BMI and Risk of Lung Cancer Among Smokers
BMI and Risk of Lung Cancer Among Smokers
Background Although obesity has been directly linked to the development of many cancers, many epidemiological studies have found that body mass index (BMI)—a surrogate marker of obesity—is inversely associated with the risk of lung cancer. These studies are difficult to interpret because of potential confounding by cigarette smoking, a major risk factor for lung cancer that is associated with lower BMI.
Methods We prospectively examined the association between BMI and the risk of lung cancer among 448 732 men and women aged 50–71 years who were recruited during 1995–1996 for the National Institutes of Health–AARP Diet and Health Study. BMI was calculated based on the participant's self-reported height and weight on the baseline questionnaire. We identified 9437 incident lung carcinomas (including 415 in never smokers) during a mean follow-up of 9.7 years through 2006. Multivariable Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) with adjustment for lung cancer risk factors, including smoking status. To address potential bias due to preexisting undiagnosed disease, we excluded potentially unhealthy participants in sensitivity analyses. All statistical tests were two-sided.
Results The crude incidence rate of lung cancer over the study follow-up period was 233 per 100 000 person-years among men and 192 per 100 000 person-years among women. BMI was inversely associated with the risk of lung cancer among both men and women (BMI ≥35 vs 22.5–24.99 kg/m: HR = 0.81, 95% CI = 0.70 to 0.94 and HR = 0.73, 95% CI = 0.61 to 0.87, respectively). The inverse association was restricted to current and former smokers and was stronger after adjustment for smoking. Among smokers, the inverse association persisted even after finely stratifying on smoking status, time since quitting smoking, and number of cigarettes smoked per day. Sensitivity analyses did not support the possibility that the inverse association was due to prevalent undiagnosed disease.
Conclusions Our results suggest that a higher BMI is associated with a reduced risk of lung cancer in current and former smokers. Our inability to attribute the inverse association between BMI and the risk of lung cancer to residual confounding by smoking or to bias suggests the need for considering other explanations.
Epidemiological studies have consistently shown that obesity is associated with an increased risk of many cancers, including thyroid, colon, renal, and endometrial cancers. Conversely, body mass index (BMI)—a surrogate indicator of obesity—has been inversely associated with the risk of incident and fatal lung cancers in many prospective investigations. Because numerous cohort studies have found that this inverse association is observed only among ever or current smokers, the finding has been attributed to potential residual confounding by cigarette smoking. However, BMI has also been found to be unrelated with the risk of lung cancer irrespective of smoking status or to be positively associated with risk among nonsmokers or ever smokers.
Previous cohort studies have been limited in their ability to explore the relationship between BMI and lung cancer for a number of reasons. Some have only examined associations with lung cancer mortality. Others have not stratified the analyses by sex or smoking status, potentially oversimplifying the relationship. Some studies have had small numbers of female current smokers, which precluded their ability to analyze associations within this important group. Furthermore, only two epidemiological studies have presented results stratified by histological subtype. The suggestion that different histological types of lung cancer may represent largely distinct diseases with divergent etiologies supports the need to examine associations between BMI and individual histological subgroups.
We attempted to clarify the relationship between BMI and incident lung cancer in the large prospective National Institutes of Health (NIH)–AARP Diet and Health Study. The large size of this cohort enabled us to evaluate detailed associations by sex and smoking status and to assess potential differences in associations according to histology.
Abstract and Introduction
Abstract
Background Although obesity has been directly linked to the development of many cancers, many epidemiological studies have found that body mass index (BMI)—a surrogate marker of obesity—is inversely associated with the risk of lung cancer. These studies are difficult to interpret because of potential confounding by cigarette smoking, a major risk factor for lung cancer that is associated with lower BMI.
Methods We prospectively examined the association between BMI and the risk of lung cancer among 448 732 men and women aged 50–71 years who were recruited during 1995–1996 for the National Institutes of Health–AARP Diet and Health Study. BMI was calculated based on the participant's self-reported height and weight on the baseline questionnaire. We identified 9437 incident lung carcinomas (including 415 in never smokers) during a mean follow-up of 9.7 years through 2006. Multivariable Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) with adjustment for lung cancer risk factors, including smoking status. To address potential bias due to preexisting undiagnosed disease, we excluded potentially unhealthy participants in sensitivity analyses. All statistical tests were two-sided.
Results The crude incidence rate of lung cancer over the study follow-up period was 233 per 100 000 person-years among men and 192 per 100 000 person-years among women. BMI was inversely associated with the risk of lung cancer among both men and women (BMI ≥35 vs 22.5–24.99 kg/m: HR = 0.81, 95% CI = 0.70 to 0.94 and HR = 0.73, 95% CI = 0.61 to 0.87, respectively). The inverse association was restricted to current and former smokers and was stronger after adjustment for smoking. Among smokers, the inverse association persisted even after finely stratifying on smoking status, time since quitting smoking, and number of cigarettes smoked per day. Sensitivity analyses did not support the possibility that the inverse association was due to prevalent undiagnosed disease.
Conclusions Our results suggest that a higher BMI is associated with a reduced risk of lung cancer in current and former smokers. Our inability to attribute the inverse association between BMI and the risk of lung cancer to residual confounding by smoking or to bias suggests the need for considering other explanations.
Introduction
Epidemiological studies have consistently shown that obesity is associated with an increased risk of many cancers, including thyroid, colon, renal, and endometrial cancers. Conversely, body mass index (BMI)—a surrogate indicator of obesity—has been inversely associated with the risk of incident and fatal lung cancers in many prospective investigations. Because numerous cohort studies have found that this inverse association is observed only among ever or current smokers, the finding has been attributed to potential residual confounding by cigarette smoking. However, BMI has also been found to be unrelated with the risk of lung cancer irrespective of smoking status or to be positively associated with risk among nonsmokers or ever smokers.
Previous cohort studies have been limited in their ability to explore the relationship between BMI and lung cancer for a number of reasons. Some have only examined associations with lung cancer mortality. Others have not stratified the analyses by sex or smoking status, potentially oversimplifying the relationship. Some studies have had small numbers of female current smokers, which precluded their ability to analyze associations within this important group. Furthermore, only two epidemiological studies have presented results stratified by histological subtype. The suggestion that different histological types of lung cancer may represent largely distinct diseases with divergent etiologies supports the need to examine associations between BMI and individual histological subgroups.
We attempted to clarify the relationship between BMI and incident lung cancer in the large prospective National Institutes of Health (NIH)–AARP Diet and Health Study. The large size of this cohort enabled us to evaluate detailed associations by sex and smoking status and to assess potential differences in associations according to histology.
