- Email: [email protected]
Shorter sleep duration in early pregnancy is associated with birth length: a prospective cohort study in Wuhan, China
Accepted Manuscript Shorter sleep duration in early pregnancy is associated with birth length: a prospective cohort study in Wuhan, China Weiye Wang, Chunrong Zhong, Yu Zhang, Li Huang, Xi Chen, Xuezhen Zhou, Renjuan Chen, Xiating Li, Mei Xiao, Liping Hao, Xuefeng Yang, Nianhong Yang, Sheng Wei PII:
S1389-9457(17)30149-1
DOI:
10.1016/j.sleep.2017.03.013
Reference:
SLEEP 3355
To appear in:
Sleep Medicine
Received Date: 25 November 2016 Revised Date:
20 February 2017
Accepted Date: 8 March 2017
Please cite this article as: Wang W, Zhong C, Zhang Y, Huang L, Chen X, Zhou X, Chen R, Li X, Xiao M, Hao L, Yang X, Yang N, Wei S, Shorter sleep duration in early pregnancy is associated with birth length: a prospective cohort study in Wuhan, China, Sleep Medicine (2017), doi: 10.1016/ j.sleep.2017.03.013. 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.
ACCEPTED MANUSCRIPT Shorter sleep duration in early pregnancy is associated with birth length: a prospective cohort study in Wuhan, China Weiye Wang a, Chunrong Zhong b, Yu Zhang b, Li Huang b, Xi Chen b, Xuezhen Zhou , Renjuan Chen b, Xiating Li b, Mei Xiao c, Liping Hao b, Xuefeng Yang b, Nianhong
RI PT
b
Yangb,*,†, Sheng Wei a,*,† a
Department of Epidemiology and Biostatistics, Ministry of Education Key Lab of
SC
Environment and Health, School of Public Health, Tongji Medical College, Huazhong
b
M AN U
University of Science and Technology, Hubei, P.R. China.
Department of Nutrition and Food Hygiene, Hubei Key Laboratory of Food
Nutrition and Safety, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Hubei, P.R. China.
Department of Obstetrics and Gynaecology, Hubei Maternal and Child Health
TE D
3
Hospital, Wuhan, Hubei, P.R. China.
* Corresponding authors. 13 Hangkong Road, Wuhan, Hubei, 430030, China. Tel.:
EP
+86 27 83650521 (N. Yang); +86 27 83692031 (S. Wei).
AC C
E-mail addresses: [email protected] (N. Yang), [email protected] (S. Wei).
†
These authors contributed equally to this work.
ACCEPTED MANUSCRIPT ABSTRACT Objectives: To examine the association between sleep duration in early pregnancy and fetal growth in a prospective cohort study of 3567 Chinese women.
RI PT
Methods: Pregnant women at 8–16 weeks of gestation were interviewed using a semi-quantitative questionnaire to assess sleep duration. Birth weight and birth length were measured by a midwife in the delivery room at birth; low birth weight (LBW)
SC
was defined as birth weight <2500 g and small for gestational age (SGA) was defined
M AN U
as <10th customized centile.
Results: The average age of participants was 28.21±3.38 years old. The mean sleep duration was 8.39±1.13 h/day. A total of 1290 women sleeping ≥9 h/day, 1563 sleeping 8 to <9 h/day, 550 sleeping 7 to <8 h/day, and 164 sleeping <7 h/day.
TE D
Compared to the sleeping 8 to <9 h/day group, birth length and birth weight of the sleeping <7 h/day group decreased by 2.42 mm (95% CI: -4.27, -0.58, p=0.010) and 42.70 g (95% CI: -103.02, 17.62, p=0.165), respectively; and risk of LBW and SGA
EP
of the sleeping <7 h/day group increased by 83% (95% CI: 0.59, 5.73, p=0.297) and
AC C
56% (95% CI: 0.84, 2.92, p=0.159), respectively; birth length of the sleeping <7 h/day group was decreased more in male babies, and among mothers without a midday napping habit or with a history of abortion (all p for interaction <0.05). Conclusions: Shorter sleep duration in early pregnancy was associated with birth length. Our findings indicate that midday napping may be a protective factor for birth length among pregnant women with shorter sleep duration.
ACCEPTED MANUSCRIPT Keywords: Sleep duration Midday napping
RI PT
Fetal growth Birth length
AC C
EP
TE D
M AN U
SC
Birth weight
ACCEPTED MANUSCRIPT Introduction Slow fetal growth is an important public health concern worldwide, and newborns with low birth length or weight often suffer negative short-term and long-term
RI PT
consequences during infancy, childhood and adulthood. Epidemiological studies have suggested that lower birth weight or shorter birth length are related to a higher risk of infection-related hospitalization throughout childhood [1]. Various chronic diseases in
SC
adult life, such as hypertension, diabetes and cancer, were also found to be
M AN U
consequences of slow fetal growth [2–4]. Many factors were reported to be related to fetal growth, such as dietary intake and lifestyle [5,6]. Additionally, short sleep duration in pregnant women has recently been recognized as a risk factor for slow fetal growth [7].
TE D
Sleep duration that is too short or too long is considered to be associated with many diseases, for example, diabetes mellitus and cardiovascular diseases, the population in most of the studies are middle-aged and elderly people [8–10]. Pregnant women may
EP
sleep less during pregnancy; Signal et al. found that pregnant women reported a mean
AC C
decrease in sleep duration of 0.5 h, compared to women in the general population [11]. There are some physiological and psychological changes during pregnancy, including rapid growth of weight, increased arousals from sleep, and decreased functional reserve capacity [12]. These changes may make pregnant women suffer from shorter sleep duration in pregnancy [13]. As far as we know, only five studies have examined the association between sleep duration in pregnancy and fetal growth, and the conclusions were inconsistent [7,14–
ACCEPTED MANUSCRIPT 17]. One study observed that women who slept for ≤8 h/day were at higher risk of having small for gestational age (SGA) neonates when compared to those who sleep for >8 h/day [14]. Other studies did not find a significant association between shorter
RI PT
sleep duration and neonate birth weight, low birth weight (LBW), or SGA [7,15–17]. However, all of these studies were focused on birth weight, none had determined the association between sleep duration in pregnancy and birth length. Birth length is
SC
another indicator of slow fetal growth, in addition to birth weight. Indeed, some
M AN U
studies indicated that only short birth length, not low birth weight, is associated with a higher risk of poor health [18,19]. Furthermore, these studies did not consider the potential effect modifiers of sleep duration.
To assess the potential impact of sleep duration of women in early pregnancy on
TE D
fetal growth, we conducted a prospective mother–child cohort study in Wuhan, central China. We further explored the potential effect modifiers of sleep duration in early
Methods
EP
pregnancy on fetal growth.
AC C
Study population
The present study was embedded in the Tongji Maternal and Child Health Cohort
(TMCHC) study. The TMCHC is an ongoing prospective cohort with pregnant women and their children residing in Wuhan, Hubei province, central China. Its primary objective is to determine the effect of environmental pollution, dietary, lifestyle and biochemical factors on the health of pregnant women and their children. Women who were 8–16 weeks pregnant were invited to participate in the TMCHC
ACCEPTED MANUSCRIPT during their first antenatal visit in Hubei Maternal and Child Health Care Hospital, or the Central Hospital of Wuhan, or Jiangan District Maternal and Child Health Care Hospital of Wuhan, beginning in January 2013. The inclusion criteria for TMCHC
RI PT
were as follows: (1) aged >18 years old, (2) residing in Wuhan, (3) no communication problems.
Participants in the present study were all from TMCHC, and the exclusion criteria
SC
were: (1) with stillborn infants and (2) with multiple pregnancy. The study was
M AN U
approved by the Ethics review committee of Tongji Medical College, Huazhong University of Science and Technology (NO. 201302). All study participants had provided written informed consent after a complete description of the study. Sleep duration
TE D
Information on sleep duration was assessed by a semi-quantitative questionnaire, which referenced to question 1 to question 4 of Pittsburg Sleep Quality Index [20]. Women were interviewed face-to-face by trained investigators when they were
EP
enrolled (at weeks 8–16 of gestation). The following questions were adopted: “When
AC C
have you usually gone to bed at night during the past month?” “How long has it usually taken you to fall asleep every night during the past month?” “When have you usually gotten up in the morning during the past month?” “How many hours did you sleep every night during the past month?” The responses of sleep duration were categorized into the following four groups: ≥9 h/day, 8 to <9 h/day, 7 to <8 h/day, and <7 h/day. Neonatal outcomes
ACCEPTED MANUSCRIPT The data on neonatal outcomes were obtained from hospital medical records, including neonatal gender and date of birth. Birth weight and birth length were measured by a midwife in the delivery room at birth. LBW was defined as birth
RI PT
weight <2500 g, and SGA was defined as <10th customized centile [21]. Gestational age at delivery was based on the self-reported last menstrual period (LMP) and delivery date. Fetal crown rump length (CRL), which was measured by ultrasound in
SC
early pregnancy, was also used to calculate gestational age at delivery for women.
M AN U
When women could not accurately remember the LMP, or the difference between LMP-based and CRL-based gestational age was more than 10 days, we chose the latter. For women who were lost to follow-up or whose child’s hospital medical records could not be found, information on neonatal outcomes was obtained by
Covariables
TE D
telephone interview.
