Gestational Weight Gain and Preterm Birth: Disparities in Adolescent Pregnancies

ORIGINAL ARTICLE Gestational weight gain and preterm birth: disparities in adolescent pregnancies

CL Woolfolk1, LM Harper2, L Flick1, K Mathews3 and JJ Chang1

OBJECTIVE: To examine racial differences in the association between gestational weight gain and preterm birth subtypes among adolescents. STUDY DESIGN: We conducted a retrospective cohort study of 211 403 adolescents using 2012 United States natality data. The outcome was preterm birth and the primary exposure was gestational weight gain. Multinomial logistic regression analyses were used to estimate adjusted odds ratios, stratiﬁed by race and body mass index (BMI). RESULTS: Black and White mothers who gained below the recommendations had increased risks for spontaneous preterm birth in all BMI categories, except obese. All Hispanic mothers who gained below the recommendations had increased risks of spontaneous preterm birth. White normal and overweight mothers and Black, Hispanic and Other normal weight mothers who exceeded the recommendations had decreased risks of spontaneous preterm birth. CONCLUSION: The effect of gestational weight gain on spontaneous and medically indicated preterm birth is modiﬁed by race and BMI. Journal of Perinatology (2016) 36, 1055–1060; doi:10.1038/jp.2016.149; published online 8 September 2016

INTRODUCTION effect modifier in the relationship between gestational weight Preterm birth (PTB), the birth of an infant before 37 weeks gain and PTB in adolescent mothers, these studies use different gestation, is a major contributor to infant morbidity and mortality. measures of gestational weight gain and reference groups, have In 2014, 1 out of 10 infants were born preterm in the United limited generalizability because they usually take place in one 11 States.1 PTB rates among African American and American Indian/ state, and fail to examine PTB by subtype. The objective of this Alaskan Native women are significantly higher compared with study was to examine racial and BMI differences in the association White women, while Asian/Pacific Islanders have significantly between gestational weight gain and spontaneous and medically lower PTB rates.2–4 Inadequate gestational weight gain, maternal indicated PTB among adolescents, using the current IOM pre-pregnancy weight and race have been identified as risk factors guidelines. for PTB.5–9 There are two subtypes of PTB, spontaneous and medically indicated, which are thought to result from different causes.6 Adolescent pregnancy is characterized by increased risks MATERIALS AND METHODS – of many adverse birth outcomes, including PTB.10 12 Because of We conducted a population-based retrospective cohort of 2012 United the additional risks associated with adolescent pregnancy, it is States (US) natality data, which includes birth certificate data of all live important to study the impact of these factors on PTB subtypes births delivered in the US. The 2012 natality data, obtained from the among adolescent mothers. National Center for Health Statistics, includes all registered births occurring fi in the US. The study used de-identified data and was exempt from Gestational weight gain is an important modi able factor that o affects PTB.6,13 In 2009, the Institute of Medicine (IOM) revised the institutional board review. Our study included all nulliparous women 20 years old, who delivered live, singleton infants 20 to 44 weeks gestation in gestational weight gain guidelines that were designed to reduce 2012. We excluded multiple gestations, congenital anomalies and subjects infant morbidity and mortality. The IOM guidelines, based on pre- with missing data on BMI. pregnancy body mass index (BMI), recommend ideal gestational Maternal BMI, calculated as weight per height (kg m − 2), was weight gain ranges to improve birth outcomes. The current categorized as underweight (BMI o18.5 kg m − 2), normal weight guidelines are based on adult BMI categories, however, adult BMI (BMI 18.5-24.9 kg m − 2), overweight (BMI 25.0-29.9 kg m − 2) and obese − categories are not always applicable to adolescents. The IOM (BMI ≥ 30 kg m 2) based on self-reported pre-pregnancy weight and recommends that adolescents use adult BMI categories until height. The self-reported maternal races were categorized as: White, there is more research to determine if separate categories are non-Hispanic; Black, non-Hispanic; Hispanic; and Other, non-Hispanic. The 14 Other, non-Hispanic maternal race category includes American Indian/ necessary. Alaskan Native, Asian, and Pacific Islander. Given the racial disparities in PTB and the increased risk of PTB The primary exposure was gestational weight gain, which was obtained among adolescents, it is important to examine whether maternal from the birth certificate. The IOM recommendations were used to classify race and BMI modify the relationship between gestational weight gestational weight gain into three categories: below, met, and exceeded gain and PTB. Although a few studies have evaluated BMI as an IOM recommendations. The IOM recommends that underweight women

