Smoking and Type 2 Diabetes: Title and subTitle BreakUnderrecognized Risks and Disease Burden

Although smoking has long been acknowledged as a major contributor to the global burden of cardiovascular disease and cancer, the adverse effect of smoking on type 2 diabetes has been generally underrecognized. Indeed, the most recent Centers for Disease Control and Prevention estimates of smoking-attributable mortality do not incorporate the impact of smoking on diabetes.¹

Diabetes exerts a considerable national and global disease burden. For US residents born in 2000, an estimated 1 in 3 will develop diabetes during their lifetime.² Approximately 21 million persons in the United States currently have diabetes, with an additional 41 million with prediabetes.³ The American Diabetes Association estimates that diabetes costs US society $132 billion annually, with the economic burden expected to increase to $192 billion by 2020.⁴ Globally, more than 150 million people have diabetes, with international prevalence expected to double by 2025.⁵

In this issue of JAMA, Willi and colleagues⁶ present the results of a systematic review and meta-analysis of 25 prospective cohort studies, comprising 1.2 million participants and 45 844 cases of diabetes, and report that both active and past smoking are significantly associated with increased risk of type 2 diabetes. Although the study is not able to analyze exact differences in diabetes risk associated with different smoking pack-years, duration, and cessation patterns, the findings demonstrate strong consistency of the association. Despite some evidence for heterogeneity, the association was overall robust and consistent across a range and variety of smoking patterns, demographics, and study characteristics. Although there is potential for residual confounding by factors such as body mass index and alcohol use, the results were similar if not stronger among studies included in the meta-analysis that used substantial multivariable adjustment for such confounders. Furthermore, because smoking is inversely related to body weight,⁷ and alcohol is associated with lower glucose levels and diabetes risk in a randomized controlled trial⁸ and prospective cohort studies,⁹ residual confounding by relative weight and alcohol would be expected to attenuate the association between smoking and diabetes. Therefore, the findings in the study by Willi et al⁶ likely reflect a conservative underestimate of the true association between smoking and type 2 diabetes.

Even though smoking is known to decrease body weight, it is associated with an increase in central adiposity,⁷ an established marker of hyperglycemia and dyslipidemia. Smoking also is well recognized to increase inflammation and oxidative stress,¹⁰ to directly damage beta-cell function,¹¹ and to impair endothelial function,¹² each of which have been strongly implicated in insulin resistance¹³ and diabetes risk.^{14 - 15} Smoking has been shown to impair insulin sensitivity and glucose tolerance.¹⁶ Overall, multiple lines of evidence support a causal nature of the association¹⁷ between smoking and type 2 diabetes.

Given the strong evidence of causality, it is important to try to quantify the burden of diabetes attributable to smoking. Using the most recent US statistics of current and former smoking prevalence,¹⁸ the estimates from the study by Willi et al,⁶ and the conventional population-attributable risk formula,¹⁹ an estimated 12% of all type 2 diabetes in the United States may be attributable to smoking. This percentage would be expected to be much higher in populations with higher smoking rates. Additionally, with type 2 diabetes accounting for 95% of the 20.6 million adults with diabetes,³ an estimated 2.3 million cases of diabetes in the United States and a corresponding $14.9 billion of the annual US $132 billion diabetes cost burden may be attributable to smoking.

Diabetes also is recognized to be a strong causal contributor to other leading causes of death, notably cardiovascular disease mortality.²⁰ A large international review indicates that 21% of all coronary heart disease deaths and 13% of all stroke deaths are attributable to high blood glucose levels.²¹ Diabetes and elevated glucose levels also are suspected as a contributor to the disease burden of cancer, associated with a 20% to 30% relative increase in overall cancer risk²² and even stronger associations for several common types of cancer. To date, such major contributions to the mortality burden have not been assigned to diabetes in tabulation of conventional cause-of-death statistics and analyses of attributable mortality. Thus, although the Centers for Disease Control and Prevention listed diabetes as the sixth leading cause of death in the United States with 74 000 deaths in 2003,²³ the total attributable national and global mortality burden is likely underestimated for diabetes, and thus the impact of smoking as well.

Given the substantial influence of smoking on diabetes risk, consideration should be given to clinical screening. Because diabetes is currently highly underdiagnosed, with an estimated 30% to 50% of all diabetes cases clinically unrecognized,³ the American Diabetes Association recommends general screening for undiagnosed diabetes every 3 years among men and women aged 45 years or older.²⁴ However, as recently as 2003, the US Preventive Services Task Force had not endorsed universal clinical screening for glucose abnormalities, with current recommendations suggesting screening only among adults with hypertension and hyperlipidemia,²⁵ although updated recommendations are pending. Given the increased incidence of type 2 diabetes associated with smoking, it is likely important and prudent for clinicians also to screen for and carefully monitor glucose levels among current and former smokers.

Furthermore, along with clinical screening, recommendations for type 2 diabetes prevention should incorporate smoking avoidance accompanied by lifestyle modification. Although a frequent concern of smoking cessation is subsequent weight gain, moderately increasing exercise can largely minimize the approximately 2 kg weight gain associated with stopping smoking,²⁶ indicating that the public health issues of smoking, exercise, and obesity are inextricably intertwined. Major population prevention of type 2 diabetes is achievable via avoidance of smoking and modification of lifestyle factors through a combination of healthy weight control, regular physical activity, moderate alcohol intake, and proper diet. An estimated 91% of all type 2 diabetes is preventable by smoking prevention and lifestyle modification.²⁷

Although global smoking-attributable mortality is expected to decline 9% between 2002 and 2030 among high-income countries, the disease burden of tobacco is expected to at least double from 3.4 million to 6.8 million in low- and middle-income countries,²⁸ exclusive of the effect of smoking on type 2 diabetes risk. The synergy between recent increasing obesity and smoking prevalence in developing nations bodes ominously for further escalation of the global burden of diabetes and other chronic diseases. In the United States, smoking tends to cluster in the low socioeconomic stratum of society and in minority groups.¹⁸ Moreover, while obesity prevalence continues to increase, the decline in smoking rates in the United States has stalled in the past several years.¹⁸

Therefore, further focus on public health and clinical interventions is needed to prevent tobacco use and promote smoking cessation. As jointly recognized and highlighted by the American Diabetes Association, the American Heart Association, and the American Cancer Society,²⁹ prevention of diabetes, cardiovascular disease, and cancer share many common agendas—a central focus of which is the avoidance of cigarette smoking.