3195.2 ± 436.0 | 0.003 |
| Birth length, cm | 49.1 ± 2.2 | 48.6 ± 2.2 | 49.4 ± 2.2 | <0.001 |
| Birth head circumference, cm | 33.5 ± 1.3 | 33.3 ± 1.3 | 33.8 ± 1.3 | <0.001 |
| Birth weight–for-GA, z score | 0.1 ± 1.0 | 0.1 ± 1.1 | 0.1 ± 1.0 | 0.82 |
| Birth length–for-GA, z score | 0.2 ± 1.1 | 0.2 ± 1.1 | 0.2 ± 1.0 | 0.98 |
| Birth head circumference–for-GA, z score | 0.1 ± 0.9 | 0.1 ± 0.9 | 0.1 ± 0.9 | 0.56 |
| Birth TBF, % | 9.7 ± 3.1 | 10.8 ± 3.0 | 8.7 ± 2.8 | <0.001 |
Values are means ± SDs or n (%). GA, gestational age; GUSTO, Growing Up in Singapore Towards healthy Outcomes; MET-min, metabolic equivalent task-minutes; TBF, total body fat.
Chi-square tests for categorical variables and independent samples Student’s t tests for continuous variables were used to compare the 2 groups; P < 0.05 was considered statistically significant.
Computed from nighttime energy intake (kilocalories)/total energy intake per day (kilocalories) × 100.
Computed from nighttime carbohydrate intake (grams) × 4/total energy intake per day (kilocalories) × 100; nighttime protein intake (grams) × 4/total energy intake per day (kilocalories) × 100; nighttime fat intake (grams) × 9/total energy intake per day (kilocalories) × 100.
Associations of maternal night-fasting interval with birth sizes and adiposity.
Table 2 shows the associations of maternal night-fasting intervals in the late second trimester of pregnancy with standardized birth sizes and with adiposity for girls and boys. In girls, each hour increase in the maternal night-fasting interval was associated with a 0.22-SD (95% CI: 0.05-, 0.40-SD; P = 0.013) increase in birth head circumference–for-GA and a 0.84% (95% CI: 0.19%, 1.49%; P = 0.012) increase in TBF at birth, after adjustment for confounders. These associations remained similar when additional adjustments were made for diet composition for food consumed during nighttime, GWG per week, and glucose management. After adjusting for plasma fasting glucose concentration, the associations of the maternal night-fasting interval with birth head circumference–for-GA and TBF were attenuated and that for birth head circumference–for-GA became nonsignificant. The effect sizes of birth head circumference–for-GA and TBF were reduced by 32% and 15%, respectively, indicating that the plasma fasting glucose concentration partially attenuated the associations. In boys, no significant associations were found with birth size and TBF (P ≥ 0.05).
TABLE 2.
Associations of maternal night-fasting interval during pregnancy with birth sizes and adiposity in the GUSTO study1
|
z Score |
||||||||
| Birth weight–for-GA |
Birth length–for-GA |
Birth head circumference–for-GA |
Birth TBF, % |
|||||
| β (95% CI) | P | β (95% CI) | P | β (95% CI) | P | β (95% CI) | P | |
| Girls (n = 182) | ||||||||
| Night fasting, h | ||||||||
| Model 12 | 0.06 (−0.06, 0.17) | 0.33 | 0.07 (−0.05, 0.19) | 0.25 | 0.14 (0.05, 0.24) | 0.003 | 0.18 (−0.15, 0.51) | 0.28 |
| Model 23 | 0.15 (−0.06, 0.36) | 0.17 | 0.10 (−0.13, 0.32) | 0.39 | 0.22 (0.05, 0.40) | 0.013 | 0.84 (0.19, 1.49) | 0.012 |
| Model 34 | 0.15 (−0.06, 0.36) | 0.16 | 0.09 (−0.13, 0.32) | 0.41 | 0.21 (0.04, 0.38) | 0.015 | 0.84 (0.19, 1.49) | 0.012 |