1Department of Epidemiology, Saint Louis University College for Public Health and Social Justice, St. Louis, MO, USA; 2Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology. University of Alabama at Birmingham, Birmingham, AL, USA and 3Department of Obstetrics, Gynecology and Women’s Health, School of Medicine, Saint Louis University in St. Louis, St. Louis, MO, USA. Correspondence: Dr JJ Chang, Department of Epidemiology, Saint Louis University College for Public Health and Social Justice, 3545 Lafayette Avenue, Room 476, St. Louis, MO 63104, USA. E-mail: [email protected] Received 15 December 2015; revised 4 July 2016; accepted 11 July 2016; published online 8 September 2016 Gestational weight gain and preterm birth CL Woolfolk et al 1056 gain 1 to 1.3 pounds per week, normal weight women gain 0.8 to 1 and the Kotelchuck index, the odds of spontaneous and medically pounds, overweight women gain 0.5 to 0.7 pounds, and obese women indicated PTB among underweight adolescents were significantly gain 0.4 to 0.6 pounds during the second and third trimesters. A weight increased among those below the IOM recommendations, gain of 1.1 to 4.4 pounds in the first trimester is assumed for each BMI.14 To compared with those who met the IOM recommendations (aOR determine the gestational weight gain, we calculated the normal weight gain range for each week using the recommended weekly weight gain for 1.87, 95% CI: 1.13, 3.07 for medically indicated and aOR 1.52, 95% each BMI. The number of gestational weeks after the first trimester for each CI: 1.25, 1.85 for spontaneous PTB). Among normal weight subject was calculated by subtracting 13 from the gestational age. adolescents, the odds of spontaneous PTB were significantly The lower limit of the normal weight gain range was calculated by adding increased among those below the IOM recommendations (aOR the lower limit of the assumed first trimester weight gain to the product fi 1.50, 95% CI: 1.33, 1.68 for spontaneous PTB); the odds of of the number of gestational weeks after the rst trimester and the lower fi limit of the recommended weight gain for that BMI. The upper limit of spontaneous PTB were signi cantly decreased among those who the normal weight gain range was calculated by adding the upper limit exceeded the IOM recommendations (aOR 0.81, 95% CI: 0.74, of the assumed first trimester weight gain to the product of the number of 0.89). Among overweight adolescents, the odds of spontaneous gestational weeks after the first trimester and the upper limit of the PTB were significantly increased among those below the IOM recommended weight gain for that BMI. Each subject’s total gestational recommendations (aOR 1.31, 95% CI: 1.03, 1.67); the odds of weight gain was then compared to the normal weight gain range for their spontaneous PTB were significantly decreased among those who BMI and gestational age. Gestational weight gain within the normal weight gain range was categorized as ‘met IOM guidelines,’ less than the normal exceeded the IOM recommendations (aOR 0.70, 95% CI: 0.58, weight gain range was categorized as ‘below IOM guidelines,’ and above 0.85). Among obese adolescents, the odds of spontaneous PTB the normal weight gain range was categorized as ‘exceeded IOM were significantly decreased among those below the IOM recommendations.’ recommendations (aOR 0.77, 95% CI: 0.61, 0.96). The outcome of interest was PTB, which included births before 37 weeks gestation. PTB was calculated using the clinical estimate of gestational age. PTB was further categorized into spontaneous and medically indicated Black maternal race PTB. Medically indicated PTB consists of all PTB where a maternal medical aORs for spontaneous and medically indicated PTB among those condition or labor induction was present. Maternal medical conditions reporting Black maternal race are presented in Table 3. After include: pre-pregnancy diabetes, gestational diabetes, pre-pregnancy adjusting for confounders, the odds of spontaneous and medically hypertension, gestational hypertension and eclampsia. Spontaneous PTB fi consists of all PTB without labor induction and maternal medical indicated PTB among underweight adolescents were signi cantly conditions (diabetes, hypertension or eclampsia).15 increased among those below the IOM