Leanness and lung cancer risk among women

Several epidemiologic studies indicate that lean individuals are at increased risk of lung cancer. Until recently the association was thought to be noncausal, reflecting either disease effects or confounding by smoking. In a population-based case-control study of Missouri women, 35-84 years of age, we assessed lung cancer risk in relation to body mass index (BMI, kg/m²) based on participants reports of adult height, contemporary and past weight. This analysis focused on interviews obtained from 392 cases and 617 controls interviewed directly. BMI at interview was strongly inversely associated with risk but clearly reflected weight loss due to disease. When we examined BMI two decades prior to interview, the inverse association was less pronounced but persisted. After adjusting for smoking, relative risks across decreasing quartiles of BMI were 1.0, 1.11, 1.41, 1.56; p for trend < 0.01. Weight gain from age 20 to two decades prior to interview was protective; risk of the disease decreased 40 percent among women in the highest category of weight gain. The relation of BMI to risk of lung cancer did not vary according to histologic classification of cases. In summary, leanness many years prior to the onset of lung cancer was associated with increased risk of the disease.