Adolescent Risk Factors for Adult Pre-Pregnancy Obesity and High Gestational Weight Gain: A Longitudinal Study

Abstract

Background: Risk factors during adolescence appear to shape adult health, but little is known about how they are associated with pregnancy health.

Objectives: We aimed to assess whether a variety of adolescent risk factors with links to adult overweight or obesity are associated with pre-pregnancy obesity (Body Mass Index [BMI] ≥ 30 kg/m²) and high gestational weight gain (GWG; > 0.5 SD for pre-pregnancy BMI category and gestational age) in a cohort of women participating since adolescence in a longitudinal cohort.

Methods: At age 11-18 years participants reported on adolescent risk factors (overweight or obesity, healthy and unhealthy home food availability, food insufficiency, family meals, depressive symptoms, body dissatisfaction, weight teasing, binge eating, unhealthy weight control behaviours and dieting). Twenty years later, participants reporting a live birth (n = 656) recalled their pre-pregnancy weight and total GWG. Modified Poisson regression models were used to estimate associations of each factor with pre-pregnancy obesity and high GWG, adjusting for sociodemographics. We used Multivariate Imputation by Chained Equations to account for outcome misclassification using internal validation data.

Results: Eighteen percent of the sample had pre-pregnancy obesity and 26% had high GWG. Adolescent overweight or obesity (RR = 4.98, 95% CI 3.27, 7.57), body dissatisfaction (RR = 1.99; 95% CI: 1.31, 3.03) and unhealthy weight control behaviours (RR = 1.70; 95% CI: 1.06, 2.74), among other factors, were associated with pre-pregnancy obesity risk. For high GWG, there were imprecise associations with adolescent overweight or obesity (RR = 1.57; 95% CI: 1.06, 2.31), binge eating (RR = 1.36; 95% CI: 0.77, 2.39) and unhealthy weight control behaviours (RR = 1.38; 95% CI: 0.84, 2.25), among others.

Conclusions: Findings suggest that some risk markers for pre-pregnancy obesity (and possibly high GWG) may be apparent as early as adolescence. Supporting adolescent health and well-being might have a role in improving weight-related health in the perinatal period.

Keywords: depressive symptoms; disordered eating; food insufficiency; gestational weight gain; life course epidemiology; maternal obesity.

FIGURE 1

LEAP analytic sample and validation subsample selection.

FIGURE 2

(a) Associations of adolescent risk factors with pre‐pregnancy obesity (BMI ≥ 30 kg/m²). Models 1 and 2 were run on non‐imputed, complete case self‐reported LEAP survey data only and did not adjust for outcome misclassification. Sample sizes for Models 1 and 2 ranged from 559 to 588. Model 1 did not adjust for covariates. Model 3 adjusted for the same set of covariates as Model 2 (race and ethnicity and adolescent socioeconomic status) and further adjusted estimates for misclassification of the outcome (pre‐pregnancy obesity) using medical record data from the validation subsample and multivariate imputation by chained equations. All estimates were adjusted for selection bias using stabilised inverse probability of selection weights. (b) Associations of adolescent risk factors with high gestational weight gain (z‐score > 0.5 SD). High gestational weight gain was defined as a gestational weight gain z‐score > +0.5 SD. Models 1 and 2 were run on non‐imputed, complete case self‐reported LEAP survey data only and did not adjust for outcome misclassification. Sample sizes for Models 1 and 2 ranged from 533 to 561. Model 1 did not adjust for covariates. Model 3 adjusted for the same set of covariates as Model 2 (race and ethnicity and adolescent socioeconomic status) and further adjusted estimates for misclassification of the outcome (high gestational weight gain) using medical record data from the validation subsample and multivariate imputation by chained equations. All estimates were adjusted for selection bias using stabilised inverse probability of selection weights.