recommendations (aOR The Pearson Χ2 test was used to assess differences in sample 2.47, 95% CI: 1.18, 5.15 for medically indicated and aOR 1.50, 95% characteristics by gestational weight gain. Effect modification by maternal CI: 1.12, 1.99 for spontaneous PTB). Among normal weight race and BMI was assessed through logistic regression using a three-way adolescents, the odds of spontaneous and medically indicated interaction term between race, BMI and gestational weight gain. A P-value PTB were significantly increased among those below the IOM fi Χ2 of 0.05 was used to assess the signi cance of the Wald statistic. recommendations (aOR 1.47, 95% CI: 1.09, 1.99 for medically We adjusted for the following confounders: maternal age (⩽15, 415), women, infants and children (WIC) status (yes, no), smoking during indicated and aOR 1.37, 95% CI: 1.20, 1.56 for spontaneous PTB); pregnancy (yes, no) and Kotelchuck prenatal care index (adequate plus, the odds of spontaneous PTB were significantly decreased among adequate, intermediate and inadequate). A multinomial logistic regression those who exceeded the IOM recommendations (aOR 0.80, analysis was used to estimate crude and adjusted odds ratios (aOR) and 95% CI: 0.71, 0.90). Among overweight adolescents, the odds of 95% confidence intervals (CI) for the relationship between gestational spontaneous PTB were significantly increased among those below fi weight gain, spontaneous PTB and medically indicated PTB, strati ed by the IOM recommendations (aOR 1.34, 95% CI: 1.03, 1.75). race and BMI. An alpha level of 0.05 was used to assess significance. All statistical analyses were performed using SAS version 9.4. Hispanic maternal race aORs for spontaneous and medically indicated PTB among those RESULTS reporting Hispanic maternal race are presented in Table 4. After A total of 3 960 796 births occurred in the US in 2012, 309 513 adjusting for confounders, the odds of spontaneous PTB among were among adolescents. After inclusion and exclusion criteria, underweight adolescents were significantly increased among 211 403 births remained. Overall, 39 097 (18.49%) were below, those below the IOM recommendations (aOR 1.33, 95% CI: 1.05, 44 871 (21.23%) met and 127 435 (60.28%) exceeded the IOM 1.69). Among normal weight adolescents, the odds of sponta- recommendations. Sample characteristics are presented in neous PTB were significantly increased among those below the Table 1. The proportions of spontaneous and medically indicated IOM recommendations (aOR 1.20, 95% CI: 1.08, 1.35); the odds of PTB were highest among mothers who gained below the IOM spontaneous PTB were significantly decreased among those who recommendations. Mothers gaining below the IOM recommenda- exceeded the IOM recommendations (aOR 0.87, 95% CI: 0.79, tions were also more likely to be Black or Other race, 15 years or 0.96). Among those overweight adolescents, the odds of younger, underweight, and receive inadequate or intermediate fi prenatal care. Mothers exceeding the IOM recommendations were spontaneous PTB were signi cantly increased among those below more likely to be White, overweight or obese, receive adequate or the IOM recommendations (aOR 1.40, 95% CI: 1.12, 1.75). Among fi adequate plus prenatal care, smoke and participate in WIC. There obese adolescents, the odds of spontaneous PTB were signi - were significant interactions between maternal race, BMI and cantly increased among those below the IOM recommendations gestational weight gain (P = 0.0019). Therefore, maternal race and (aOR 1.50, 95% CI: 1.12, 2.02). BMI are effect modifiers in the relationship between gestational weight gain and PTB. Logistic regression analyses stratified by Other maternal race maternal race and BMI are presented in Tables 2, 3, 4 and 5. aORs for spontaneous and medically indicated PTB among those reporting Other maternal race are presented in Table 5. After White maternal race adjusting for confounders, the odds of spontaneous PTB among aOR for spontaneous and medically indicated PTB among those normal weight adolescents were significantly decreased among reporting White maternal race are presented in Table 2. After those who exceeded the IOM recommendations (aOR 0.62, 95% adjusting for maternal age, WIC status, smoking during pregnancy CI: 0.46, 0.83).