| Model 45 | 0.15 (−0.06, 0.37) | 0.17 | 0.09 (−0.15, 0.32) | 0.47 | 0.23 (0.05, 0.42) | 0.014 | 0.80 (0.14, 1.45) | 0.018 |
| Model 56 | 0.16 (−0.05, 0.37) | 0.13 | 0.11 (−0.12, 0.33) | 0.35 | 0.22 (0.05, 0.40) | 0.013 | 0.86 (0.21, 1.51) | 0.011 |
| Model 67 | 0.12 (−0.09, 0.33) | 0.24 | −0.01 (−0.23, 0.22) | 0.98 | 0.15 (−0.02, 0.33) | 0.09 | 0.73 (0.04, 1.41) | 0.037 |
| Boys (n = 202) | ||||||||
| Night fasting, h | ||||||||
| Model 12 | 0.02 (−0.10, 0.13) | 0.78 | −0.05 (−0.17, 0.07) | 0.40 | 0.06 (−0.05, 0.16) | 0.31 | 0.10 (−0.24, 0.43) | 0.58 |
| Model 23 | −0.10 (−0.33, 0.12) | 0.37 | −0.11 (−0.32, 0.11) | 0.32 | −0.13 (−0.33, 0.07) | 0.20 | −0.16 (−0.74, 0.42) | 0.58 |
| Model 34 | −0.06 (−0.30, 0.17) | 0.59 | −0.10 (−0.32, 0.12) | 0.36 | −0.09 (−0.29, 0.11) | 0.36 | −0.04 (−0.63, 0.54) | 0.89 |
| Model 45 | −0.13 (−0.39, 0.13) | 0.33 | −0.20 (−0.44, 0.04) | 0.11 | −0.10 (−0.32, 0.12) | 0.35 | −0.19 (−0.87, 0.49) | 0.58 |
| Model 56 | −0.03 (−0.23, 0.17) | 0.75 | −0.05 (−0.25, 0.15) | 0.62 | −0.13 (−0.34, 0.07) | 0.20 | −0.14 (−0.71, 0.44) | 0.64 |
| Model 67 | −0.03 (−0.26, 0.20) | 0.78 | −0.11 (−0.33, 0.11) | 0.31 | −0.12 (−0.33, 0.09) | 0.27 | 0.05 (−0.55, 0.64) | 0.88 |
Data were analyzed by using multivariate general linear models; P < 0.05 was considered statistically significant. GA, gestational age; GUSTO, Growing Up in Singapore Towards healthy Outcomes; TBF, total body fat.
Model 1: crude.
Model 2: adjusted for maternal age, ethnicity, education, parity, BMI, employment status, physical activity, bedtime, eating frequency, total calories, and percentage of nighttime calories (1900–0659).
Model 3: adjusted for maternal age, ethnicity, education, parity, BMI, employment status, physical activity, bedtime, eating frequency, total calories, percentage of energy from protein consumed during nighttime, percentage of energy from fat consumed during nighttime, and percentage of energy from carbohydrates consumed during nighttime.
Model 4: adjusted for model 2 + gestational weight gain per week.
Model 5: adjusted for model 2 + glucose management (diet or insulin treatment).
Model 6: adjusted for model 2 + fasting plasma glucose concentration.
In girls, compared with the lowest tertile of the maternal nighttime fasting interval (mean: 8.3 h; range: 4.8–9.3 h), the highest tertile of the maternal nighttime fasting interval (mean: 11.4 h; range: 10.8–12.0 h) was associated with higher birth head circumference–for-GA z score (β: 0.52 SD; 95% CI: −0.01, 1.05 SD; P = 0.05) (Figure 2A) and higher percentage of TBF at birth (β: 2.00%; 95% CI: −0.05%, 4.04%; P = 0.06) (Figure 2B). In a sensitivity analysis with stratification by sex, results remained similar after excluding women who engaged in night shifts (Supplemental Table 1).
FIGURE 2.
Associations of the maternal nighttime fasting interval in tertiles with (A) birth head circumference–for-GA and (B) neonatal percentage total body fat in girls. Error bars represent upper bounds of the 95% CIs. Tertile fasting interval range (minimum–maximum): tertile 1 = 4.8–9.3 h (n = 59); tertile 2 = 9.3–10.8 h (n = 60); and tertile 3 = 10.8–12.0 h (n = 63). Tertile median fasting interval: tertile 1 = 8.7 h; tertile 2 = 10.0 h; and tertile 3 = 11.3 h. GA, gestational age.
