Association Between Smoking and Total Energy Expenditure
Association Between Smoking and Total Energy Expenditure
Background The association between smoking and total energy expenditure (TEE) is still controversial. We examined this association in a multi-country study where TEE was measured in a subset of participants by the doubly labeled water (DLW) method, the gold standard for this measurement.
Methods This study includes 236 participants from five different African origin populations who underwent DLW measurements and had complete data on the main covariates of interest. Self-reported smoking status was categorized as either light (<7 cig/day) or high (≥7 cig/day). Lean body mass was assessed by deuterium dilution and physical activity (PA) by accelerometry.
Results The prevalence of smoking was 55% in men and 16% in women with a median of 6.5 cigarettes/day. There was a trend toward lower BMI in smokers than non-smokers (not statistically significant). TEE was strongly correlated with fat-free mass (men: 0.70; women: 0.79) and with body weight (0.59 in both sexes). Using linear regression and adjusting for body weight, study site, age, PA, alcohol intake and occupation, TEE was larger in high smokers than in never smokers among men (difference of 298 kcal/day, p = 0.045) but not among women (162 kcal/day, p = 0.170). The association became slightly weaker in men (254 kcal/day, p = 0.058) and disappeared in women (−76 kcal/day, p = 0.380) when adjusting for fat-free mass instead of body weight.
Conclusion There was an association between smoking and TEE among men. However, the lack of an association among women, which may be partly related to the small number of smoking women, also suggests a role of unaccounted confounding factors.
Tobacco use is the leading avoidable cause of death worldwide. The belief that smoking may be a weight control agent constitutes a reason to initiate or to pursue smoking, especially among women and adolescents. Nevertheless, findings on the effects of smoking on body weight are inconsistent in the literature. Body weight was shown to be lower in smokers compared to nonsmokers in many population-based studies conducted a few decades ago, however, several recent cohort studies report no body weight lowering effects. A recent Mendelian randomization analysis suggested that smoking causes lower body mass index (BMI). It is therefore important to update the current literature investigating the association between smoking and body weight, in particular with regards to claims about the benefits of smoking related to weight control, which may undermine national tobacco control programs.
Two broad mechanisms may underlie an effect of smoking on body weight. First it is reported that smoking increases total energy expenditure (TEE) through increasing the resting metabolic rate (RMR) and/or decreasing energy intake as a result of suppressing appetite. A number of experimental studies showed that a moderate increase in energy expenditure occurred for short periods of time right after cigarette smoking, although no effect was found in another study. More recently, two cross-sectional studies examined the association between smoking and TEE in smokers vs. non smokers using the doubly labeled water (DLW) technique, the gold standard to assess TEE. Warwick et al. found no association but the study relied on a very small sample size (i.e., 21 participants). The second study, by Bradley et al. involved 304 participants and also did not find an association between TEE and smoking but smoking was not categorized by the intensity of the exposure, e.g. as light or heavy smokers.
In this paper, we examine the association between smoking, BMI and TEE measured with DLW in a subset of participants of an ongoing study in several populations of African origin (Modeling the Epidemiological Transition Study—METS). We hypothesized that smokers would have a higher TEE and a lower BMI compared to never smokers.
Abstract and Introduction
Abstract
Background The association between smoking and total energy expenditure (TEE) is still controversial. We examined this association in a multi-country study where TEE was measured in a subset of participants by the doubly labeled water (DLW) method, the gold standard for this measurement.
Methods This study includes 236 participants from five different African origin populations who underwent DLW measurements and had complete data on the main covariates of interest. Self-reported smoking status was categorized as either light (<7 cig/day) or high (≥7 cig/day). Lean body mass was assessed by deuterium dilution and physical activity (PA) by accelerometry.
Results The prevalence of smoking was 55% in men and 16% in women with a median of 6.5 cigarettes/day. There was a trend toward lower BMI in smokers than non-smokers (not statistically significant). TEE was strongly correlated with fat-free mass (men: 0.70; women: 0.79) and with body weight (0.59 in both sexes). Using linear regression and adjusting for body weight, study site, age, PA, alcohol intake and occupation, TEE was larger in high smokers than in never smokers among men (difference of 298 kcal/day, p = 0.045) but not among women (162 kcal/day, p = 0.170). The association became slightly weaker in men (254 kcal/day, p = 0.058) and disappeared in women (−76 kcal/day, p = 0.380) when adjusting for fat-free mass instead of body weight.
Conclusion There was an association between smoking and TEE among men. However, the lack of an association among women, which may be partly related to the small number of smoking women, also suggests a role of unaccounted confounding factors.
Introduction
Tobacco use is the leading avoidable cause of death worldwide. The belief that smoking may be a weight control agent constitutes a reason to initiate or to pursue smoking, especially among women and adolescents. Nevertheless, findings on the effects of smoking on body weight are inconsistent in the literature. Body weight was shown to be lower in smokers compared to nonsmokers in many population-based studies conducted a few decades ago, however, several recent cohort studies report no body weight lowering effects. A recent Mendelian randomization analysis suggested that smoking causes lower body mass index (BMI). It is therefore important to update the current literature investigating the association between smoking and body weight, in particular with regards to claims about the benefits of smoking related to weight control, which may undermine national tobacco control programs.
Two broad mechanisms may underlie an effect of smoking on body weight. First it is reported that smoking increases total energy expenditure (TEE) through increasing the resting metabolic rate (RMR) and/or decreasing energy intake as a result of suppressing appetite. A number of experimental studies showed that a moderate increase in energy expenditure occurred for short periods of time right after cigarette smoking, although no effect was found in another study. More recently, two cross-sectional studies examined the association between smoking and TEE in smokers vs. non smokers using the doubly labeled water (DLW) technique, the gold standard to assess TEE. Warwick et al. found no association but the study relied on a very small sample size (i.e., 21 participants). The second study, by Bradley et al. involved 304 participants and also did not find an association between TEE and smoking but smoking was not categorized by the intensity of the exposure, e.g. as light or heavy smokers.
In this paper, we examine the association between smoking, BMI and TEE measured with DLW in a subset of participants of an ongoing study in several populations of African origin (Modeling the Epidemiological Transition Study—METS). We hypothesized that smokers would have a higher TEE and a lower BMI compared to never smokers.