Table 1. Characteristics of the study population by gestational weight gain category, N = 211 403

Overall N (%) Below IOM Met IOM Exceeded IOM P-value recommendations recommendations recommendations (N = 39 097) N (%) (N = 44,871) N (%) (N = 127 435) N (%)

Maternal race o0.0001 White 85 858 (40.6) 12 946 (33.11) 16 747 (37.32) 56 165 (44.07) Black 46 086 (21.8) 10 393 (26.58) 9779 (21.79) 25 914 (20.34) Hispanic 72421 (34.3) 14 303 (36.58) 16 832 (37.51) 41 286 (32.40) Other 7038 (3.3) 1455 (3.72) 1513 (3.37) 4070 (3.19) Maternal BMI o0.0001 Underweight 16 811 (7.9) 5071 (12.97) 5784 (12.89) 5956 (4.67) Normal weight 118 366 (56.0) 22 725 (58.12) 27 524 (61.34) 68 117 (53.45) Overweight 45 426 (21.5) 5991 (15.32) 6676 (14.88) 32 759 (25.71) Obese 30 800 (14.6) 5310 (13.58) 4887 (10.89) 20 603 (16.17) Maternal age o0.0001 ⩽ 15 11 577 (5.5) 2552 (6.53) 2616 (5.83) 6409 (5.03) 415 199 826 (94.5) 36 545 (93.47) 42 255 (94.17) 12 1026 (94.97) Kotelchuck prenatal o0.0001 care adequacy Inadequate 60 517 (28.6) 13 854 (35.43) 13 342 (29.73) 33 321 (26.15) Intermediate 19 082 (9.0) 3717 (9.51) 4247 (9.46) 11 118 (8.72) Adequate 70 058 (33.1) 11 241 (28.75) 14 832 (33.05) 43 985 (34.52) Adequate plus 61 746 (29.2) 10 285 (26.31) 12 450 (27.75) 39 011 (30.61) Smoking status o0.0001 No 189 806 (89.8) 35 727 (91.38) 40 809 (90.95) 113 270 (88.88) Yes 21 597 (10.2) 3370 (8.62) 4062 (9.05) 14 165 (11.12) WIC status o0.0001 No 40 634 (19.2) 8013 (20.50) 8834 (19.69) 23 787 (18.67) Yes 170 769 (80.8) 31 084 (79.50) 36 037 (80.31) 103 648 (81.33) PTB o0.0001 Medically 2798 (1.3) 541 (1.38) 516 (1.15) 1749 (1.37) indicated PTB Spontaneous PTB 14 831 (7.0) 3767 (9.64) 3275 (7.30) 7781 (6.11) Term birth 19 3774 (91.7) 34 789 (88.98) 41 080 (91.55) 117 905 (92.52) Abbreviations: BMI, body mass index; IOM, Institute of Medicine; PTB, preterm birth; WIC, women, infants and children.

Table 2. Odds ratiosa by BMI and gestational weight gain categories for white adolescents

Medically indicated PTB Spontaneous PTB

BMI Gestational weight gainb N (%) aORa 95% CI N (%) aORa 95% CI

Underweight Below IOM recommendations 37 (1.9) 1.87 1.13, 3.07 248 (12.6) 1.52 1.25, 1.85 Exceeded IOM recommendations 47 (1.4) 1.27 0.79, 2.05 263 (7.9) 0.87 0.72, 1.05 Normal weight Below IOM recommendations 97 (1.4) 1.1 0.82, 1.46 713 (10.2) 1.5 1.33, 1.68 Exceeded IOM recommendations 401 (1.3) 1.03 0.83, 1.27 1814 (5.6) 0.81 0.74, 0.89 Overweight Below IOM recommendations 24 (1.2) 0.97 0.54, 1.76 175 (9.1) 1.31 1.03, 1.67 Exceeded IOM recommendations 179 (1.3) 1.04 0.67, 1.60 715 (5.3) 0.7 0.58, 0.85 Obese Below IOM recommendations 22 (1.1) 0.79 0.43, 1.47 143 (7.0) 1.03 0.78, 1.34 Exceeded IOM recommendations 137 (1.6) 1.14 0.72, 1.82 449 (5.3) 0.77 0.61, 0.96 Abbreviations: aOR, adjusted odds ratio; BMI, body mass index; CI, conﬁdence interval; IOM, Institute of Medicine; N, number of preterm subjects; PTB, preterm birth; WIC, women, infants and children. aAdjusted for maternal age, WIC status, smoking status and the Kotelchuck prenatal care index. bReference group = met IOM recommendations.

