Diet and Tobacco Use: Analysis of Data From the Diabetic Control and Complications Trial, a Randomized Study

David K. Cundiff

In This Article


Before applying the results of the DCCT to a general population, we need to address the possible influence of the patients' type 1 diabetes in their diets and serum lipid profiles. Although the DCCT dietitians strongly advised all the study subjects to reduce dietary fat and saturated fat to less than 30% and 10% of calories, respectively, the subjects averaged about 37% total fat and 13% saturated fat intake. This is similar if not higher than the percentage of calories as fat consumed by the average American.[38,39] Likewise, the lipid profiles of the study subjects mirror those of age-matched average Americans.[40] Therefore, at least in the DCCT subjects, type 1 diabetes does not appear to have significantly altered the dietary fat intake or the lipid profiles compared with nondiabetics.

The influence of smoking on diet is independent of age (Tables 5a and 5b). However, diet analyses and lipid profiles of age and sex matched nondiabetic smokers and nonsmokers and people older than age 40 years should be done to help discover any possible differences.

In the DCCT database, both male and female smokers ate a more animal-based diet (ie, more fats and cholesterol and less fiber) vs the more plant-based diet of those who refrained from tobacco use. Tobacco use and higher intakes of dietary fat are both strongly associated with cardiovascular disease and some cancers. Tobacco's direct effect in the etiology of atherosclerosis and cancer may be supplemented by the indirect effect of its relationship with higher dietary fat and cholesterol and lower fiber intake.

Several potential mechanisms for the association of tobacco use and higher animal vs plant food intake are possible. Tobacco using may reflect a general lack of concern about a healthy lifestyle that extends to eating habits. The adverse effects of tobacco use on taste and smell of food may lead smokers to compensate by increasing intake of animal fat and salt while reducing plant-based foods.

Former tobacco users of both sexes ate more like never users than current users. This suggests that the dietary preferences relate more to the tobacco itself than to other individual differences between those inclined vs not inclined to use tobacco.

Male smokers consumed significantly more calories without increased BMI compared with nonusers. Given that 1 pound equals about 3500 calories, tobacco use reduced weight gain in males by at least 12 pounds per year. Female tobacco users ate consistently but not statistically significantly more calories than nonusers but still gained significantly less weight (Tables 1a and 3a). Accordingly, any effect of tobacco use in controlling weight derives from increasing the metabolism or other mechanism besides calorie reduction. The higher HbA1c and associated glucosuria in tobacco users may partly account for the lack of inordinate weight gain.

In a prospective study of male physicians, Manson and colleagues[41] found up to a 2.1-fold risk of developing type 2 diabetes associated with cigarette smoking. Since the development of type 2 diabetes has been associated with a high-fat diet,[42] the higher calorie and fat and lower fiber diets associated with tobacco use in men may be the causative factor that increases incidence of type 2 diabetes.

Tobacco users of both sexes consistently had lipid profiles that were associated with greater cardiovascular risk than those of nontobacco-users (Tables 1a and 1b). The lipid profiles of former tobacco users resembled those of nonusers rather than current users (Tables 2a, 2b, 4a, and 4b). This suggests that the change to a more plant-based diet associated with quitting tobacco leads to lower-risk lipid profiles. This correlation of smoking with higher-risk lipid profiles is consistent with findings of Poletto and colleagues.[43] In a group of 238 smoking and nonsmoking 18-year-old men, they found higher serum cholesterols in the smokers.

This study's limitations include being a retrospective analysis of data that was not planned to analyze the effect of tobacco use on diet. More frequent dietary analyses (eg, monthly instead of only every 2 years) and tobacco histories (eg, monthly instead of yearly) would have strengthened the findings. Nondiabetic smokers and nonsmokers should also be studied to confirm the relationships in this study.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.