A series of covariates were obtained from self-reported questionnaires at enrollment
EP
as follows: maternal age when enrolled (year), self-reported pre-pregnancy weight
AC C
(kg), ethnicity (Han Chinese/others), education level, smoking (active smoking or smoking before pregnancy, defined as ≥3 times a week, yes/no, respectively), alcohol consumption (active drinking or drinking before pregnancy, defined as ≥3 times a week, yes/no, respectively), average personal income (per month, CNY), abortion history (yes/no), parity (primiparous/multiparous) and midday napping (h/day). Education level was recoded according to the number of completed schooling years (≤12/12–15/≥16). Maternal height (cm) was measured at enrollment. Maternal weight
ACCEPTED MANUSCRIPT was measured on admission to hospital while awaiting delivery to calculate gestational weight gain (GWG). Statistical analyses
RI PT
Distributions of demographic characteristics were described using mean ± SDs for continuous variables, and using frequencies and percentages for categorical variables. One-way analysis of variance (ANOVA) or Pearson χ2-test were used to compare
SC
demographic characteristics of pregnant women grouped by sleep duration. Multiple
M AN U
linear regression analyses were applied to examine the association of sleep duration in pregnancy with birth weight or birth length, and the results are presented as coefficients (β) with 95% confidence intervals (CIs). Binomial logistic regression analyses were used to assess the association between sleep duration in pregnancy and
TE D
LBW or SGA, and the results are presented as odds ratios (ORs) with 95% CIs. Covariates that were related to sleep duration and fetal growth reported in the literature were chosen as potential confounders, including maternal age, education
EP
level, alcohol consumption, smoking, ethnicity, height, pre-pregnancy body mass
AC C
index (BMI), gestational weight gain, gestational age at delivery, average personal income, parity, fetal gender, and midday napping. Stratified analysis was used to explore the potential modification effect. The
following factors were considered as potential effect modifiers: pre-pregnancy BMI (<24, ≥24 kg/m2), age (<30/≥30 years), parity (primiparous/multiparous), fetal gender (boy/girl), midday napping (yes/no), and abortion history (yes/no). The Wald test was used to test the potential effect modification by adding interaction terms of these
ACCEPTED MANUSCRIPT covariates with sleep duration in early pregnancy. All statistical analyses were performed with R software (The R Project for Statistical Computing, version 3.3.1), and were conducted assuming a two-sided
RI PT
alternative hypothesis; p<0.05 was considered to be statistically significant. Results
A total of 4539 pregnant women who were eligible for inclusion in this cohort were
SC
recruited. Of those, 492 were lost to follow-up, 196 had stillbirths and 106 had twins,
M AN U
and 178 were excluded because of missing information on covariates, leaving 3567 women for the final analysis. Fig. 1 shows the processes of participants’ inclusion. The average age of the participants was 28.21±3.38 years old. The mean sleep duration in early pregnancy was 8.39±1.13 h/day. Average birth length was
TE D
501.61±13.21 mm, and average birth weight was 3339.04±435.56 g. The proportions of LBW and SGA were 2.2% and 6.4%, respectively. Pregnant women were divided into four groups according to their sleep duration in early pregnancy, with 1290
EP
(36.2%) pregnant women sleeping ≥9 h/day, 1563 (43.8%) sleeping 8 to <9 h/day, 550
AC C
(15.4%) sleeping 7 to <8 h/day, and 164 (4.6%) sleeping <7 h/day. The differences of age, pre-pregnancy weight, pre-pregnancy BMI, educational level, midday-napping and parity among the four groups were significant (all p<0.05). Details of the demographic and clinical characteristics are shown in Table 1. The one-way ANOVA shows that compared to the group sleeping 8 to <9 h/day, birth length of the group sleeping <7 h/day decreased by 1.88 mm, birth weight of the sleeping <7 h/day group decreased by 5.40 g, but the differences were not statistically
ACCEPTED MANUSCRIPT significant (all p>0.05). Compared to the group sleeping 8 to <9 h/day, the proportions of LBW and SGA in the group sleeping <7 h/day were lower, however, the Pearson χ2-test shows that differences were not statistical significant (all p>0.05).
RI PT
Details are shown in Table 2. After adjustment for a series of covariates, compared to the group sleeping 8 to <9 h/day, birth length of the group sleeping <7 h/day decreased by 2.42 mm (95% CI:
SC
-4.27, -0.58, p=0.010); birth weight of the sleeping <7 h/day group decreased by
M AN U
42.70 g (95% CI: -103.02, 17.62, p=0.165), the risk of LBW of the sleeping <7 h/day group increased 83% (95% CI: 0.59, 5.73, p=0.297); and the risk of SGA of the sleeping <7 h/day group increased 56% (95% CI: 0.84, 2.92, p=0.159). Details are shown in Table 3.
TE D
Stratified analyses were used to assess the potential effect modification for the impact of shorter sleep duration on birth length. Compared to the group sleeping 8 to <9 h/day, birth length of the sleeping <7 h/day group decreased by 8.03 mm (95% CI:
EP
-12.78, -3.29, p=0.001) among mothers without a midday napping habit, the effect
AC C
was stronger than among mothers with a midday napping habit, p-values for interaction were 0.011; birth length of sleeping <7 h/day group decreased 5.29 mm (95% CI: -7.89, -2.68, p<0.001) in male babies, the effect was stronger than in female babies, p-values for interaction were 0.002. Birth length of the group sleeping <7 h/day decreased by 5.59 mm (95% CI: -8.59, -2.60, p<0.001) among mothers with a history of abortion, and the effect was stronger than among mothers without an abortion history, the p-value for interaction was 0.008. Furthermore, pre-pregnancy
ACCEPTED MANUSCRIPT BMI, age and parity were also considered as potential effect modifiers; however, no statistically significant interaction was observed (all p for interaction >0.05). The
AC C
EP
TE D
M AN U
SC
RI PT
results of the stratified analyses are shown in Table 4.
ACCEPTED MANUSCRIPT Discussion In this large prospective cohort study, we observed that shorter sleep duration (<7 h/day) in early pregnancy was associated with shorter birth length. The negative
RI PT
impact of shorter sleep duration in early pregnancy on birth length was stronger in male babies, and among mothers with a history of abortion or without a midday napping habit.
SC
To the best of our knowledge, the present study is the first to report on the
M AN U
association between sleep duration in early pregnancy and birth length. This is clinically important because newborns with shorter birth length, not lower birth weight, have a higher risk of hospitalization due to any cause [22]. As mentioned earlier, only five studies had examined the association of sleep duration on fetal
TE D
growth. Their outcomes were birth weight, LBW and SGA, and our results were in line with all except one study that was conducted in Sri Lanka [14]. The Sri Lankan study found that when compared to the group sleeping >8 h/night, pregnant women
EP
sleeping ≤8 h/night were at higher risk of SGA. It should be noted that the result
AC C
above was based on the condition that SGA was defined as birth weights <5th centiles; when defined as <10th centiles, the association was not statistically significant [14]. In this study, with a much larger population size, we found no evidence that sleep duration was associated with birth weight, LBW or SGA. Some biological mechanisms may explain the effect of shorter sleep duration on birth length. Increase in inflammation may play an essential role in the impact of shorter sleep duration on birth length. Shorter sleep duration was reported to be
ACCEPTED MANUSCRIPT associated with higher circulating concentrations of inflammatory cytokines, including interleukin (IL)-6, IL-8 and tumor necrosis factor alpha (TNF-α), which may lead to increased inflammation [23–25]. This aberrant inflammatory response
RI PT
leads to decreased nitric oxide bio-availability, inhibited spiral artery remodeling, deficient trophoblast invasion, and may alter uteroplacental hemodynamics, and then leads to increased risk of fetal growth restriction [26,27]. Insulin resistance is another
SC
possible mechanism. Through increased sympathetic nervous activity, shorter sleep
M AN U
duration may lead to insulin resistance in pregnant women [28]. Since insulin resistance may result in slower fetal growth, shorter sleep duration may have a negative impact on birth length [29].
This study provides new evidence that the adverse effect of shorter sleep duration
TE D
in early pregnancy on birth length was stronger among mothers without a midday napping habit. Midday napping is a common practice in China, and it is traditionally considered to be a healthy lifestyle [30]. The effect of midday napping on health
EP
remains a matter of debate. It has been reported that midday napping, especially
AC C
long-duration midday napping, is associated with many chronic diseases, while other studies found that shift naps may play a beneficial role for night workers with shorter sleep duration [31,32]. In the present study, we found that compared to mothers without a napping habit, the impact of shorter sleep duration in early pregnancy on birth length was much weaker among mothers with a midday napping habit. This indicates that midday napping may play a beneficial role for newborns; for pregnant women who sleep less during early pregnancy, a midday napping habit may reduce
ACCEPTED MANUSCRIPT the impact of shorter sleep duration on the birth length of their children. The underlying biological mechanisms may due to the decreased inflammation. As mentioned earlier, increased inflammation caused by shorter sleep duration may lead
RI PT
to fetal growth restriction [24,26,27]; midday napping after a night of sleep loss can return the level of inflammatory biomarker, such as IL-6, to normal [33,34]. Our findings should be replicated and confirmed in further longitudinal research.
SC
Our analysis also suggested that the association of shorter sleep duration with birth
M AN U
length was stronger in male babies and among mothers with a history of abortion; the biological mechanisms remain poorly understood. Fetal gender may act as a possible modifier as follows: shorter sleep duration leads to increased inflammation in mothers and then results in adverse effects on the fetal supply of oxygen and nutrients [35].
TE D
Male babies’ growth rate is faster, and their supply of nutrients and oxygen is larger than in girls, therefore, the adverse impacts of shorter maternal sleep duration were more serious in male babies [36,37]. These preliminary findings need to be confirmed
EP
in other populations.