DISCUSSION of both PTB subtypes. However, in Hispanic adolescents, there was This study found signiﬁcant relationships between gestational only an increased risk of spontaneous PTB among those who were weight gain relative to the current IOM recommendations and the underweight, while no signiﬁcant association was observed in risk of PTB and this relationship varies by maternal race and BMI. underweight Other race mothers. This study also distinguished between spontaneous and medically Dietz et al.16 found a strong association between very low indicated PTB and lends further support to the notion that causes gestational weight gain (o0.12 kg per week) and PTB among of PTB are heterogeneous. Some results of this analysis are underweight women, although they used different measures of consistent with other studies. Among Black and White under- gestational weight gain and included both adult and adolescent weight adolescents, we found that gestational weight gain below mothers in the study sample. In addition, Salihu et al.6 found an the IOM recommendations was associated with an increased risk increased risk of spontaneous PTB with very low gestational

Table 3. Odds ratiosa by BMI and gestational weight gain categories for black adolescents

Medically indicated PTB Spontaneous PTB

BMI Gestational weight gainb N (%) aORa 95% CI N (%) aORa 95% CI

Underweight Below IOM recommendations 27 (2.4) 2.47 1.18, 5.15 163 (14.4) 1.5 1.12, 1.99 Exceeded IOM recommendations 21 (2.4) 2.14 0.99, 4.61 82 (9.3) 0.83 0.59, 1.16 Normal weight Below IOM recommendations 114 (1.9) 1.47 1.09, 1.99 677 (11.5) 1.37 1.20, 1.56 Exceeded IOM recommendations 227 (1.8) 1.3 0.99, 1.69 905 (7.0) 0.8 0.71, 0.90 Overweight Below IOM recommendations 33 (1.9) 1.13 0.65, 1.96 173 (10.2) 1.34 1.03, 1.75 Exceeded IOM recommendations 107 (1.5) 0.86 0.54, 1.35 469 (6.6) 0.89 0.71, 1.12 Obese Below IOM recommendations 21 (1.2) 0.83 0.41, 1.66 150 (9.0) 1.32 0.98, 1.77 Exceeded IOM recommendations 84 (1.6) 1.17 0.67, 2.05 355 (7.0) 0.98 0.75, 1.27 Abbreviations: aOR, adjusted odds ratio; BMI, body mass index; CI, conﬁdence interval; IOM, Institute of Medicine; N, number of preterm subjects; PTB, preterm birth; WIC, women, infants and children. aAdjusted for maternal age, WIC status, smoking status and the Kotelchuck prenatal care index. bReference group = met IOM recommendations.

Table 4. Odds ratiosa by BMI and gestational weight gain categories for hispanic adolescents

Medically indicated PTB Spontaneous PTB

BMI Gestational weight gainb N (%) aORa 95% CI N (%) aORa 95% CI

Underweight Below IOM recommendations 18 (1.0) 0.85 0.45, 1.62 177 (10.0) 1.33 1.05, 1.69 Exceeded IOM recommendations 20 (1.3) 1.1 0.59, 2.04 119 (7.6) 0.96 0.74, 1.25 Normal weight Below IOM recommendations 85 (0.9) 1.11 0.82, 1.50 716 (8.0) 1.2 1.08, 1.35 Exceeded IOM recommendations 248 (1.1) 1.27 0.99, 1.62 1295 (5.9) 0.87 0.79, 0.96 Overweight Below IOM recommendations 251 (1.2) 1.27 0.73, 2.23 188 (8.8) 1.4 1.12, 1.75 Exceeded IOM recommendations 134 (1.2) 1.23 0.81, 1.89 645 (5.7) 0.91 0.76, 1.08 Obese Below IOM recommendations 23 (1.6) 1.22 0.67, 2.24 114 (8.1) 1.5 1.12, 2.02 Exceeded IOM recommendations 98 (1.5) 1.16 0.72, 1.85 417 (6.5) 1.2 0.94, 1.53 Abbreviations: aOR, adjusted odds ratio; BMI, body mass index; CI, conﬁdence interval; IOM, Institute of Medicine; PTB, preterm birth; N, number of preterm subjects; WIC, women, infants and children. aAdjusted for maternal age, WIC status, smoking status and the Kotelchuck prenatal care index. bReference group = met IOM recommendations.

