Relapse to Smoking and Postpartum Weight Retention Among Women Who Quit Smoking During Pregnancy

Abstract

Postpartum weight retention contributes to obesity risk in women. Given that most women who quit smoking as a result of pregnancy will resume smoking within 6 months postpartum and that there is a robust association between smoking and weight, we sought to evaluate postpartum weight retention as a function of postpartum smoking status among women who had quit smoking during pregnancy. Women (N = 183) with biochemically confirmed cigarette abstinence at the end of pregnancy were recruited between February 2003 and November 2006. Women self-reported demographic information and weight before pregnancy. Smoking status and weight were documented at the end of pregnancy and at 6, 12, and 24 weeks postpartum. Breastfeeding was reported at 6 weeks postpartum. Differences in weight retention by relapse status at each assessment were evaluated. To examine weight retention in the presence of conceptually relevant covariates, mixed models with log-transformed weight data were used. At 24 weeks postpartum, 34.6% of women remained abstinent. Women who remained abstinent throughout the 24-week period retained 4.7 ± 2.1 kg more than did women who had relapsed by 6 weeks postpartum, P = 0.03. This difference in postpartum weight retention was significant after controlling for relevant covariates (age, race, breastfeeding, and pregravid BMI). Resumption of smoking within the first 6 weeks following childbirth is associated with decreased postpartum weight retention, even after controlling for breastfeeding and pregravid weight. Interventions to sustain smoking abstinence postpartum might be enhanced by components designed to minimize weight retention.

Pregnancy-related weight gain contributes to obesity risk among women, and weight gained during pregnancy is a consistent predictor of postpartum weight retention (1). Excessive gestational weight gain (GWG), or GWG that exceeds guidelines established by the Institute of Medicine, predicts maternal weight 15 (1) and 21 (2) years later, even after controlling for prepregnancy weight. In addition to GWG, pregravid weight (3), age (4), and breastfeeding (5) also have been related to postpartum weight retention, although lactation has not been consistently associated with postpartum weight loss (6).

Given the relationship between smoking and body weight (7), postpartum smoking relapse may also be salient to understanding postpartum weight retention. Approximately 65% of women who quit smoking during pregnancy will resume smoking within 6 months postpartum (8), and smoking-related weight concerns decrease the likelihood of remaining quit postpartum (9). Moreover, both a complete cessation of smoking (10,11) as well as a reduction in the number of cigarettes smoked during pregnancy (12) have been associated with increased GWG, and women concerned about cessation-related weight gain are less likely to make a prenatal quit attempt (13).

Despite the association between prenatal cessation and GWG, little is known about the relationship between postpartum smoking relapse and weight retention. Although some research on postpartum weight has evaluated perinatal smoking status (3,5), the nature of the relationship between smoking relapse and postpartum weight is unclear as prior studies have combined women who may never have smoked with those who may have quit and relapsed. Accordingly, we prospectively evaluated the relationship between postpartum smoking relapse and weight among women who had quit smoking consequent to pregnancy. We sought to determine the relationship between smoking status and weight retention during the first 6 months postpartum and examine the relative impact of smoking status in the context of other variables related to postpartum weight retention.

METHODS

Pregnant women who had quit smoking were recruited for a study concerning the prediction of postpartum smoking relapse. Women were eligible if they: (i) were at least 28 weeks pregnant, (ii) reported smoking at least eight cigarettes/day for at least 1 month before pregnancy, (iii) were not currently smoking, as verified by a carbon monoxide (CO) reading of 8 ppm or less at the end of pregnancy, and (iv) were at least 14 years of age. Additional information has been reported elsewhere (9) and in Supplementary Table S1 online.

Women were interviewed in the third trimester of pregnancy about previous smoking and provided an expired-air sample for CO content. Women were assessed again at 6, 12, and 24 weeks postpartum. A subsample of participants (n =56) was enrolled in a pilot study and did not complete a 24-week postpartum assessment; these women were followed through 12 weeks postpartum. Participants were compensated for completing assessments.

Assessments

Demographic information was collected at the end of pregnancy. Current breastfeeding was self-reported at 6 weeks postpartum.

Weight/weight retention

At the end of pregnancy, women self-reported current and pregravid weights. Prepregnancy BMI (kg/m²) was calculated, and GWG was estimated as the difference between prepregnancy weight and current weight at the time of the third trimester assessment. Postpartum weight retention, the primary-dependent variable, was defined as the difference between weight at each postpartum assessment and self-reported weight before pregnancy.

Postpartum smoking

At each visit after delivery, women were interviewed using timeline follow back methodology, and CO was measured. Postpartum relapse was defined as the self-report of smoking for 7 consecutive days after delivery and before assessment, or a biochemical indication of smoking at the assessment. A CO reading of less than or equal to 8 ppm was used to confirm self-reported nonsmoking. Women who missed an appointment or were lost to follow-up were considered to have relapsed, except when a subsequent biochemical assessment confirmed nonsmoking.

Analysis

We first evaluated differences in weight retention at each assessment point by relapse status. Given the positive skew in weight retention, nonparametric Wilcoxon rank-sum tests were used to evaluate differences between women who had and had not resumed smoking at each postpartum visit. Next, we categorized women who had a 24-week assessment as: maintained continuous abstinence for 24 weeks (abstainers, n = 44), relapsed by 6 weeks (early-relapsers, n = 52), relapsed between 6 and 12 weeks (middle-relapsers, n = 17) and relapsed between 12 and 24 weeks postpartum (late-relapsers, n = 14). We then compared weight changes using mixed models with log-transformed data.