AC C
The main strength of this study is the prospective cohort study design. The prospective design allows us to obtain abundant and accurate covariate information, which may avoid possible confounding bias. In addition, the population size of this study is much larger than that of all previous similar studies. There are some limitations in this study that deserve acknowledgment. First, sleep duration was measured by a self-reported questionnaire, which is less accurate than objective measures such as actigraphy or polysomnography; this may cause a small
ACCEPTED MANUSCRIPT measurement error and misclassification. However, it is reported that in population-based epidemiological studies, such as the present study, self-reports of sleep duration are highly correlated with objective measures of sleep duration [38].
RI PT
Second, although the sample size was large in the present study, the number of pregnant women sleeping <7 h/day in subgroups for stratified analyses was relatively small, and requires replication to increase the confidence in our findings in future
SC
cohort studies. Third, although we adjusted for various variables that were associated
M AN U
with fetal growth, there may be some undiscovered potential confounders. Conclusions
The present prospective cohort study observed an adverse impact of shorter sleep duration in early pregnancy on birth length, and the impact was stronger on male
TE D
babies, and among mothers with a history of abortion or without a midday napping habit. Our findings suggest that regulating sleep duration in pregnancy is vital for slow fetal growth prevention, and midday napping may be a protective factor for fetal
EP
growth among pregnant women with shorter sleep duration. Future studies are needed
AC C
to confirm these findings, especially the potential effect modification, and to explore the potential biological mechanisms. Conflict of interests
The authors have no conflicts of interest to declare. Acknowledgements We gratefully acknowledge all the participants who took part in this study for their cooperation. We would like to thank all members of the TMCHC study Group.
ACCEPTED MANUSCRIPT WW, SW and NY participated in conception and study design. WW, CZ, YZ, LH, XC, XZ, RC, XL, MX collected data. SW, XY, LHao and NY supervised the study conduct. WW, SW contributed in the statistical analysis. WW, SW wrote the manuscript. All of
RI PT
authors have read and approved the final version of the manuscript. This work was supported by National Program on Basic Research Project of China (NO.2013FY114200). The funding sources had no role in study design, or in study
SC
conduct, or in the collection, analysis, interpretation of data, or in the writing of the
M AN U
manuscript, or in the decision to submit the manuscript for publication.
References
[1] Miller JE, Hammond GC, Strunk T, et al. Association of gestational age and
TE D
growth measures at birth with infection-related admissions to hospital throughout childhood: a population-based, data-linkage study from Western Australia. Lancet Infect Dis 2016;16:952-61.
EP
DOI: 10.1016/S1473-3099(16)00150-X
AC C
[2] Li Y, Ley SH, VanderWeele TJ, et al. Joint association between birth weight at term and later life adherence to a healthy lifestyle with risk of hypertension: a prospective cohort study. BMC Med 2015;13:175. DOI: 10.1186/s12916-015-0409-1 [3] Zimmermann E, Gamborg M, Sorensen TI, et al. Sex Differences in the Association Between Birth Weight and Adult Type 2 Diabetes. Diabetes 2015;64:4220-5.
ACCEPTED MANUSCRIPT DOI: 10.2337/db15-0494 [4] Yang TO, Reeves GK, Green J, et al. Birth weight and adult cancer incidence: large prospective study and meta-analysis. Ann Oncol 2014;25:1836-43.
RI PT
DOI: 10.1093/annonc/mdu214 [5] Chen LW, Wu Y, Neelakantan N, et al. Maternal caffeine intake during pregnancy
meta-analysis. BMC Med 2014;12:174.
M AN U
DOI: 10.1186/s12916-014-0174-6
SC
is associated with risk of low birth weight: a systematic review and dose-response
[6] Grivell RM, Yelland LN, Deussen A, et al. Antenatal dietary and lifestyle advice for women who are overweight or obese and the effect on fetal growth and adiposity: the LIMIT randomised trial. BJOG 2016;123:233-43.
TE D
DOI: 10.1111/1471-0528.13777
[7] Howe LD, Signal TL, Paine SJ, et al. Self-reported sleep in late pregnancy in relation to birth size and fetal distress: the E Moe, Mama prospective cohort study.
EP
BMJ Open 2015;5:e008910.
AC C
[8] Shan Z, Ma H, Xie M, et al. Sleep duration and risk of type 2 diabetes: a meta-analysis of prospective studies. Diabetes Care 2015;38(3):529-37. DOI: 10.2337/dc14-2073 [9] Covassin N, Singh P. Sleep Duration and Cardiovascular Disease Risk: Epidemiologic and Experimental Evidence. Sleep Med Clin 2016;11:81-9. DOI: 10.1016/j.jsmc.2015.10.007 [10] Gangwisch JE. A review of evidence for the link between sleep duration and
ACCEPTED MANUSCRIPT hypertension. Am J Hypertens 2014;27:1235-42. DOI: 10.1093/ajh/hpu071 [11] Signal TL, Paine SJ, Sweeney B, et al. Prevalence of abnormal sleep duration and
RI PT
excessive daytime sleepiness in pregnancy and the role of socio-demographic factors: comparing pregnant women with women in the general population. Sleep Med 2014;15:1477-83.
SC
DOI: 10.1016/j.sleep.2014.07.007
M AN U
[12] Pien GW, Schwab RJ. Sleep disorders during pregnancy. Sleep 2004;27:1405-17. [13] Mindell JA, Cook RA, Nikolovski J. Sleep patterns and sleep disturbances across pregnancy. Sleep Med 2015;16:483-8. DOI: 10.1016/j.sleep.2014.12.006
TE D
[14] 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.
EP
DOI: 10.1111/j.1479-828X.2009.01010.x
AC C
[15] Micheli K, Komninos I, Bagkeris E, et al. Sleep patterns in late pregnancy and risk of preterm birth and fetal growth restriction. Epidemiology 2011;22:738-44. DOI: 10.1097/EDE.0b013e31822546fd [16] Zafarghandi N, Hadavand S, Davati A, et al. The effects of sleep quality and duration in late pregnancy on labor and fetal outcome. J Matern Fetal Neonatal Med 2012;25:535-7. DOI: 10.3109/14767058.2011.600370
ACCEPTED MANUSCRIPT [17] Owusu JT, Anderson FJ, Coleman J, et al. Association of maternal sleep practices with pre-eclampsia, low birth weight, and stillbirth among Ghanaian women. Int J Gynaecol Obstet 2013;121:261-5.
RI PT
DOI: 10.1016/j.ijgo.2013.01.013 [18] Andersen AM, Osler M. Birth dimensions, parental mortality, and mortality in early adult age: a cohort study of Danish men born in 1953. Int J Epidemiol
SC
2004;33:92-9.
M AN U
DOI: 10.1093/ije/dyg195
[19] Kajantie E, Osmond C, Barker DJ, et al. Size at birth as a predictor of mortality in adulthood: a follow-up of 350 000 person-years. Int J Epidemiol 2005;34:655-63. DOI: 10.1093/ije/dyi048
TE D
[20] Buysse DJ, Reynolds CF 3rd, Monk TH, et al. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiat Res 1989;28:193-213.
EP
DOI: 10.1016/0165-1781(89)90047-4
AC C
[21] Chen Y, Wu L, Zou L, et al. Update on the birth weight standard and its diagnostic value in small for gestational age (SGA) infants in China. J Matern Fetal Neonatal Med 2016:1-7. DOI: 10.1080/14767058.2016.1186636 [22] Kato T, Yorifuji T, Inoue S, et al. Association of birth length and risk of hospitalisation among full-term babies in Japan. Paediatr Perinat Epidemiol 2013;27:361-70.
ACCEPTED MANUSCRIPT DOI: 10.1111/ppe.12062 [23] Okun ML, Hall M, Coussons-Read ME. Sleep disturbances increase interleukin-6 production during pregnancy: implications for pregnancy complications.
RI PT
Reprod Sci 2007;14:560-7. DOI: 10.1177/1933719107307647
[24] Okun ML, Luther JF, Wisniewski SR, et al. Disturbed sleep and inflammatory
SC
cytokines in depressed and nondepressed pregnant women: an exploratory analysis of
M AN U
pregnancy outcomes. Psychosom Med 2013;75:670-81. DOI: 10.1097/PSY.0b013e31829cc3e7
[25] Patel SR, Zhu X, Storfer-Isser A, et al. Sleep duration and biomarkers of inflammation. Sleep 2009;32:200-4.
TE D
[26] Romero R, Espinoza J, Goncalves LF, et al. Inflammation in preterm and term labour and delivery. Semin Fetal Neonatal Med 2006;11:317-26. DOI: 10.1016/j.siny.2006.05.001
EP
[27] Cotechini T, Komisarenko M, Sperou A, et al. Inflammation in rat pregnancy
AC C
inhibits spiral artery remodeling leading to fetal growth restriction and features of preeclampsia. J Exp Med 2014;211:165-79. DOI: 10.1084/jem.20130295 [28] Spiegel K, Leproult R, Van Cauter E. Impact of sleep debt on metabolic and endocrine function. Lancet 1999;354:1435-9. DOI: 10.1016/S0140-6736(99)01376-8 [29] Lain KY, Catalano PM. Factors that affect maternal insulin resistance and modify
ACCEPTED MANUSCRIPT fetal growth and body composition. Metab Syndr Relat Disord 2006;4:91-100. DOI: 10.1089/met.2006.4.91 [30] Lan TY, Lan TH, Wen CP, et al. Nighttime sleep, Chinese afternoon nap, and
RI PT
mortality in the elderly. Sleep 2007;30:1105-10. [31] Lucassen EA. To nap or not to nap: is the Wujiao a healthy habit? Sleep Med 2013;14:929-30.