Table 5. Odds ratiosa by BMI and gestational weight gain categories for other adolescents

Medically indicated PTB Spontaneous PTB

BMI Gestational weight gainb N (%) aORa 95% CI N (%) aORa 95% CI

Underweight Below IOM recommendations 3 (1.6) NA NA 20 (10.3) 1.04 0.48, 2.23 Exceeded IOM recommendations 2 (1.1) NA NA 17 (9.1) 1.05 0.49, 2.25 Normal weight Below IOM recommendations 7 (0.9) 0.78 0.29, 2.05 83 (10.1) 0.99 0.71, 1.39 Exceeded IOM recommendations 26 (1.2) 1.01 0.48, 2.12 136 (6.1) 0.62 0.46, 0.83 Overweight Below IOM recommendations 3 (1.3) 0.97 0.19, 4.92 19 (8.0) 1.08 0.52, 2.25 Exceeded IOM recommendations 11 (1.1) 0.67 0.18, 2.50 64 (6.2) 0.74 0.41, 1.34 Obese Below IOM recommendations 0 (0.0) NA NA 10 (5.0) 1.06 0.38, 2.95 Exceeded IOM recommendations 3 (0.5) 0.25 0.05, 1.17 40 (6.4) 1.16 0.50, 2.69 Abbreviations: aOR, adjusted odds ratio; BMI, body mass index; CI, conﬁdence interval; IOM, Institute of Medicine; N, number of preterm subjects; NA, not applicable; PTB, preterm birth; WIC, women, infants and children. aAdjusted for maternal age, WIC status, smoking status and the Kotelchuck prenatal care index. bReference group = met IOM recommendations.

weight gain, regardless of BMI in an adolescent population. Harper et al.,11 examined the current IOM guidelines in adolescent This analysis found that weight gain below the IOM recommenda- mothers and found an increased risk of PTB among those below tions was generally associated with an increased risk of the IOM recommendations, regardless of BMI. However, Harper spontaneous PTB in Black, White and Hispanic mothers. However, et al.11 did not examine PTB by subtype and was restricted to in obese mothers, the association only remained significant in White and Black adolescent mothers. Hispanic mothers. The risk of medically indicated PTB was not The effect of excessive gestational weight gain on PTB is not significantly increased among obese and overweight mothers clear. Although one previous study found no significant associa- who gained below the IOM recommendations. In contrast, tion between excessive weight gain and PTB,6 other studies found

