Objective
The objective of this study was to examine whether teen birth was independently associated with overweight and obesity in a US cohort.
Study Design
We examined whether teen birth is independently associated with overweight and obesity in a multiyear US cohort using the 2001-2010 National Health and Nutrition Examination Survey, a nationally representative cross-sectional survey of the US civilian, noninstitutionalized population. We performed multinomial logistic regression adjusting for survey cohort, age at survey, race, education, and parity. We included women 20-59 years old at the time of survey, with at least 1 live birth, not currently or recently pregnant (unweighted, n = 5220; weighted, n = 48.4 million). Our outcome measure was the effect of teen birth on subsequent overweight and obesity.
Results
In bivariate analyses, women with a teen birth were significantly more likely than women without a teen birth to be overweight (relative risk ratios [RRRs], 1.61; 95% confidence interval [CI], 1.37–1.90) or obese (RRR, 1.84; 95% CI, 1.56–2.16) at the time of the survey. In multivariate models, women with a teen birth remained significantly more likely to be overweight (adjusted RRR, 1.33; 95% CI, 1.10–1.62) or obese (adjusted RRR, 1.32; 95% CI, 1.09–1.61) than women without a teen birth.
Conclusion
For women in the United States, giving birth as a teen is associated with subsequent overweight/obese status later in life. To inform clinical and policy interventions with the goal to improve the long-term health of teenage mothers, future studies must examine modifiable physiological and sociomedical reasons for early child-bearing and later risk of obesity.
For Editors’ Commentary, see Contents
Despite decades of clinical and research efforts, obesity continues to be a prevalent, debilitating, and expensive public health problem. Rates of obesity in adults and children have not improved and in some groups continue to climb. Obesity is associated with serious physical, psychological, and social problems including cardiovascular disease, lower quality of life, and stigma.
Pregnancy is a strong risk factor for new or persistent obesity, with minority women having a higher incidence of parity-related obesity than white women. In a recent study of the National Longitudinal Survey of Youth, the 5 year incidence of obesity was 11.3 per 100 in parous women, compared with 4.5 per 100 in nulliparous women.
Primary prevention efforts are especially important in population-level obesity management because studies have shown that once obesity is established, it is difficult to reverse and continues through adulthood. Furthermore, obesity in adolescence is significantly associated with an increased risk of severe obesity in adulthood, with the prevalence of severe obesity increasing much faster than obesity.
Although multiple studies have focused on overweight and obese children and adolescents, one population that has not yet been studied as a high-risk group for obesity is women who have given birth as teenagers. Approximately 30% of teen women in the United States become pregnant and 20% give birth by age 20 years. Teenage mothers have both sociodemographic and physiological risk factors for obesity. Sociodemographic risks among women for obesity include black or Hispanic race/ethnicity, poverty, and low educational attainment.
Physiological risks among teenage mothers include greater gestational weight gain and greater postpartum weight retention than adults. Studies have also shown that after 28 weeks’ gestation, growing adolescents continue to accrue fat rather than mobilize fat stores like nongrowing adolescents and adults. Furthermore, a recent US longitudinal study showed that women who give birth during adolescence and young adulthood have substantially greater increments in overall and central adiposity than adolescents who do not give birth.
We are unaware of prior studies that have examined, at the individual or population level, whether a history of teen birth is a risk factor for obesity in later adult life. If teen birth is a risk factor, teen mothers would be a newly identified high-risk group for prevention of long term obesity.
The purpose of this study was to examine whether teen birth is independently associated with overweight and obesity in a US cohort, using the National Health and Nutrition Examination Survey (NHANES) from 2001 to 2010.
Materials and Methods
NHANES is a nationally representative cross-sectional survey designed to assess the health and nutritional status of the US civilian, noninstitutionalized population. NHANES became a continuous survey in 1999 and data are released in 2 year cycles. The NHANES survey includes a personal interview in the household and a detailed physical examination in a Mobile Examination Center (MEC). Additional data on the survey design, questionnaires, and laboratory methods are available online.
To examine the association between teen birth and overweight and obesity, data were analyzed from five 2-year cycles from 2001 to 2010. The overall examination survey response rates for each cycle ranged from 72% to 81%. During 2001-2010, 8956 women aged 20-59 years participated in the reproductive questionnaire and MEC examination.
This study sample was restricted to women aged 20-59 years at the time of the survey, whose measured weight and height at the time of participation in the MEC led to calculated body mass index (BMI) greater than 18.5 kg/m 2 , who had at least 1 child and who were not currently pregnant or recently pregnant (within 12 months). We limited the ages to 20-59 years at the time of the survey to mitigate cohort differences in health care and society that might have been present with inclusion of older women and also to avoid ages 60 years and beyond during which conditions such as cancer can be associated with weight loss. These criteria yielded a sample of 5256 women, of whom 5220 (99.3%) had complete data including BMI, age at first birth, current age, race, education, and parity.