SC
DOI: 10.1016/j.sleep.2013.05.010
M AN U
[32] Rotenberg L, Silva-Costa A, Vasconcellos-Silva PR, et al. Work schedule and self-reported hypertension - the potential beneficial role of on-shift naps for night workers. Chronobiol Int 2016;33:697-705. DOI: 10.3109/07420528.2016.1167721
TE D
[33] Vgontzas AN, Pejovic S, Zoumakis E, et al. Daytime napping after a night of sleep loss decreases sleepiness, improves performance, and causes beneficial changes in cortisol and interleukin-6 secretion. Am J Physiol Endocrinol Metab
EP
2007;292:E253-61.
AC C
DOI: 10.1152/ajpendo.00651.2005 [34] Faraut B, Boudjeltia KZ, Dyzma M, et al. Benefits of napping and an extended duration of recovery sleep on alertness and immune cells after acute sleep restriction. Brain Behav Immun 2011;25:16-24. DOI: 10.1016/j.bbi.2010.08.001 [35] Izci-Balserak B, Pien GW. Sleep-disordered breathing and pregnancy: potential mechanisms and evidence for maternal and fetal morbidity. Curr Opin Pulm Med
ACCEPTED MANUSCRIPT 2010;16:574-82. DOI: 10.1097/MCP.0b013e32833f0d55 [36] Edwards A, Megens A, Peek M, et al. Sexual origins of placental dysfunction.
DOI: 10.1016/S0140-6736(99)05061-8
RI PT
Lancet 2000;355:203-4.
prematurity. BJOG 2005;112:140-4.
M AN U
DOI: 10.1111/j.1471-0528.2004.00308.x
SC
[37] Ghidini A, Salafia CM. Gender differences of placental dysfunction in severe
[38] Lauderdale DS, Knutson KL, Yan LL, et al. Self-reported and measured sleep duration: how similar are they? Epidemiology 2008;19:838-45.
TE D
DOI: 10.1097/EDE.0b013e318187a7b0
AC C
EP
Fig. 1. Flow chart of study population for analysis.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 1. Demographic and clinical characteristics of study population. Characteristics Total Sleep duration (h/day) (N=3567) ≥9 8 to <9 (N=1290) (N=1563) Age (years) 28.21±3.38 27.61±3.17 28.38±3.48 Height (cm) 160.46±4.95 160.29±4.99 160.44±4.99 Pre-pregnancy weight (kg) 53.51±7.54 52.79±7.45 53.51±7.50 2 Pre-pregnancy BMI (kg/m ) 20.77±2.67 20.53±2.60 20.78±2.65 Gestational weight gain (kg) 16.09±4.43 16.17±4.42 16.15±4.42 Education level (schooling years) ≤12 442 (12.4%) 200 (15.5%) 169 (10.8%) 12–15 937 (26.3%) 390 (30.2%) 378 (24.2%) ≥16 2188 (61.3%) 700 (54.3%) 1016 (65.0%) a Average personal income (CNY ) ≤2999 229 (6.4%) 102 (7.9%) 94 (6.0%) 3000–4999 1134 (31.8%) 425 (32.9%) 482 (30.8%) 5000–9999 1504 (42.2%) 528 (40.9%) 668 (42.7%) ≥10000 700 (19.6%) 235 (18.2%) 319 (20.4%) Ethnicity Han Chinese 3475 (97.4%) 1255 (97.3%) 1523 (97.4%) Others 92 (2.6%) 35 (2.7%) 40 (2.6%) Smoking Yes 119 (3.3%) 41 (3.2%) 45 (2.9%) No 3448 (96.7%) 1249 (96.8%) 1518 (97.1%) Alcohol consumption Yes 57 (1.6%) 22 (1.7%) 20 (1.3%) No 3510 (98.4%) 1268 (98.3%) 1543 (98.7%) Midday-napping No 611 (17.1%) 261 (20.2%) 246 (15.7%) Yes 2956 (82.9%) 1029 (79.8%) 1317 (84.3%) Parity Primiparous 3044 (85.3%) 1132 (87.8%) 1326 (84.8%) Multiparous 523 (14.7%) 158 (12.2%) 237 (15.2%) Abortion history No 2317 (65.0%) 814 (63.1%) 1034 (66.2%) Yes 1250 (35.0%) 476 (36.9%) 529 (33.8%) Gestational age (weeks) 39.63±1.49 39.69±1.47 39.60±1.50 Fetal gender Male 1911 (53.6%) 694 (53.8%) 825 (52.8%) Female 1656 (46.4%) 596 (46.2%) 738 (47.2%) a
Chinese Yuan; 1 CNY ≈ 0.15 US Dollars.
7 to <8 (N=550) 28.90 ± 3.3 160.74±4.8 54.56±7.47 21.11±2.70 15.72±4.45
50 (9.1%) 123 (22.4% 377 (68.5%
26 (4.7%) 173 (31.5% 238 (43.3% 113 (20.5%
538 (97.8% 12 (2.2%)
26 (4.7%) 524 (95.3%
10 (1.8%) 540 (98.2%
79 (14.4%) 471 (85.6%
451 (82.0% 99 (18.0%)
370 (67.3% 180 (32.7% 39.63±1.54
297 (54.0% 253 (46.0%
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 2. Univariate association between sleep duration in early pregnancy and birth length, birth weight, low birth weight or small for gestational age. Characteristics Sleep duration (h/day) ≥9 8 to <9 7 to <8 <7 (N=1290) (N=1563) (N=550) (N=164) * Birth length (mm) 501.76±13.45 501.33±12.95 502.58±12.50 499.88±15.65 Birth weight (g) 3339.58±432.50 3329.36±436.19 3369.54±442.77 3324.76±428.19 LBW Yes 29 (2.2%) 34 (2.2%) 11 (2.0%) 5 (3.0%) No 1261 (97.8%) 1529 (97.8%) 539 (98.0%) 159 (97.0%) SGA Yes 85 (6.6%) 100 (6.4%) 30 (5.5%) 14 (8.5%) No 1205 (93.4%) 1463 (93.6%) 520 (94.5%) 150 (91.5%) Sleeping 8 to <9 h/day group was used as reference group. LBW, low birth weight; SGA, small for gestational age. * p<0.05.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 3. Multivariate association between sleep duration in early pregnancy and birth length, birth weight, low birth weight or small for gestational age. Birth length Birth weight Sleep duration (h/day) N LBW (OR (95% CI (β (95% CI), mm) (β (95% CI), g) ≥9 1290 0.34 (-0.50, 1.18) 13.12 (-14.46, 40.70) 1.10 (0.59, 2.05) 8 to <9 (reference 1563 0.00 (0.00, 0.00) 0.00 (0.00, 0.00) 1.00 (1.00, 1.00) group) 7 to <8 550 0.59 (-0.50, 1.69) 16.76 (-19.29, 52.81) 1.12 (0.48, 2.62) ** <7 164 -2.42 (-4.27, -0.58) -42.70 (-103.02, 17.62) 1.83 (0.59, 5.73) All models were adjusted for maternal age, education level, alcohol consumption, smoking, ethnicity, height, pre-pregnancy body mass index, gestational weight gain, gestational age, average personal income, parity, fetal gender, and midday napping. CI, confidence interval; LBW, low birth weight; OR, odds ratio; SGA, small for gestational age. ** p<0.01.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 4. Stratified analyses: potential effect modification for the impact of short sleep duration on birth length. Stratification Category N1 a N2 b β (95% CI) c pd pint e factors Age <30 years 1083 101 -1.38 (-3.73, 0.96) 0.308 0.243 ≥30 years 480 63 -3.59 (-6.61, -0.57) 0.020 Midday-napping No 246 25 -8.03 (-12.78, -3.29) 0.001 0.011 Yes 1317 139 -1.40 (-3.39, 0.59) 0.169 2 Pre-pregnancy <24 kg/m 1382 133 -2.47 (-4.48, -0.45) 0.017 0.569 BMI ≥24 kg/m2 181 31 -2.16 (-7.07, 2.75) 0.389 Parity Primiparous 1326 135 -1.92 (-3.96, 0.12) 0.065 0.307 Multiparous 237 29 -5.01 (-9.33, -0.70) 0.023 Fetal gender Boy 825 95 -5.29 (-7.89, -2.68) <0.001 0.002 Girl 738 69 1.03 (-1.48, 3.54) 0.422 Abortion history No 1034 99 -0.40 (-2.74, 1.94) 0.736 0.008 Yes 529 65 -5.59 (-8.59, -2.60) <0.001 All covariates are as follows: maternal age, education level, alcohol consumption, smoking, ethnicity, height, pre-pregnancy body mass index, gestational weight gain, gestational age, average personal income, parity, fetal gender and midday napping. Analyses that stratified by abortion history were adjusted for all covariates mentioned above, each of the else groups were adjusted for all covariates except itself. BMI, Body mass index; CI, confidence interval. a Sample size of the sleeping 8 to <9 h/day group in each subgroup. b Sample size of the sleeping <7 h/day group in each subgroup. c Change (95% CI) in mean birth length (mm, sleeping <7 h/day vs sleeping 8 to <9 h/day). d p-value for the effect in each subgroup. e p-value for interaction based on Wald test.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights: •
Shorter sleep duration in early pregnancy was associated with birth length.