Discussion
In this Asian mother-offspring cohort, infant girls born to mothers with a longer late second trimester night-fasting interval during pregnancy had larger head circumferences and greater TBF percentages at birth, after accounting for various demographic and lifestyle factors. These associations remained after additional adjustment for diet composition of food consumed at night, GWG, and GDM management status, but were attenuated somewhat after adjustment for maternal fasting plasma glucose concentration. In boys, no associations were observed with birth size and TBF percentage. These findings suggest that the nighttime maternal fasting interval can influence fetal brain growth and adiposity in a sex-specific manner.
The negative impact of chronodisruption during pregnancy on fetal growth has been demonstrated in epidemiological studies, where shift-work exposure during pregnancy was associated with low-birth-weight delivery (4, 5), although modest or null findings were observed in other studies (6, 28). We found no significant association between the maternal night-fasting interval and birth weight in this study, however, the effect size was in a positive direction. To our knowledge, relatively few studies involving circadian variables have considered infant head circumference and adiposity as clinical outcomes. We are aware of only one relevant study relating to these outcomes, which involved mothers from the Southampton Women’s Survey in the United Kingdom, where night-shift work during pregnancy was not found to be associated with being small for GA and head circumference (28). The former finding is in line with our results, whereas the latter finding contradicts our observations. To our knowledge, no outcome data are available from literature pertaining to neonatal TBF.
The biological mechanisms related to our findings are currently unclear. The attenuation effect of fasting plasma glucose concentration on fetal growth may suggest the role of maternoplacental glucose transfer to meet fetal nutrient demand and development (29). In showing that the association between the maternal night-fasting interval and fetal growth was partially influenced by fasting plasma glucose concentration, it is suggested that this association can be also effected by other mechanisms. In animal models, chronodisruption during pregnancy can reduce placental melatonin secretion, which might in turn elevate oxidative stress amounts (9) and compromise placental nutrient transfer, thereby restricting fetal growth and altering fat deposition (10, 30). Alternatively, our findings may be interpreted with the use of the predictive adaptive response model (31). A longer night-fasting interval may create a short-term starvation status that restricts nutrient availability to the fetus. Over time, the most adaptive response to such an environment would be to store energy and slow down metabolism, which would lead to increased adipose tissue accumulation by the fetus (32). To protect brain growth, peripheral insulin sensitivity is downregulated to get more glucose into the brain (32). Such prioritization of nutrient allocation for fetal fat deposition and brain growth could be determined by fetal liver blood flow distribution (29). Otherwise, it is possible that variations in the length of the night-fasting interval may result in alterations in the proportions and rates of transfer of different nutrients to the fetus, contributing to different fetal growth patterns.
Sexually dimorphic responses to in utero programming of metabolic function have been reported in mothers with unbalanced nutrition or maternal stress (10, 33). Most human and animal studies reported that males rather than females were more developmentally vulnerable to prenatal insults (11, 29), although this phenomenon could be context and stage specific (33). Our findings show that infant girls are more susceptible to the effects of the maternal night-fasting interval, which is in line with a study that used a rat model and showed that female offspring exposed to chronic phase shift in utero had increased adiposity and poor glucose tolerance by 12 mo of age (34). The basis underlying these sex-specific programming effects remains elusive (10), but could reflect interactions between the maternal circadian timing system, nutritional mechanisms, and factors associated with fetal development that are sexually dimorphic.
The present study has the advantage of a prospective design, with exposure assessed at a prenatal stage and most outcomes ascertained objectively. An important strength of the study is the use of 3-d food diaries to assess usual dietary intake and the night-fasting interval, which helped to minimize recall bias. A wide array of sociodemographic, health, and lifestyle factors that could confound associations was taken into account and controlled for in the analyses. Given the paucity of existing data on prenatal fasting patterns and neonatal outcomes, our study provides useful, informative data on this relation.