Journal of Perinatology (2016), 1055 – 1060 © 2016 Nature America, Inc., part of Springer Nature. Gestational weight gain and preterm birth CL Woolfolk et al 1059 a reduced risk of PTB among mothers with excessive weight The current IOM guidelines are based on adult BMI categories; gain.9,11,16 Harper et al.11 found that among White and Black because BMI in adolescence varies by age, adolescent mothers adolescent mothers, exceeding the IOM recommendations was may be misclassified as a result. However, obstetricians usually protective against all PTB, regardless of BMI. However, the authors counsel adolescent patients using adult BMI categories. Another noted that because the IOM categories are based on adult BMI limitation is that the Other maternal race category included categories, it is possible that adolescents may have been American Indian, Alaskan Native, Asian and Pacific Islander races. misclassified into lower BMI categories, therefore the protective Differences in PTB may exist between these races, however, we effect observed against PTB may not be accurate. On the contrary, were not able to evaluate them separately due to sample size our analysis found that exceeding the IOM recommendations was restrictions. The Other maternal race category also had a small not always protective against spontaneous and medically number of medically indicated PTB among obese mothers. indicated PTB. In fact, there was no significant association Therefore, we were unable to obtain stable risk estimates within between exceeding the IOM recommendations and medically this category. The underlying causes of extreme PTB may differ indicated PTB; exceeding the IOM recommendations was only from very PTB or late PTB. However, due to the small number of protective against spontaneous PTB in some instances. Based on cases after stratifying by maternal BMI and race, we were unable our analysis, the effect of excessive weight gain on PTB varies by to examine PTB as a categorical variable that distinguishes race, BMI and PTB subtype. extreme, very and late PTB. Smoking status is often self-reported Gestational weight gain is a reflection of the mother’s and may be under-reported because smoking during pregnancy is nutritional status during pregnancy, while BMI is a reflection of not a socially acceptable behavior. Residual confounding may also the mother’s nutritional status before pregnancy; both factors 17 be present because the data set lacked information on a number play a role in PTB. Inadequate gestational weight gain was not of important potential confounders, including drug and alcohol fi always associated with signi cantly increased risks of the PTB use, infection, socioeconomic status, stress and lack of social subtypes in overweight and obese mothers; this suggests support, all of which can negatively affect BMI and gestational that overweight and obese mothers may have access to stored weight gain. To account for residual confounding due to the lack fat that is beneficial to mothers who fail to meet gestational 16 of data on socioeconomic status, WIC status was used as proxy for weight gain recommendations. In addition, the study results socioeconomic status. suggest that overweight and obese adolescents may need to be Despite the limitations, this analysis adds to the body of regarded as a high risk group and placed under close medical literature on gestational weight gain and PTB among adolescents. surveillance during prenatal care to help prevent adverse birth To our knowledge, no other studies have examined whether outcomes. maternal race/ethnicity modifies the effect of gestational weight The majority of the adolescents in this analysis exceeded the gain on PTB among adolescent mothers by pre-pregnancy BMI. IOM gestational weight gain recommendations. Although protec- The effect of gestational weight gain on spontaneous and tive against PTB subtypes in some instances, excessive gestational medically indicated PTB is modified by maternal race and pre- weight gain coupled with high BMI can increase the risk of pregnancy BMI. Clinicians should encourage adolescent mothers postpartum weight retention, which can ultimately lead to to meet the IOM gestational weight gain recommendations and increased adolescent obesity rates. Furthermore, Harper et al.11 closely monitor gestational weight gain during pregnancy. Failure found increased risks of large for gestational age, cesarean to meet the IOM recommendations is generally the most harmful delivery and preeclampsia among adolescents who exceeded the among underweight adolescents. Although exceeding the IOM IOM recommendations. Therefore, it is important to examine other recommendations is protective of spontaneous PTB in some outcomes associated with exceeding the IOM gestational weight instances, other maternal and perinatal outcomes should be gain recommendations. This study has many strengths. The large sample size allowed examined before clinicians consider advising adolescent mothers stratification by a number of maternal race/ethnicity subgroups to exceed the IOM recommendations. Future studies should and maternal pre-pregnancy BMI simultaneously and the examine maternal outcomes associated with excessive weight evaluation of a variety of confounders. The sample was nationally gain in adolescents, examine maternal nutrition and likelihood of representative, which increases the external validity of the study. smoking in relation to gestational weight gain and PTB, and The findings from this study should also be interpreted with examine the association between patterns of gestational weight caution due to some limitations. First, selection bias may be gain and PTB. present because ~ 16% of the sample was excluded due to missing or unknown data on BMI. We conducted sensitivity CONFLICT OF INTEREST analyses comparing subjects with known and missing BMI, and fl observed statistically significant differences between those with The authors declare no con ict of interest. known BMIs and those with missing BMI. Mothers with missing BMI information were more likely to be non-smokers, o15 years REFERENCES of age, Black or Other maternal race, gain above the IOM 1 Centers for Disease Control and Prevention. Preterm Birth [Internet]. Georgia: recommendations, receive inadequate prenatal care, and to not National Center for Chronic Disease Prevention and Health Promotion; receive WIC. In addition, the exposure, gestational weight gain, 4 December 2015 (cited on 15 December 2015). Available from http://www.cdc. was calculated based on the total weight gain during pregnancy. gov/reproductivehealth/maternalinfanthealth/pretermbirth.htm. We lacked information on the rate of weight gain and when the 2 MacDorman MF. Race and ethnic disparities in fetal mortality, preterm birth, and weight gain occurred. Studies have shown that patterns of infant mortality in the United States: an overview. Semin Perinatol 2011; 35: weight gain are important,9,11,16 however, we were only able to 200–208. evaluate the total gestational weight gain. Finally, pre-pregnancy 3 Hwang M, Shrestha A, Yazzie S, Jackson ML. Preterm birth among American weight and maternal race are self-reported and are subject to bias. Indian/Alaskan natives in Washington and Montana: comparison with non- 17 – Although we attempted to separate medically indicated Hispanic Whites. Matern Child Health J 2013; :1908 1912. 4 Wise LA, Palmer JR, Heffner LJ, Rosenberg L. Prepregnancy body size, gestational and spontaneous PTB, due to the retrospective nature of the weight gain, and risk of preterm birth in African-American women. Epidemiology study and limitations of the data available some births may have 2010; 21:243–252. been misclassified. However, the misclassification of maternal 5 Rudra CB, Frederick IO, Williams MA. Pre-pregnancy body mass index and weight pre-pregnancy BMI and PTB are likely non-differential which gain during pregnancy in relation to preterm delivery subtypes. Acta Obstet would bias the point estimates toward the null value. Gynecol Scand 2008; 87:510–517.

