摘要：Objectives. In non-American Indian/Alaska Native groups, current smoking prevalence is similar for those with or without diabetes (26%) We analyzed current smoking prevalence in American Indian/Alaska Natives by diabetes status. Methods. Data were extracted from Indian Health Service clinic visit information from 1998 to 2003. After consolidation into unique patient records, the sample comprised 71221 patients aged 14 years or older with both diabetes and current smoking information. Results. Cross-sectional results indicated that diabetic American Indian/Alaska Natives were significantly more likely than those without diabetes to be current smokers (29.8% vs 18.8%; P <.01). Smoking rates were 2 to 3 times higher among diabetic American Indians and Alaska Natives for each age category ( P <.001), and current smokers with diabetes were more likely than nonsmokers to have glycosylated hemoglobin A1c levels at 8.0% or higher ( P <.05). Conclusions. American Indian/Alaska Natives with diabetes at all sites and age categories were found to smoke at significantly higher rates than those without diabetes. Smoking cessation programs should target diabetic patients to more effectively prevent complications and promote successful management of diabetes in American Indians/Alaska Natives. Several clinical and large prospective studies have reported significant links between smoking and the development of diabetes, 1 , 2 micro- and macrovascular complications, 3 and impairment of metabolic control. 4 , 5 Foy 6 reported a linear dose relationship of incident diabetes with increasing number of pack years of smoking, and an odds ratio of 5.7 ( P >.001) for current smokers with 20 or more pack years of smoking. Smoking is an independent risk factor for cardiovascular disease, and acute and chronic tobacco exposure significantly impairs glucose tolerance and increases insulin resistance. Conversely, improvement in insulin sensitivity and elevated levels of high-density lipoprotein cholesterol both occur after smoking cessation. 5 However, most studies in the general population have shown the prevalence of tobacco use to be similar among those with or without diabetes (26%), but many of these studies have used self-report of diabetes and tobacco use. 7 Diabetes is the fourth leading cause of death for American Indian and Alaska Native (AIAN) populations and is the major independent risk factor for cardiovascular disease, the leading cause of death among American Indians and Alaska Natives. 8 AIAN populations have the highest reported tobacco use of any US ethnic group—40.4% reported from 1999–2001 Centers for Disease Control and Prevention (CDC) data, 9 and 32% reported from 2005 CDC data. 10 However, data that show smoking prevalence stratified by type 2 diabetes status are scarce. Although Indian Health Service (IHS) data are the most comprehensive nationally based AIAN health data available, the data report age, gender, socioeconomic status, and mortality rates but not disease or lifestyle behavior prevalence rates. 8 In other national databases, only small samples of American Indians and Alaska Natives are included and therefore prevalence estimates are not representative or reliable. Many American Indians and Alaska Natives included in these data sets are misclassified as another race/ethnicity and vice versa (i.e., many people who do not have legitimate AIAN heritage self-identify as American Indian and Alaska Native). 11 , 12 The Strong Heart Study, a longitudinal cohort study of cardiovascular disease among American Indians and Alaska Natives, has not reported current smoking by diabetes status but has reported similar rates of current smoking for those with incident cardiovascular disease (44.8%) compared with those with no incident cardiovascular disease (40.2%). 13 Only 1 study of cardiovascular disease risk factors among American Indians and Alaska Natives reported smoking prevalence rates by diabetes status that were similar to the Strong Heart Study (34% vs 41%). 14 However, that study used a telephone interview survey to assess self-reported diabetes and smoking prevalence rates. Gilmer et al. 15 showed that for every 1% rise in glycosylated hemoglobin A1c levels (HbA1c; a more stable glucose measure over a 2- to 3-month period than single glucose measures), there was a 15% increase in the cost of medical care. Their report identified the clinical threshold of HbA1c level at 8.0%, above which complication rates and costs significantly intensified. One way to maintain better glucose control and prevent complications is to eliminate smoking. However, the prevalence of smoking in diabetic American Indians and Alaska Natives who have HbA1c levels at 8.0% or higher has not been adequately reported. We sought to describe current smoking prevalence by diabetes status among a large group of American Indian/Alaska Natives, as well as in diabetic American Indians and Alaska Natives whose glucose was not well controlled. Data were collected in the clinic setting and extracted from existing IHS and tribally owned health facility patient visit data from 1998 to 2003.