Finally, we evaluated the relationship of smoking status to weight retention after controlling for variables previously shown to be related to weight retention (age, GWG, pregravid BMI, and breastfeeding). Our measure of GWG was taken several weeks prior to delivery, and there was wide variation in the timing of assessment (0–17 weeks, 7.3 weeks on average). Given this variation, and because weight gain occurs throughout the third trimester, we included both GWG and the number of weeks before delivery weight in the model. Thus, we evaluated smoking status in a mixed model of postpartum weight retention that included GWG, weeks before delivery, age, breastfeeding at 6 weeks and pregravid BMI.

RESULTS

At 6 weeks postpartum, 53.6% (98/183) of women were abstinent. Rates of abstinence were 37.7% (69/183) and 34.6% (44/127) at 12 and 24 weeks postpartum, respectively. Women retained 9.4 ± 9.4, 9.5 ± 9.8, and 9.1 ± 10.1 kg at 6, 12, and 24 weeks postpartum, respectively. As shown in Figure 1, at each postpartum assessment, women who remained abstinent retained more weight than did those who resumed smoking. Abstinent women were 2.9 ± 1.8 kg (P = 0.01), and 3.6 ± 2.0, kg (P = 0.02) heavier than women who had resumed smoking at 12 and 24 weeks, respectively.

Figure 1.

Postpartum weight retention by weeks among women who are and are not abstinent.

Weight retention in abstainers, early-, middle-, and late-relapsers

Although all women retained a significant amount of weight over time, early-relapsers retained 4.7 ±2.1 kg less (6.4 kg ± 1.6 vs. 11.1 kg ± 1.5) than did abstainers, P = 0.01. Women who were middle-relapsers (9.4 kg ± 2.4) and late-relapsers (10.8 kg ± 2.7) also retained a substantial amount of weight, but the difference between their weight retention and that of abstainers was not significant (P > 0.54).

Correlates of weight retention

Next, we examined smoking relapse in the context of other factors that have been related to postpartum weight retention. Age and breastfeeding were not related to postpartum weight retention, but pregravid BMI and GWG were related to postpartum weight retention. Women who were heavier prepregnancy retained more weight postpartum (P = 0.02), and those who gained more weight during pregnancy retained more weight after delivery (P < 0.0001). Nonetheless, relapse status remained significantly related to weight retention after adjustment for all covariates at 12 (P = 0.03) and 24 weeks (P = 0.01) postpartum. Similarly, in the analysis comparing early-, middle-, and late-relapsers to abstainers, the difference between early-relapsers and abstainers remained significant (P = 0.002) after covariate adjustment.

DISCUSSION

Women who successfully sustain smoking abstinence during the first 6 months postpartum retain more weight than do those who resume smoking, regardless of pregravid weight or GWG. Differences between those who do and do not remain abstinent emerge early postpartum. In addition, although all of the women who quit smoking before or during pregnancy retained a substantial amount of weight postpartum, the finding that sustained smoking abstinence is associated with significantly more postpartum weight retention suggests that efforts to sustain smoking abstinence or minimize postpartum weight need to address smoking and weight-related behaviors simultaneously

The amount of postpartum weight retention among women who had quit smoking prenatally was significantly more than amounts reported in samples not selected on the basis of smoking status. It is possible that the greater weight retention among women who quit smoking during pregnancy is as a result of weight accrued consequent to smoking cessation (7) as well as from GWG. Correlates of postpartum weight retention, notably GWG and prepregnancy weight, were related to postpartum weight among women who had quit smoking as they have been in other samples (3). However, smoking relapse remained significantly related to weight retention after controlling for these factors. Thus, women who have quit smoking as a result of pregnancy may be particularly vulnerable to weight gain and at risk for onset or exacerbation of obesity.

Strengths of the study include the prospective nature of the data, diverse sample, biochemical confirmation of smoking status at each assessment, and control for many of the known correlates of postpartum weight in determining the relative contribution of smoking status to weight retention. In addition, these data provide unique information about changes in weight following smoking resumption outside of pregnancy, as women were followed for 6 months postpartum, regardless of smoking status. Nonetheless, results need to be considered in the context of limitations that typify research on perinatal smoking and weight. First, although the use of self-reported pregravid weights is common (14,15), pregravid weights were collected at the end of pregnancy, and the timing may have increased the inaccuracy of recall. However, the use of measured weight and an analytic strategy focused on changes in weight over time, rather than absolute weight values, help to mitigate the effects of self-report bias. Second, the use of CO to document nonsmoking status is imperfect and may miss individuals who smoke infrequently. Finally, the measures of weight and smoking status occurred at specific points postpartum, and we were unable to determine the precise temporal relationship between weight and smoking relapse. More frequent sampling postpartum and collection of additional variables related to weight gain may provide useful information on the timing of relapse and associated weight changes.

In summary, these data suggest that attending to both smoking and weight may be necessary to change women's health behaviors postpartum. Efforts to sustain smoking abstinence postpartum may need to target eating behavior and physical activity to minimize weight retention and do so in the context of other issues that affect women during and after pregnancy. Similarly, interventions to prevent excessive GWG and minimize postpartum weight retention might be enhanced by consideration of the impact of smoking.