However, we also recognized and considered the following limitations. First, neonatal TBF was estimated by using a formula-derived equation; the formula, however, has been validated by infant body composition measurements with the use air displacement plethysmography (25). Second, our study lacked measures of plasma insulin, glucocorticoids, and oxidative stress amounts, which would have been useful to mechanistically explain the associations of the maternal night-fasting interval with fetal growth and adiposity. Third, dietary assessment was only collected at one time point in the late second trimester, which restricted our ability to evaluate the trimester-specific effect of the maternal night-fasting interval on fetal growth. Similarly, maternal glucose concentrations were measured at one time point during pregnancy, so we were not able to assess the effect of glucose metabolism at different stages of pregnancy. Fourth, the study was conducted in a Southeast Asian setting with a Singaporean mother-offspring cohort and, hence, generalization to other populations should be made with caution. Lastly, despite our efforts to control for multiple covariates, the possibility of residual confounding by unmeasured or poorly measured covariates remains.
In conclusion, sex-dependent fetal growth and adiposity development in response to a fasting period were observed, such that infant girls born to mothers with a longer night-fasting interval in the late second trimester of pregnancy were found to have increased head circumferences and TBF percentages at birth. Monitoring maternal nighttime fasting might be a new approach to potentially modify neonatal outcomes and future health. However, more studies are needed to replicate these findings and to understand their biological mechanisms. Also, randomized controlled trials and long-term studies are warranted to confirm whether a lengthened maternal night-fasting interval is beneficial or harmful to the offspring. Because GUSTO is an ongoing prospective study, it will allow us to assess the longer-term consequences on child growth, development, and health. Our finding thus supports an important line of research into the developmental programming effects of chrononutrition.
Acknowledgments
We thank the GUSTO (Growing Up in Singapore Towards healthy Outcomes) study group, which includes Allan Sheppard, Amutha Chinnadurai, Anne Eng Neo Goh, Anne Rifkin-Graboi, Anqi Qiu, Arijit Biswas, Bee Wah Lee, Birit FP Broekman, Boon Long Quah, Borys Shuter, Chai Kiat Chng, Cheryl Ngo, Choon Looi Bong, Christiani Jeyakumar Henry, Cornelia Yin Ing Chee, Yam Thiam Daniel Goh, Doris Fok, George Seow Heong Yeo, Helen Chen, Hugo PS van Bever, Iliana Magiati, Inez Bik Yun Wong, Ivy Yee-Man Lau, Jeevesh Kapur, Jenny L Richmond, Joanna D Holbrook, Joshua J Gooley, Kenneth Kwek, Krishnamoorthy Niduvaje, Leher Singh, Lin Su, Lourdes Mary Daniel, Marielle V Fortier, Mark Hanson, Mary Rauff, Mei Chien Chua, Michael Meaney, Neerja Karnani, Oon Hoe Teoh, PC Wong, Pratibha Agarwal, Rob M van Dam, Salome A Rebello, Seang-Mei Saw, Shang Chee Chong, Shirong Cai, Shu-E Soh, Sok Bee Lim, Chin-Ying Stephen Hsu, Walter Stunkel, Wee Meng Han, Wei Pang and Yiong Huak Chan. The authors’ responsibilities were as follows—KMG, PDG, KHT, LP-CS, and Y-SC: designed the GUSTO cohort study; SLL and FY: designed the present study; MTC and PHW: contributed to the dietary data collection and analyses; SLL, PHW, M-TT, IMA, PN, and FM-R: performed the data management and analysis; YBC: advised on the statistical analysis; SLL, PHW, M-TT, JKYC, YBC, NL, VSR, YSL, MF-FC, and FY: interpreted the findings and revised drafts of the paper; SLL: wrote the paper; and all authors: read and approved the final manuscript.
Footnotes
Abbreviations used: GA, gestational age; GDM, gestational diabetes mellitus; GUSTO, Growing Up in Singapore Towards healthy Outcomes; GWG, gestational weight gain.