•
The association was much stronger among mothers without midday napping
No evidence was found that sleep duration was associated with birth weight.
AC C
EP
TE D
M AN U
SC
•
RI PT
habit.
S1389-9457(17)30149-1
DOI:
10.1016/j.sleep.2017.03.013
Reference:
SLEEP 3355
To appear in:
Sleep Medicine
Received Date: 25 November 2016 Revised Date:
20 February 2017
Accepted Date: 8 March 2017
Please cite this article as: Wang W, Zhong C, Zhang Y, Huang L, Chen X, Zhou X, Chen R, Li X, Xiao M, Hao L, Yang X, Yang N, Wei S, Shorter sleep duration in early pregnancy is associated with birth length: a prospective cohort study in Wuhan, China, Sleep Medicine (2017), doi: 10.1016/ j.sleep.2017.03.013. 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.
ACCEPTED MANUSCRIPT Shorter sleep duration in early pregnancy is associated with birth length: a prospective cohort study in Wuhan, China Weiye Wang a, Chunrong Zhong b, Yu Zhang b, Li Huang b, Xi Chen b, Xuezhen Zhou , Renjuan Chen b, Xiating Li b, Mei Xiao c, Liping Hao b, Xuefeng Yang b, Nianhong
RI PT
b
Yangb,*,†, Sheng Wei a,*,† a
Department of Epidemiology and Biostatistics, Ministry of Education Key Lab of
SC
Environment and Health, School of Public Health, Tongji Medical College, Huazhong
b
M AN U
University of Science and Technology, Hubei, P.R. China.
Department of Nutrition and Food Hygiene, Hubei Key Laboratory of Food
Nutrition and Safety, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Hubei, P.R. China.
Department of Obstetrics and Gynaecology, Hubei Maternal and Child Health
TE D
3
Hospital, Wuhan, Hubei, P.R. China.
* Corresponding authors. 13 Hangkong Road, Wuhan, Hubei, 430030, China. Tel.:
EP
+86 27 83650521 (N. Yang); +86 27 83692031 (S. Wei).
AC C
E-mail addresses: [email protected] (N. Yang), [email protected] (S. Wei).
†
These authors contributed equally to this work.
ACCEPTED MANUSCRIPT ABSTRACT Objectives: To examine the association between sleep duration in early pregnancy and fetal growth in a prospective cohort study of 3567 Chinese women.
RI PT
Methods: Pregnant women at 8–16 weeks of gestation were interviewed using a semi-quantitative questionnaire to assess sleep duration. Birth weight and birth length were measured by a midwife in the delivery room at birth; low birth weight (LBW)
SC
was defined as birth weight <2500 g and small for gestational age (SGA) was defined
M AN U
as <10th customized centile.
Results: The average age of participants was 28.21±3.38 years old. The mean sleep duration was 8.39±1.13 h/day. A total of 1290 women sleeping ≥9 h/day, 1563 sleeping 8 to <9 h/day, 550 sleeping 7 to <8 h/day, and 164 sleeping <7 h/day.
TE D
Compared to the sleeping 8 to <9 h/day group, birth length and birth weight of the sleeping <7 h/day group decreased by 2.42 mm (95% CI: -4.27, -0.58, p=0.010) and 42.70 g (95% CI: -103.02, 17.62, p=0.165), respectively; and risk of LBW and SGA
EP
of the sleeping <7 h/day group increased by 83% (95% CI: 0.59, 5.73, p=0.297) and
AC C
56% (95% CI: 0.84, 2.92, p=0.159), respectively; birth length of the sleeping <7 h/day group was decreased more in male babies, and among mothers without a midday napping habit or with a history of abortion (all p for interaction <0.05). Conclusions: Shorter sleep duration in early pregnancy was associated with birth length. Our findings indicate that midday napping may be a protective factor for birth length among pregnant women with shorter sleep duration.
ACCEPTED MANUSCRIPT Keywords: Sleep duration Midday napping
RI PT
Fetal growth Birth length
AC C
EP
TE D
M AN U
SC
Birth weight
ACCEPTED MANUSCRIPT Introduction Slow fetal growth is an important public health concern worldwide, and newborns with low birth length or weight often suffer negative short-term and long-term
RI PT
consequences during infancy, childhood and adulthood. Epidemiological studies have suggested that lower birth weight or shorter birth length are related to a higher risk of infection-related hospitalization throughout childhood [1]. Various chronic diseases in
SC
adult life, such as hypertension, diabetes and cancer, were also found to be
M AN U
consequences of slow fetal growth [2–4]. Many factors were reported to be related to fetal growth, such as dietary intake and lifestyle [5,6]. Additionally, short sleep duration in pregnant women has recently been recognized as a risk factor for slow fetal growth [7].
TE D
Sleep duration that is too short or too long is considered to be associated with many diseases, for example, diabetes mellitus and cardiovascular diseases, the population in most of the studies are middle-aged and elderly people [8–10]. Pregnant women may
EP
sleep less during pregnancy; Signal et al. found that pregnant women reported a mean
AC C
decrease in sleep duration of 0.5 h, compared to women in the general population [11]. There are some physiological and psychological changes during pregnancy, including rapid growth of weight, increased arousals from sleep, and decreased functional reserve capacity [12]. These changes may make pregnant women suffer from shorter sleep duration in pregnancy [13]. As far as we know, only five studies have examined the association between sleep duration in pregnancy and fetal growth, and the conclusions were inconsistent [7,14–
ACCEPTED MANUSCRIPT 17]. One study observed that women who slept for ≤8 h/day were at higher risk of having small for gestational age (SGA) neonates when compared to those who sleep for >8 h/day [14]. Other studies did not find a significant association between shorter
RI PT
sleep duration and neonate birth weight, low birth weight (LBW), or SGA [7,15–17]. However, all of these studies were focused on birth weight, none had determined the association between sleep duration in pregnancy and birth length. Birth length is
SC
another indicator of slow fetal growth, in addition to birth weight. Indeed, some
M AN U
studies indicated that only short birth length, not low birth weight, is associated with a higher risk of poor health [18,19]. Furthermore, these studies did not consider the potential effect modifiers of sleep duration.
To assess the potential impact of sleep duration of women in early pregnancy on
TE D
fetal growth, we conducted a prospective mother–child cohort study in Wuhan, central China. We further explored the potential effect modifiers of sleep duration in early
Methods
EP
pregnancy on fetal growth.
AC C
Study population
The present study was embedded in the Tongji Maternal and Child Health Cohort
(TMCHC) study. The TMCHC is an ongoing prospective cohort with pregnant women and their children residing in Wuhan, Hubei province, central China. Its primary objective is to determine the effect of environmental pollution, dietary, lifestyle and biochemical factors on the health of pregnant women and their children. Women who were 8–16 weeks pregnant were invited to participate in the TMCHC
ACCEPTED MANUSCRIPT during their first antenatal visit in Hubei Maternal and Child Health Care Hospital, or the Central Hospital of Wuhan, or Jiangan District Maternal and Child Health Care Hospital of Wuhan, beginning in January 2013. The inclusion criteria for TMCHC
RI PT
were as follows: (1) aged >18 years old, (2) residing in Wuhan, (3) no communication problems.
Participants in the present study were all from TMCHC, and the exclusion criteria
SC
were: (1) with stillborn infants and (2) with multiple pregnancy. The study was
M AN U
approved by the Ethics review committee of Tongji Medical College, Huazhong University of Science and Technology (NO. 201302). All study participants had provided written informed consent after a complete description of the study. Sleep duration
TE D
Information on sleep duration was assessed by a semi-quantitative questionnaire, which referenced to question 1 to question 4 of Pittsburg Sleep Quality Index [20]. Women were interviewed face-to-face by trained investigators when they were
EP
enrolled (at weeks 8–16 of gestation). The following questions were adopted: “When
AC C
have you usually gone to bed at night during the past month?” “How long has it usually taken you to fall asleep every night during the past month?” “When have you usually gotten up in the morning during the past month?” “How many hours did you sleep every night during the past month?” The responses of sleep duration were categorized into the following four groups: ≥9 h/day, 8 to <9 h/day, 7 to <8 h/day, and <7 h/day. Neonatal outcomes
ACCEPTED MANUSCRIPT The data on neonatal outcomes were obtained from hospital medical records, including neonatal gender and date of birth. Birth weight and birth length were measured by a midwife in the delivery room at birth. LBW was defined as birth
RI PT
weight <2500 g, and SGA was defined as <10th customized centile [21]. Gestational age at delivery was based on the self-reported last menstrual period (LMP) and delivery date. Fetal crown rump length (CRL), which was measured by ultrasound in
SC
early pregnancy, was also used to calculate gestational age at delivery for women.
M AN U
When women could not accurately remember the LMP, or the difference between LMP-based and CRL-based gestational age was more than 10 days, we chose the latter. For women who were lost to follow-up or whose child’s hospital medical records could not be found, information on neonatal outcomes was obtained by
Covariables
TE D
telephone interview.