For each participant, the age at first birth was abstracted from the dataset and used to determine whether the participant had a teen birth. Teen birth is defined in this study as having a live birth between age 13 years and age 19 years, inclusive. Other variables abstracted included survey cohort (2001-2002, 2003-2004, 2005-2006, 2007-2008, 2009-2010), age at time of survey, race/ethnicity (grouped as non-Hispanic white, non-Hispanic black, Mexican-American, and other, which includes multiracial), education (grouped less than ninth grade, ninth to 11th grade, high school graduate or general education degree (GED), some college, and college graduate), and parity (grouped as 1, 2, 3, 4, or ≥5 live births). These variables were included in our model because they have been shown to be strongly associated with both obesity and teen birth and need to be adjusted for to avoid confounding. To avoid overfitting or overadjusting our model, we did not include variables in the causal pathway of obesity such as diet and physical activity.
The outcome of interest was overweight (BMI 25.0-29.9 kg/m 2 ) vs normal weight (BMI 18.5-24.9 kg/m 2 ) and obese (BMI ≥30 kg/m 2 ) vs normal weight (BMI 18.5-24.9 kg/m 2 ) at the time of the survey. The exposure of interest was teen birth (yes/no). We performed descriptive statistics and then multinomial logistic regression to determine the bivariate association between BMI and teen birth. Subsequently we added available sociodemographic variables (current age, race/ethnicity, education) as well as the survey cohort variable to take into account BMI trends over time by cohort. Finally, we added the parity variable to assess independent associations with obesity.
To account for the complex, multistage probability survey design, analyses were conducted using STATA (version 12; Stata Corp, College Station, TX), adjusting for sample clustering and applying sampling weights to permit national inferences. Results are presented as unadjusted and adjusted relative risk ratios (aRRRs). Statistical significance was defined as an alpha of less than 0.05. All findings are presented using weighted data, except where otherwise indicated. Institutional review board approval was not required for the analysis of this publically available, deidentified data.
Results
The baseline characteristics in our sample of women (unweighted, n = 5220; weighted, n = 48.4 million) are presented in Table 1 . Women with a teen birth differed in several respects, as expected, from women without a teen birth. Women with a teen birth were more likely to be non-Hispanic black, or Mexican-American and more likely to report less than high school education than women without a teen birth. Women with a teen birth were also significantly more likely to have higher parity (2.70; 95% confidence interval [CI], 2.63–2.76) than women without a teen birth (2.12; 95% CI, 2.08–2.17).
| Characteristic | Non-teen birth, % (95% CI) | Teen birth, % (95% CI) | Student t test, P value | χ 2 , P value |
|---|---|---|---|---|
| n | 3422 | 1798 | ||
| Total n | 34.8 million | 13.6 million | ||
| Percent of sample | 65.6 | 34.4 | ||
| Mean age | 43.5 (43.1–43.9) | 40.7 (40.0–41.4) | < .001 | |
| Race/ethnicity | < .001 | |||
| White, non-Hispanic | 72.5 (69.3–75.5) | 54.0 (48.3–59.6) | ||
| Black, non-Hispanic | 9.6 (8.2–11.3) | 22.3 (18.9–26.1) | ||
| Mexican-American | 7.0 (5.8– 8.5) | 12.5 (10.1–15.6) | ||
| Other | 10.8 (8.1–12.8) | 11.2 (8.9–14.0) | ||
| Education | < .001 | |||
| Less than ninth grade | 3.3 (2.7–4.0) | 9.6 (8.2–11.2) | ||
| Ninth to 11th grade | 8.4 (7.3–9.7) | 24.7 (22.6–27.0) | ||
| HS grad or GED | 22.0 (20.1–24.0) | 32.1 (29.5–34.8) | ||
| Some college | 36.0 (33.9–38.2) | 29.0 (26.4–31.8) | ||
| College graduate and above | 30.3 (28.1–32.6) | 4.6 (3.3–6.3) | ||
| Parity | < .001 | |||
| 1 | 27.2 (25.4–29.1) | 14.2 (12.3–16.4) | ||
| 2 | 43.1 (41.0–45.2) | 34.0 (31.3–36.8) | ||
| 3 | 22.1 (20.5–23.8) | 29.4 (26.8–32.2) | ||
| 4 | 5.3 (4.6–6.3) | 12.6 (10.7–14.8) | ||
| ≥5 | 2.3 (1.6–3.2) | 9.8 (8.5–11.3) |
At the time of the survey, significantly more women with a teen birth were obese (44.2%; 95% CI, 41.5–47.0) compared with women without a teen birth (35.2%; 95% CI, 33.2–37.3; P < .001). Significantly fewer women with a teen birth were normal weight (26.1%; 95% CI, 23.3–29.0) compared with women without a teen birth (37.9%; 95% CI, 35.9–40.0; P < .001). There was no significant difference in the prevalence of overweight by teen birth status ( Figure ).