References
- 1.Oike H, Oishi K, Kobori M. Nutrients, clock genes, and chrononutrition. Curr Nutr Rep 2014;3:204–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Stenvers DJ, Jonkers CF, Fliers E, Bisschop PH, Kalsbeek A. Nutrition and the circadian timing system. Prog Brain Res 2012;199:359–76. [DOI] [PubMed] [Google Scholar]
- 3.Longo VD, Panda S. Fasting, circadian rhythms, and time-restricted feeding in healthy lifespan. Cell Metab 2016;23:1048–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Abeysena C, Jayawardana P, De A, Seneviratne R. Maternal sleep deprivation is a risk factor for small for gestational age: a cohort study. Aust N Z J Obstet Gynaecol 2009;49:382–7. [DOI] [PubMed] [Google Scholar]
- 5.Lin YC, Chen MH, Hsieh CJ, Chen PC. Effect of rotating shift work on childbearing and birth weight: a study of women working in a semiconductor manufacturing factory. World J Pediatr 2011;7:129–35. [DOI] [PubMed] [Google Scholar]
- 6.Palmer KT, Bonzini M, Harris EC, Linaker C, Bonde JP. Work activities and risk of prematurity, low birth weight and pre-eclampsia: an updated review with meta-analysis. Occup Environ Med 2013;70:213–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Park HJ, Lee JJ, Kim JM, Lee HA, Kim SH, Kim Y. A study of snack consumption, night-eating habits, and nutrient intake in gestational diabetes mellitus. Clin Nutr Res 2013;2:42–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Watanabe S, Akiyama S, Hanita T, Li H, Nakagawa M, Kaneshi Y, Ohta H; Japan RED Filter Study Group. Designing artificial environments for preterm infants based on circadian studies on pregnant uterus. Front Endocrinol (Lausanne) 2013;4:113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Valenzuela FJ, Vera J, Venegas C, Pino F, Lagunas C. Circadian system and melatonin hormone: risk factors for complications during pregnancy. Obstet Gynecol Int 2015;2015:825802. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lecoutre S, Breton C. Maternal nutritional manipulations program adipose tissue dysfunction in offspring. Front Physiol 2015;6:158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.de Almeida Faria J, de Araujo TM, Mancuso RI, Meulman J, da Silva Ferreira D, Batista TM, Vettorazzi JF, da Silva PM, Rodrigues SC, Kinote A, et al. Day-restricted feeding during pregnancy and lactation programs glucose intolerance and impaired insulin secretion in male rat offspring. Acta Physiol (Oxf) 2016;217:240–53. [DOI] [PubMed] [Google Scholar]
- 12.Soh SE, Tint MT, Gluckman PD, Godfrey KM, Rifkin-Graboi A, Chan YH, Stünkel W, Holbrook JD, Kwek K, Chong YS, et al. Cohort profile: Growing Up in Singapore Towards healthy Outcomes (GUSTO) birth cohort study. Int J Epidemiol 2014;43:1401–9. [DOI] [PubMed] [Google Scholar]
- 13.IPAQ Research Committee. Guidelines for data processing and analysis of the International Physical Activity Questionnaire (IPAQ) [Internet]. c2005. [cited 2017 Jan 15]. Available from: http://www.institutferran.org/documentos/scoring_short_ipaq_april04.pdf.
- 14.Padmapriya N, Shen L, Soh SE, Shen Z, Kwek K, Godfrey KM, Gluckman PD, Chong YS, Saw SM, Müller-Riemenschneider F. Physical activity and sedentary behavior patterns before and during pregnancy in a multi-ethnic sample of Asian women in Singapore. Matern Child Health J 2015;19:2523–35. [DOI] [PubMed] [Google Scholar]
- 15.Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res 1989;28:193–213. [DOI] [PubMed] [Google Scholar]
- 16.Cheung YB. Statistical analysis of human growth and development. Boca Raton (FL): CRC Press; 2014. [Google Scholar]
- 17.Health Promotion Board Singapore. Singapore: healthy living and disease prevention information [Internet]. c2012 [cited 2015 Apr 20]. Available from: http://www.hpb.gov.sg/HOPPortal/.
- 18.USDA Agricultural Research Servce. USDA Food Composition Databases. USDA National Nutrient Database for Standard Reference [Internet]. c2011. [cited 2016 Jan 30]. Available from: http://ndb.nal.usda.gov/.