© 2016 Nature America, Inc., part of Springer Nature. Journal of Perinatology (2016), 1055 – 1060 Gestational weight gain and preterm birth CL Woolfolk et al 1060 6 Salihu HM, Luke S, Alio AP, Deutsch A, Marty PJ. The impact of obesity on 13 DeVader SR, Neeley HL, Myles TD, Leet TL. Evaluation of gestational weight gain spontaneous and medically indicated preterm birth among adolescent mothers. guidelines for women with normal prepregnancy body mass index. Obstet Arch Gynecol Obstet 2010; 282:127–134. Gynecol 2007; 110:745–751. 7 Cnattingius S, Villamor E, Johansson S, Bonamy AKE, Persson M, Wilkstrom AK 14 Institute of Medicine. Weight gain during pregnancy: reexamining the et al. Maternal obesity and risk of preterm delivery. JAMA 2013; 309: 2362–2370. guidelines [Internet]. District of Columbia: The National Academies Press; 2009 8 Nohr EA, Bech BH, Vaeth M, Rasmussen KM, Henriksen TB, Olsen J. Obesity, (cited on 15 Dec 2015). Available from https://iom.nationalacademies.org/ gestational weight gain and preterm birth: a study within the Danish National ~ /media/Files/Report%20Files/2009/Weight-Gain-During-Pregnancy-Reexamin Birth Cohort. Paediatr Perinat Epidemiol 2007; 21:5–14. ing-the-Guidelines/Report%20Brief%20-%20Weight%20Gain%20During%20Preg 9 Schieve LA, Cogswell ME, Scanlon KS, Perry G, Ferre C, Blackmore-Prince C et al. nancy.pdf. Prepregnancy body mass index and pregnancy weight gain: associations with 15 Reddy UM, Ko CW, Raju TN, Willinger M. Delivery indications at late-preterm preterm delivery. Obstet Gynecol 2000; 96:194–200. gestations and infant mortality rates in the United States. Pediatrics 2009; 124: 10 Haeri S, Guichard I, Baker AM, Saddlemire S, Boggess KA. The effect of teenage 234–240. maternal obesity on perinatal outcomes. Obstet Gynecol 2009; 113:300–304. 16 Dietz PM, Callaghan WM, Cogswell ME, Morrow B, Ferre C, Schieve LA. 11 Harper LM, Chang JJ, Macones GA. Adolescent pregnancy and gestational weight Combined effects of prepregnancy body mass index and weight gain gain: do the Institute of Medicine recommendations apply? Am J Obstet Gynecol during pregnancy on the risk of preterm delivery. Epidemiology 2006; 17: 2011; 205: 140.e141–e148. 170–177. 12 Nielsen JN, O'Brien KO, Witter FR, Chang SC, Mancini J, Nathanson MS et al. High 17 Siega-Riz AM, Adair LS, Hobel CJ. Institute of Medicine maternal weight gain gestational weight gain does not improve birth weight in a cohort of African recommendations and pregnancy outcome in a predominantly Hispanic popu- American adolescents. Am J Clin Nutr 2006; 84:183–189. lation. Obstet Gynecol 1994; 84:565–573.