A series of covariates were obtained from self-reported questionnaires at enrollment
EP
as follows: maternal age when enrolled (year), self-reported pre-pregnancy weight
AC C
(kg), ethnicity (Han Chinese/others), education level, smoking (active smoking or smoking before pregnancy, defined as ≥3 times a week, yes/no, respectively), alcohol consumption (active drinking or drinking before pregnancy, defined as ≥3 times a week, yes/no, respectively), average personal income (per month, CNY), abortion history (yes/no), parity (primiparous/multiparous) and midday napping (h/day). Education level was recoded according to the number of completed schooling years (≤12/12–15/≥16). Maternal height (cm) was measured at enrollment. Maternal weight
ACCEPTED MANUSCRIPT was measured on admission to hospital while awaiting delivery to calculate gestational weight gain (GWG). Statistical analyses
RI PT
Distributions of demographic characteristics were described using mean ± SDs for continuous variables, and using frequencies and percentages for categorical variables. One-way analysis of variance (ANOVA) or Pearson χ2-test were used to compare
SC
demographic characteristics of pregnant women grouped by sleep duration. Multiple
M AN U
linear regression analyses were applied to examine the association of sleep duration in pregnancy with birth weight or birth length, and the results are presented as coefficients (β) with 95% confidence intervals (CIs). Binomial logistic regression analyses were used to assess the association between sleep duration in pregnancy and
TE D
LBW or SGA, and the results are presented as odds ratios (ORs) with 95% CIs. Covariates that were related to sleep duration and fetal growth reported in the literature were chosen as potential confounders, including maternal age, education
EP
level, alcohol consumption, smoking, ethnicity, height, pre-pregnancy body mass
AC C
index (BMI), gestational weight gain, gestational age at delivery, average personal income, parity, fetal gender, and midday napping. Stratified analysis was used to explore the potential modification effect. The
following factors were considered as potential effect modifiers: pre-pregnancy BMI (<24, ≥24 kg/m2), age (<30/≥30 years), parity (primiparous/multiparous), fetal gender (boy/girl), midday napping (yes/no), and abortion history (yes/no). The Wald test was used to test the potential effect modification by adding interaction terms of these
ACCEPTED MANUSCRIPT covariates with sleep duration in early pregnancy. All statistical analyses were performed with R software (The R Project for Statistical Computing, version 3.3.1), and were conducted assuming a two-sided
RI PT
alternative hypothesis; p<0.05 was considered to be statistically significant. Results
A total of 4539 pregnant women who were eligible for inclusion in this cohort were
SC
recruited. Of those, 492 were lost to follow-up, 196 had stillbirths and 106 had twins,
M AN U
and 178 were excluded because of missing information on covariates, leaving 3567 women for the final analysis. Fig. 1 shows the processes of participants’ inclusion. The average age of the participants was 28.21±3.38 years old. The mean sleep duration in early pregnancy was 8.39±1.13 h/day. Average birth length was
TE D
501.61±13.21 mm, and average birth weight was 3339.04±435.56 g. The proportions of LBW and SGA were 2.2% and 6.4%, respectively. Pregnant women were divided into four groups according to their sleep duration in early pregnancy, with 1290
EP
(36.2%) pregnant women sleeping ≥9 h/day, 1563 (43.8%) sleeping 8 to <9 h/day, 550
AC C
(15.4%) sleeping 7 to <8 h/day, and 164 (4.6%) sleeping <7 h/day. The differences of age, pre-pregnancy weight, pre-pregnancy BMI, educational level, midday-napping and parity among the four groups were significant (all p<0.05). Details of the demographic and clinical characteristics are shown in Table 1. The one-way ANOVA shows that compared to the group sleeping 8 to <9 h/day, birth length of the group sleeping <7 h/day decreased by 1.88 mm, birth weight of the sleeping <7 h/day group decreased by 5.40 g, but the differences were not statistically
ACCEPTED MANUSCRIPT significant (all p>0.05). Compared to the group sleeping 8 to <9 h/day, the proportions of LBW and SGA in the group sleeping <7 h/day were lower, however, the Pearson χ2-test shows that differences were not statistical significant (all p>0.05).
RI PT
Details are shown in Table 2. After adjustment for a series of covariates, compared to the group sleeping 8 to <9 h/day, birth length of the group sleeping <7 h/day decreased by 2.42 mm (95% CI:
SC
-4.27, -0.58, p=0.010); birth weight of the sleeping <7 h/day group decreased by
M AN U
42.70 g (95% CI: -103.02, 17.62, p=0.165), the risk of LBW of the sleeping <7 h/day group increased 83% (95% CI: 0.59, 5.73, p=0.297); and the risk of SGA of the sleeping <7 h/day group increased 56% (95% CI: 0.84, 2.92, p=0.159). Details are shown in Table 3.
TE D
Stratified analyses were used to assess the potential effect modification for the impact of shorter sleep duration on birth length. Compared to the group sleeping 8 to <9 h/day, birth length of the sleeping <7 h/day group decreased by 8.03 mm (95% CI:
EP
-12.78, -3.29, p=0.001) among mothers without a midday napping habit, the effect
AC C
was stronger than among mothers with a midday napping habit, p-values for interaction were 0.011; birth length of sleeping <7 h/day group decreased 5.29 mm (95% CI: -7.89, -2.68, p<0.001) in male babies, the effect was stronger than in female babies, p-values for interaction were 0.002. Birth length of the group sleeping <7 h/day decreased by 5.59 mm (95% CI: -8.59, -2.60, p<0.001) among mothers with a history of abortion, and the effect was stronger than among mothers without an abortion history, the p-value for interaction was 0.008. Furthermore, pre-pregnancy
ACCEPTED MANUSCRIPT BMI, age and parity were also considered as potential effect modifiers; however, no statistically significant interaction was observed (all p for interaction >0.05). The
AC C
EP
TE D
M AN U
SC
RI PT
results of the stratified analyses are shown in Table 4.
ACCEPTED MANUSCRIPT Discussion In this large prospective cohort study, we observed that shorter sleep duration (<7 h/day) in early pregnancy was associated with shorter birth length. The negative
RI PT
impact of shorter sleep duration in early pregnancy on birth length was stronger in male babies, and among mothers with a history of abortion or without a midday napping habit.
SC
To the best of our knowledge, the present study is the first to report on the
M AN U
association between sleep duration in early pregnancy and birth length. This is clinically important because newborns with shorter birth length, not lower birth weight, have a higher risk of hospitalization due to any cause [22]. As mentioned earlier, only five studies had examined the association of sleep duration on fetal
TE D
growth. Their outcomes were birth weight, LBW and SGA, and our results were in line with all except one study that was conducted in Sri Lanka [14]. The Sri Lankan study found that when compared to the group sleeping >8 h/night, pregnant women
EP
sleeping ≤8 h/night were at higher risk of SGA. It should be noted that the result
AC C
above was based on the condition that SGA was defined as birth weights <5th centiles; when defined as <10th centiles, the association was not statistically significant [14]. In this study, with a much larger population size, we found no evidence that sleep duration was associated with birth weight, LBW or SGA. Some biological mechanisms may explain the effect of shorter sleep duration on birth length. Increase in inflammation may play an essential role in the impact of shorter sleep duration on birth length. Shorter sleep duration was reported to be
ACCEPTED MANUSCRIPT associated with higher circulating concentrations of inflammatory cytokines, including interleukin (IL)-6, IL-8 and tumor necrosis factor alpha (TNF-α), which may lead to increased inflammation [23–25]. This aberrant inflammatory response
RI PT
leads to decreased nitric oxide bio-availability, inhibited spiral artery remodeling, deficient trophoblast invasion, and may alter uteroplacental hemodynamics, and then leads to increased risk of fetal growth restriction [26,27]. Insulin resistance is another
SC
possible mechanism. Through increased sympathetic nervous activity, shorter sleep
M AN U
duration may lead to insulin resistance in pregnant women [28]. Since insulin resistance may result in slower fetal growth, shorter sleep duration may have a negative impact on birth length [29].
This study provides new evidence that the adverse effect of shorter sleep duration
TE D
in early pregnancy on birth length was stronger among mothers without a midday napping habit. Midday napping is a common practice in China, and it is traditionally considered to be a healthy lifestyle [30]. The effect of midday napping on health
EP
remains a matter of debate. It has been reported that midday napping, especially
AC C
long-duration midday napping, is associated with many chronic diseases, while other studies found that shift naps may play a beneficial role for night workers with shorter sleep duration [31,32]. In the present study, we found that compared to mothers without a napping habit, the impact of shorter sleep duration in early pregnancy on birth length was much weaker among mothers with a midday napping habit. This indicates that midday napping may play a beneficial role for newborns; for pregnant women who sleep less during early pregnancy, a midday napping habit may reduce
ACCEPTED MANUSCRIPT the impact of shorter sleep duration on the birth length of their children. The underlying biological mechanisms may due to the decreased inflammation. As mentioned earlier, increased inflammation caused by shorter sleep duration may lead
RI PT
to fetal growth restriction [24,26,27]; midday napping after a night of sleep loss can return the level of inflammatory biomarker, such as IL-6, to normal [33,34]. Our findings should be replicated and confirmed in further longitudinal research.
SC
Our analysis also suggested that the association of shorter sleep duration with birth
M AN U
length was stronger in male babies and among mothers with a history of abortion; the biological mechanisms remain poorly understood. Fetal gender may act as a possible modifier as follows: shorter sleep duration leads to increased inflammation in mothers and then results in adverse effects on the fetal supply of oxygen and nutrients [35].
TE D
Male babies’ growth rate is faster, and their supply of nutrients and oxygen is larger than in girls, therefore, the adverse impacts of shorter maternal sleep duration were more serious in male babies [36,37]. These preliminary findings need to be confirmed
EP
in other populations.