In an unadjusted multinomial logistic regression, women with a teen birth were significantly more likely than women without a teen birth to be overweight (relative risk ratios [RRRs], 1.61; 95% CI, 1.37–1.90; P < .001) or obese (RRR, 1.84; 95% CI, 1.56–2.16; P < .001) at the time of the survey ( Tables 2 and 3 ). In multivariate models, women with a teen birth remained significantly more likely than women without a teen birth to be overweight (adjusted RRR, 1.33; 95% CI, 1.10–1.62; P = .004) or obese (aRRR, 1.32; 95% CI, 1.09–1.61; P = .005). When the BMI variable was further stratified into grade 1 (BMI of 30 to <35 kg/m 2 ), grade 2 (BMI of 35 to <40 kg/m 2 ), and grade 3 (BMI of ≥40 kg/m 2 ), the effect of teen birth was generally unchanged (data not shown).
| Variable | Unadjusted, RRR (95% CI, P value) | Adjusted for sociodemographic factors, RRR (95% CI, P value) | Adjusted for sociodemographic factors and parity, RRR (95% CI, P value) |
|---|---|---|---|
| Overweight vs normal weight | |||
| Teen birth | 1.61 (1.37–1.90, < .001) a | 1.34 (1.11–1.60, .002) a | 1.33 (1.10–1.62, .004) a |
| Current age | 1.02 (1.01–1.03, < .001) a | 1.02 (1.01–1.03, < .001) a | |
| Race/ethnicity | |||
| White, non-Hispanic | 1.00 Referent | 1.00 Referent | |
| Black, non-Hispanic | 2.27 (1.72–2.99, < .001) a | 2.30 (1.75–3.04, < .001) a | |
| Mexican-American | 1.93 (1.53–2.44, < .001) a | 1.95 (1.54–2.46, < .001) a | |
| Other | 1.27 (0.96–1.69, .095) | 1.28 (0.97–1.70, .085) | |
| Education | |||
| Less than 9th grade | 1.00 Referent | 1.00 Referent | |
| Ninth to 11th grade | 0.76 (0.51–1.13, .166) | 0.75 (0.50–1.13, .166) | |
| HS grad or GED | 0.75 (0.51–1.10, .134) | 0.75 (0.51–1.10, .136) | |
| Some college | 0.81 (0.56–1.18, .278) | 0.82 (0.56–1.19, .291) | |
| College graduate and above | 0.49 (0.33–0.73, .001) a | 0.49 (0.33–0.72, .001) a | |
| Parity | |||
| 1 | 1.00 Referent | ||
| 2 | 1.02 (0.82–1.27, .844) | ||
| 3 | 1.17 (0.92–1.49, .204) | ||
| 4 | 0.80 (0.52–1.24, .319) | ||
| ≥5 | 1.12 (0.71–1.76, .618) | ||
| Variable | Unadjusted, RRR (95% CI, P value) | Adjusted for sociodemographic factors, RRR (95% CI, P value) | Adjusted for sociodemographic factors and parity, RRR (95% CI, P value) |
|---|---|---|---|
| Obese vs normal weight | |||
| Teen birth | 1.84 (1.56–2.16, < .001) a | 1.38 (1.14–1.66, .001) a | 1.32 (1.09–1.61, .005) a |
| Current age | 1.02 (1.02–1.03, < .001) a | 1.02 (1.01–1.03, < .001) a | |
| Race/ethnicity | |||
| White, non-Hispanic | 1.00 Referent | 1.00 Referent | |
| Black, non-Hispanic | 3.30 (2.59–4.20, < .001) a | 3.30 (2.59–4.21, < .001) a | |
| Mexican-American | 1.85 (1.50–2.27, < .001) a | 1.81 (1.47–2.23, < .001) a | |
| Other | 0.84 (0.62–1.12, .232) | 0.83 (0.62–1.11, .213) | |
| Education | |||
| Less than 9th grade | 1.00 Referent | 1.00 Referent | |
| Ninth to 11th grade | 0.90 (0.61–1.35, .616) | 0.93 (0.63–1.38, .716) | |
| HS graduate or GED | 0.96 (0.64–1.44, .845) | 1.01 (0.68–1.49, .964) | |
| Some college | 0.91 (0.62–1.33, .619) | 0.95 (0.66–1.38, .792) | |
| College graduate and above | 0.42 (0.28–0.64, < .001) a | 0.44 (0.29–0.66, < .001) a | |
| Parity | |||
| 1 | 1.00 Referent | ||
| 2 | 1.07 (0.86–1.32, .540) | ||
| 3 | 1.27 (1.04–1.54, .019) a | ||
| 4 | 1.12 (0.80–1.56, .502) | ||
| ≥5 | 1.38 (0.94–2.01, .096) | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