- 19.Gibney MJ, Wolever TM. Periodicity of eating and human health: present perspective and future directions. Br J Nutr 1997;77:S3–5. [DOI] [PubMed] [Google Scholar]
- 20.Cheng TS, Loy SL, Toh JY, Cheung YB, Chan JK, Godfrey KM, Gluckman PD, Saw SM, Chong YS, Lee YS, et al. Predominantly nighttime feeding and weight outcomes in infants. Am J Clin Nutr 2016;104:380–8. [DOI] [PubMed] [Google Scholar]
- 21.Loy SL, Chan JK, Wee PH, Colega MT, Cheung YB, Godfrey KM, Godfrey KM, Kwek K, Saw SM, Chong YS, et al. Maternal circadian eating time and frequency are associated with blood glucose concentrations during pregnancy. J Nutr 2017;147:70–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 1998;15:539–53. [DOI] [PubMed] [Google Scholar]
- 23.WHO. WHO Child Growth Standards: training course on child growth assessment - interpreting growth indicators. Geneva (Switzerland): WHO Press; 2008. [Google Scholar]
- 24.Aris IM, Gandhi M, Cheung YB, Soh SE, Tint MT, Gluckman PD, Lee YS, Yap FK, Chong YS. A new population-based reference for gestational age-specific size-at-birth of Singapore infants. Ann Acad Med Singapore 2014;43:439–47. [PubMed] [Google Scholar]
- 25.Aris IM, Soh SE, Tint MT, Liang S, Chinnadurai A, Saw SM, Kwek K, Godfrey KM, Gluckman PD, Chong YS, et al. Body fat in Singaporean infants: development of body fat prediction equations in Asian newborns. Eur J Clin Nutr 2013;67:922–7. [DOI] [PubMed] [Google Scholar]
- 26.Marinac CR, Natarajan L, Sears DD, Gallo LC, Hartman SJ, Arredondo E, Patterson RE. Prolonged nightly fasting and breast cancer risk: findings from NHANES (2009–2010). Cancer Epidemiol Biomarkers Prev 2015;24:783–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Patterson RE, Laughlin GA, LaCroix AZ, Hartman SJ, Natarajan L, Senger CM, Martinez ME, Villasenor A, Sears DD, Marinac CR, et al. Intermittent fasting and human metabolic health. J Acad Nutr Diet 2015;115:1203–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bonzini M, Coggon D, Godfrey K, Inskip H, Crozier S, Palmer KT. Occupational physical activities, working hours and outcome of pregnancy: findings from the Southampton Women’s Survey. Occup Environ Med 2009;66:685–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Godfrey KM, Haugen G, Kiserud T, Inskip HM, Cooper C, Harvey NC, Crozier SR, Robinson SM, Davies L; Southampton Women’s Survey Study Group, et al. Fetal liver blood flow distribution: role in human developmental strategy to prioritize fat deposition versus brain development. PLoS One 2012;7:e41759. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Muhlhausler B, Smith SR. Early-life origins of metabolic dysfunction: role of the adipocyte. Trends Endocrinol Metab 2009;20:51–7. [DOI] [PubMed] [Google Scholar]
- 31.Gluckman PD, Hanson MA. Living with the past: evolution, development, and patterns of disease. Science 2004;305:1733–6. [DOI] [PubMed] [Google Scholar]
- 32.Entringer S. Exposure to prenatal psychosocial stress: implications for long-term disease susceptibility. Göttingen (Germany): Cuvillier Verlag; 2007. [Google Scholar]
- 33.Tarrade A, Panchenko P, Junien C, Gabory A. Placental contribution to nutritional programming of health and diseases: epigenetics and sexual dimorphism. J Exp Biol 2015;218:50–8. [DOI] [PubMed] [Google Scholar]
- 34.Varcoe TJ, Wight N, Voultsios A, Salkeld MD, Kennaway DJ. Chronic phase shifts of the photoperiod throughout pregnancy programs glucose intolerance and insulin resistance in the rat. PLoS One 2011;6:e18504. [DOI] [PMC free article] [PubMed] [Google Scholar]