AC C
The main strength of this study is the prospective cohort study design. The prospective design allows us to obtain abundant and accurate covariate information, which may avoid possible confounding bias. In addition, the population size of this study is much larger than that of all previous similar studies. There are some limitations in this study that deserve acknowledgment. First, sleep duration was measured by a self-reported questionnaire, which is less accurate than objective measures such as actigraphy or polysomnography; this may cause a small
ACCEPTED MANUSCRIPT measurement error and misclassification. However, it is reported that in population-based epidemiological studies, such as the present study, self-reports of sleep duration are highly correlated with objective measures of sleep duration [38].
RI PT
Second, although the sample size was large in the present study, the number of pregnant women sleeping <7 h/day in subgroups for stratified analyses was relatively small, and requires replication to increase the confidence in our findings in future
SC
cohort studies. Third, although we adjusted for various variables that were associated
M AN U
with fetal growth, there may be some undiscovered potential confounders. Conclusions
The present prospective cohort study observed an adverse impact of shorter sleep duration in early pregnancy on birth length, and the impact was stronger on male
TE D
babies, and among mothers with a history of abortion or without a midday napping habit. Our findings suggest that regulating sleep duration in pregnancy is vital for slow fetal growth prevention, and midday napping may be a protective factor for fetal
EP
growth among pregnant women with shorter sleep duration. Future studies are needed
AC C
to confirm these findings, especially the potential effect modification, and to explore the potential biological mechanisms. Conflict of interests
The authors have no conflicts of interest to declare. Acknowledgements We gratefully acknowledge all the participants who took part in this study for their cooperation. We would like to thank all members of the TMCHC study Group.
ACCEPTED MANUSCRIPT WW, SW and NY participated in conception and study design. WW, CZ, YZ, LH, XC, XZ, RC, XL, MX collected data. SW, XY, LHao and NY supervised the study conduct. WW, SW contributed in the statistical analysis. WW, SW wrote the manuscript. All of
RI PT
authors have read and approved the final version of the manuscript. This work was supported by National Program on Basic Research Project of China (NO.2013FY114200). The funding sources had no role in study design, or in study
SC
conduct, or in the collection, analysis, interpretation of data, or in the writing of the
M AN U
manuscript, or in the decision to submit the manuscript for publication.
References
[1] Miller JE, Hammond GC, Strunk T, et al. Association of gestational age and
TE D
growth measures at birth with infection-related admissions to hospital throughout childhood: a population-based, data-linkage study from Western Australia. Lancet Infect Dis 2016;16:952-61.
EP
DOI: 10.1016/S1473-3099(16)00150-X
AC C
[2] Li Y, Ley SH, VanderWeele TJ, et al. Joint association between birth weight at term and later life adherence to a healthy lifestyle with risk of hypertension: a prospective cohort study. BMC Med 2015;13:175. DOI: 10.1186/s12916-015-0409-1 [3] Zimmermann E, Gamborg M, Sorensen TI, et al. Sex Differences in the Association Between Birth Weight and Adult Type 2 Diabetes. Diabetes 2015;64:4220-5.
ACCEPTED MANUSCRIPT DOI: 10.2337/db15-0494 [4] Yang TO, Reeves GK, Green J, et al. Birth weight and adult cancer incidence: large prospective study and meta-analysis. Ann Oncol 2014;25:1836-43.
RI PT
DOI: 10.1093/annonc/mdu214 [5] Chen LW, Wu Y, Neelakantan N, et al. Maternal caffeine intake during pregnancy
meta-analysis. BMC Med 2014;12:174.
M AN U
DOI: 10.1186/s12916-014-0174-6
SC
is associated with risk of low birth weight: a systematic review and dose-response
[6] Grivell RM, Yelland LN, Deussen A, et al. Antenatal dietary and lifestyle advice for women who are overweight or obese and the effect on fetal growth and adiposity: the LIMIT randomised trial. BJOG 2016;123:233-43.
TE D
DOI: 10.1111/1471-0528.13777
[7] Howe LD, Signal TL, Paine SJ, et al. Self-reported sleep in late pregnancy in relation to birth size and fetal distress: the E Moe, Mama prospective cohort study.
EP
BMJ Open 2015;5:e008910.
AC C
[8] Shan Z, Ma H, Xie M, et al. Sleep duration and risk of type 2 diabetes: a meta-analysis of prospective studies. Diabetes Care 2015;38(3):529-37. DOI: 10.2337/dc14-2073 [9] Covassin N, Singh P. Sleep Duration and Cardiovascular Disease Risk: Epidemiologic and Experimental Evidence. Sleep Med Clin 2016;11:81-9. DOI: 10.1016/j.jsmc.2015.10.007 [10] Gangwisch JE. A review of evidence for the link between sleep duration and
ACCEPTED MANUSCRIPT hypertension. Am J Hypertens 2014;27:1235-42. DOI: 10.1093/ajh/hpu071 [11] Signal TL, Paine SJ, Sweeney B, et al. Prevalence of abnormal sleep duration and
RI PT
excessive daytime sleepiness in pregnancy and the role of socio-demographic factors: comparing pregnant women with women in the general population. Sleep Med 2014;15:1477-83.
SC
DOI: 10.1016/j.sleep.2014.07.007
M AN U
[12] Pien GW, Schwab RJ. Sleep disorders during pregnancy. Sleep 2004;27:1405-17. [13] Mindell JA, Cook RA, Nikolovski J. Sleep patterns and sleep disturbances across pregnancy. Sleep Med 2015;16:483-8. DOI: 10.1016/j.sleep.2014.12.006
TE D
[14] 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.
EP
DOI: 10.1111/j.1479-828X.2009.01010.x
AC C
[15] Micheli K, Komninos I, Bagkeris E, et al. Sleep patterns in late pregnancy and risk of preterm birth and fetal growth restriction. Epidemiology 2011;22:738-44. DOI: 10.1097/EDE.0b013e31822546fd [16] Zafarghandi N, Hadavand S, Davati A, et al. The effects of sleep quality and duration in late pregnancy on labor and fetal outcome. J Matern Fetal Neonatal Med 2012;25:535-7. DOI: 10.3109/14767058.2011.600370
ACCEPTED MANUSCRIPT [17] Owusu JT, Anderson FJ, Coleman J, et al. Association of maternal sleep practices with pre-eclampsia, low birth weight, and stillbirth among Ghanaian women. Int J Gynaecol Obstet 2013;121:261-5.
RI PT
DOI: 10.1016/j.ijgo.2013.01.013 [18] Andersen AM, Osler M. Birth dimensions, parental mortality, and mortality in early adult age: a cohort study of Danish men born in 1953. Int J Epidemiol
SC
2004;33:92-9.
M AN U
DOI: 10.1093/ije/dyg195
[19] Kajantie E, Osmond C, Barker DJ, et al. Size at birth as a predictor of mortality in adulthood: a follow-up of 350 000 person-years. Int J Epidemiol 2005;34:655-63. DOI: 10.1093/ije/dyi048
TE D
[20] Buysse DJ, Reynolds CF 3rd, Monk TH, et al. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiat Res 1989;28:193-213.
EP
DOI: 10.1016/0165-1781(89)90047-4
AC C
[21] Chen Y, Wu L, Zou L, et al. Update on the birth weight standard and its diagnostic value in small for gestational age (SGA) infants in China. J Matern Fetal Neonatal Med 2016:1-7. DOI: 10.1080/14767058.2016.1186636 [22] Kato T, Yorifuji T, Inoue S, et al. Association of birth length and risk of hospitalisation among full-term babies in Japan. Paediatr Perinat Epidemiol 2013;27:361-70.
ACCEPTED MANUSCRIPT DOI: 10.1111/ppe.12062 [23] Okun ML, Hall M, Coussons-Read ME. Sleep disturbances increase interleukin-6 production during pregnancy: implications for pregnancy complications.
RI PT
Reprod Sci 2007;14:560-7. DOI: 10.1177/1933719107307647
[24] Okun ML, Luther JF, Wisniewski SR, et al. Disturbed sleep and inflammatory
SC
cytokines in depressed and nondepressed pregnant women: an exploratory analysis of
M AN U
pregnancy outcomes. Psychosom Med 2013;75:670-81. DOI: 10.1097/PSY.0b013e31829cc3e7
[25] Patel SR, Zhu X, Storfer-Isser A, et al. Sleep duration and biomarkers of inflammation. Sleep 2009;32:200-4.
TE D
[26] Romero R, Espinoza J, Goncalves LF, et al. Inflammation in preterm and term labour and delivery. Semin Fetal Neonatal Med 2006;11:317-26. DOI: 10.1016/j.siny.2006.05.001
EP
[27] Cotechini T, Komisarenko M, Sperou A, et al. Inflammation in rat pregnancy
AC C
inhibits spiral artery remodeling leading to fetal growth restriction and features of preeclampsia. J Exp Med 2014;211:165-79. DOI: 10.1084/jem.20130295 [28] Spiegel K, Leproult R, Van Cauter E. Impact of sleep debt on metabolic and endocrine function. Lancet 1999;354:1435-9. DOI: 10.1016/S0140-6736(99)01376-8 [29] Lain KY, Catalano PM. Factors that affect maternal insulin resistance and modify
ACCEPTED MANUSCRIPT fetal growth and body composition. Metab Syndr Relat Disord 2006;4:91-100. DOI: 10.1089/met.2006.4.91 [30] Lan TY, Lan TH, Wen CP, et al. Nighttime sleep, Chinese afternoon nap, and
RI PT
mortality in the elderly. Sleep 2007;30:1105-10. [31] Lucassen EA. To nap or not to nap: is the Wujiao a healthy habit? Sleep Med 2013;14:929-30.
SC
DOI: 10.1016/j.sleep.2013.05.010
M AN U
[32] Rotenberg L, Silva-Costa A, Vasconcellos-Silva PR, et al. Work schedule and self-reported hypertension - the potential beneficial role of on-shift naps for night workers. Chronobiol Int 2016;33:697-705. DOI: 10.3109/07420528.2016.1167721
TE D
[33] Vgontzas AN, Pejovic S, Zoumakis E, et al. Daytime napping after a night of sleep loss decreases sleepiness, improves performance, and causes beneficial changes in cortisol and interleukin-6 secretion. Am J Physiol Endocrinol Metab
EP
2007;292:E253-61.
AC C
DOI: 10.1152/ajpendo.00651.2005 [34] Faraut B, Boudjeltia KZ, Dyzma M, et al. Benefits of napping and an extended duration of recovery sleep on alertness and immune cells after acute sleep restriction. Brain Behav Immun 2011;25:16-24. DOI: 10.1016/j.bbi.2010.08.001 [35] Izci-Balserak B, Pien GW. Sleep-disordered breathing and pregnancy: potential mechanisms and evidence for maternal and fetal morbidity. Curr Opin Pulm Med
ACCEPTED MANUSCRIPT 2010;16:574-82. DOI: 10.1097/MCP.0b013e32833f0d55 [36] Edwards A, Megens A, Peek M, et al. Sexual origins of placental dysfunction.
DOI: 10.1016/S0140-6736(99)05061-8
RI PT
Lancet 2000;355:203-4.
prematurity. BJOG 2005;112:140-4.
M AN U
DOI: 10.1111/j.1471-0528.2004.00308.x
SC
[37] Ghidini A, Salafia CM. Gender differences of placental dysfunction in severe
[38] Lauderdale DS, Knutson KL, Yan LL, et al. Self-reported and measured sleep duration: how similar are they? Epidemiology 2008;19:838-45.
TE D
DOI: 10.1097/EDE.0b013e318187a7b0
AC C
EP
Fig. 1. Flow chart of study population for analysis.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 1. Demographic and clinical characteristics of study population. Characteristics Total Sleep duration (h/day) (N=3567) ≥9 8 to <9 (N=1290) (N=1563) Age (years) 28.21±3.38 27.61±3.17 28.38±3.48 Height (cm) 160.46±4.95 160.29±4.99 160.44±4.99 Pre-pregnancy weight (kg) 53.51±7.54 52.79±7.45 53.51±7.50 2 Pre-pregnancy BMI (kg/m ) 20.77±2.67 20.53±2.60 20.78±2.65 Gestational weight gain (kg) 16.09±4.43 16.17±4.42 16.15±4.42 Education level (schooling years) ≤12 442 (12.4%) 200 (15.5%) 169 (10.8%) 12–15 937 (26.3%) 390 (30.2%) 378 (24.2%) ≥16 2188 (61.3%) 700 (54.3%) 1016 (65.0%) a Average personal income (CNY ) ≤2999 229 (6.4%) 102 (7.9%) 94 (6.0%) 3000–4999 1134 (31.8%) 425 (32.9%) 482 (30.8%) 5000–9999 1504 (42.2%) 528 (40.9%) 668 (42.7%) ≥10000 700 (19.6%) 235 (18.2%) 319 (20.4%) Ethnicity Han Chinese 3475 (97.4%) 1255 (97.3%) 1523 (97.4%) Others 92 (2.6%) 35 (2.7%) 40 (2.6%) Smoking Yes 119 (3.3%) 41 (3.2%) 45 (2.9%) No 3448 (96.7%) 1249 (96.8%) 1518 (97.1%) Alcohol consumption Yes 57 (1.6%) 22 (1.7%) 20 (1.3%) No 3510 (98.4%) 1268 (98.3%) 1543 (98.7%) Midday-napping No 611 (17.1%) 261 (20.2%) 246 (15.7%) Yes 2956 (82.9%) 1029 (79.8%) 1317 (84.3%) Parity Primiparous 3044 (85.3%) 1132 (87.8%) 1326 (84.8%) Multiparous 523 (14.7%) 158 (12.2%) 237 (15.2%) Abortion history No 2317 (65.0%) 814 (63.1%) 1034 (66.2%) Yes 1250 (35.0%) 476 (36.9%) 529 (33.8%) Gestational age (weeks) 39.63±1.49 39.69±1.47 39.60±1.50 Fetal gender Male 1911 (53.6%) 694 (53.8%) 825 (52.8%) Female 1656 (46.4%) 596 (46.2%) 738 (47.2%) a
Chinese Yuan; 1 CNY ≈ 0.15 US Dollars.
7 to <8 (N=550) 28.90 ± 3.3 160.74±4.8 54.56±7.47 21.11±2.70 15.72±4.45
50 (9.1%) 123 (22.4% 377 (68.5%
26 (4.7%) 173 (31.5% 238 (43.3% 113 (20.5%
538 (97.8% 12 (2.2%)
26 (4.7%) 524 (95.3%
10 (1.8%) 540 (98.2%
79 (14.4%) 471 (85.6%
451 (82.0% 99 (18.0%)
370 (67.3% 180 (32.7% 39.63±1.54
297 (54.0% 253 (46.0%
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 2. Univariate association between sleep duration in early pregnancy and birth length, birth weight, low birth weight or small for gestational age. Characteristics Sleep duration (h/day) ≥9 8 to <9 7 to <8 <7 (N=1290) (N=1563) (N=550) (N=164) * Birth length (mm) 501.76±13.45 501.33±12.95 502.58±12.50 499.88±15.65 Birth weight (g) 3339.58±432.50 3329.36±436.19 3369.54±442.77 3324.76±428.19 LBW Yes 29 (2.2%) 34 (2.2%) 11 (2.0%) 5 (3.0%) No 1261 (97.8%) 1529 (97.8%) 539 (98.0%) 159 (97.0%) SGA Yes 85 (6.6%) 100 (6.4%) 30 (5.5%) 14 (8.5%) No 1205 (93.4%) 1463 (93.6%) 520 (94.5%) 150 (91.5%) Sleeping 8 to <9 h/day group was used as reference group. LBW, low birth weight; SGA, small for gestational age. * p<0.05.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 3. Multivariate association between sleep duration in early pregnancy and birth length, birth weight, low birth weight or small for gestational age. Birth length Birth weight Sleep duration (h/day) N LBW (OR (95% CI (β (95% CI), mm) (β (95% CI), g) ≥9 1290 0.34 (-0.50, 1.18) 13.12 (-14.46, 40.70) 1.10 (0.59, 2.05) 8 to <9 (reference 1563 0.00 (0.00, 0.00) 0.00 (0.00, 0.00) 1.00 (1.00, 1.00) group) 7 to <8 550 0.59 (-0.50, 1.69) 16.76 (-19.29, 52.81) 1.12 (0.48, 2.62) ** <7 164 -2.42 (-4.27, -0.58) -42.70 (-103.02, 17.62) 1.83 (0.59, 5.73) All models were adjusted for maternal age, education level, alcohol consumption, smoking, ethnicity, height, pre-pregnancy body mass index, gestational weight gain, gestational age, average personal income, parity, fetal gender, and midday napping. CI, confidence interval; LBW, low birth weight; OR, odds ratio; SGA, small for gestational age. ** p<0.01.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 4. Stratified analyses: potential effect modification for the impact of short sleep duration on birth length. Stratification Category N1 a N2 b β (95% CI) c pd pint e factors Age <30 years 1083 101 -1.38 (-3.73, 0.96) 0.308 0.243 ≥30 years 480 63 -3.59 (-6.61, -0.57) 0.020 Midday-napping No 246 25 -8.03 (-12.78, -3.29) 0.001 0.011 Yes 1317 139 -1.40 (-3.39, 0.59) 0.169 2 Pre-pregnancy <24 kg/m 1382 133 -2.47 (-4.48, -0.45) 0.017 0.569 BMI ≥24 kg/m2 181 31 -2.16 (-7.07, 2.75) 0.389 Parity Primiparous 1326 135 -1.92 (-3.96, 0.12) 0.065 0.307 Multiparous 237 29 -5.01 (-9.33, -0.70) 0.023 Fetal gender Boy 825 95 -5.29 (-7.89, -2.68) <0.001 0.002 Girl 738 69 1.03 (-1.48, 3.54) 0.422 Abortion history No 1034 99 -0.40 (-2.74, 1.94) 0.736 0.008 Yes 529 65 -5.59 (-8.59, -2.60) <0.001 All covariates are as follows: maternal age, education level, alcohol consumption, smoking, ethnicity, height, pre-pregnancy body mass index, gestational weight gain, gestational age, average personal income, parity, fetal gender and midday napping. Analyses that stratified by abortion history were adjusted for all covariates mentioned above, each of the else groups were adjusted for all covariates except itself. BMI, Body mass index; CI, confidence interval. a Sample size of the sleeping 8 to <9 h/day group in each subgroup. b Sample size of the sleeping <7 h/day group in each subgroup. c Change (95% CI) in mean birth length (mm, sleeping <7 h/day vs sleeping 8 to <9 h/day). d p-value for the effect in each subgroup. e p-value for interaction based on Wald test.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights: •
Shorter sleep duration in early pregnancy was associated with birth length.
•
The association was much stronger among mothers without midday napping
No evidence was found that sleep duration was associated with birth weight.
AC C
EP
TE D
M AN U
SC
•
RI PT
habit.