Findings from the Third National Health and Nutrition Examination Survey (NHANES III)
OBJECTIVE—To describe diet and exercise practices from a nationally representative sample of U.S. adults with type 2 diabetes.
METHODS—We analyzed data from 1,480 adults older than 17 years with a self-reported diagnosis of type 2 diabetes in the Third National Health and Nutrition Examination Survey (NHANES III). Fruit and vegetable consumption was obtained from a food frequency questionnaire; the percentages of total calories from fat and saturated fat were obtained from a 24-h food recall. Physical activity was based on self report during the month before the survey.
RESULTS—Of individuals with type 2 diabetes, 31% reported no regular physical activity and another 38% reported less than recommended levels of physical activity. Sixty-two percent of respondents ate fewer than five servings of fruits and vegetables per day. Almost two thirds of the respondents consumed >30% of their daily calories from fat and >10% of total calories from saturated fat. Mexican Americans and individuals over the age of 65 years ate a higher number of fruits and vegetables and a lower percentage of total calories from fat. Lower income and increasing age were associated with physical inactivity. Thirty-six percent of the sample were overweight and another 46% were obese.
CONCLUSIONS—The majority of individuals with type 2 diabetes were overweight, did not engage in recommended levels of physical activity, and did not follow dietary guidelines for fat and fruit and vegetable consumption. Additional measures are needed to encourage regular physical activity and improve dietary habits in this population.
- BRFSS, Behavioral Risk Factor Surveillance System
- NHANES III, the Third National Health and Nutrition Examination Survey
Diet and exercise are considered important components of the treatment strategy for adults with type 2 diabetes. Appropriate use of diet and exercise can improve insulin sensitivity and glycemic control and decrease the need for oral medications or insulin (1,2). Although there is some controversy over the optimal diet for adults with type 2 diabetes (high fiber, glycemic index approaches, low versus moderate fat) (3,4), there is a consensus to increase consumption of fruits and vegetables and decrease daily consumption of saturated fats (4,5). Regular, moderate-intensity physical activity for 30 min at least 5 times per week is recommended for most Americans (6).
Previous studies suggest that individuals with diabetes may not follow recommended guidelines for diet and exercise, although there have been no nationally representative U.S. studies examining nutritional intake among adults with type 2 diabetes. In a survey of >2,000 adults with diabetes, the most frequently reported barriers in diabetes self-management were adherence to diet and exercise (7). Only 60% of individuals with diabetes in the National Health Interview Survey report that they “follow a diabetic diet” (8), and several non-U.S. studies have reported that actual nutrient intake among individuals with diabetes may be suboptimal (9–12). In addition, the majority of individuals in the U.S., including those with diabetes, are not meeting national physical activity goals (13).
The purpose of this study is to describe both dietary intake and physical activity from a nationally representative sample of adults with type 2 diabetes. We assessed compliance with national guidelines for physical activity (6,14) and total fat, saturated fat, and fruit and vegetable consumption (4,5). In addition, we examined the association of socioeconomic factors with diet and exercise practices. This information will be critical to assess the effectiveness of current clinical practice and help identify groups at high risk for poor dietary intake and physical inactivity.
RESEARCH DESIGN AND METHODS
The Third National Health and Nutrition Examination Survey (NHANES III), was conducted by the National Center for Health Statistics at 89 U.S. survey locations between 1988 and 1994. The survey used a stratified, multistage probability cluster design with oversampling of Mexican Americans, African Americans, and the elderly to ensure minimum sample sizes of these populations. The survey consists of multiple components including a household interview, a physical examination, and laboratory tests. Descriptions of the standardized protocols used for participant recruitment, interviews, and examinations have been previously published (15).
Information on a medical history of diabetes was obtained during the household interview among all adults over the age of 17 years (n = 20,005). A total of 1,608 adults reported a diagnosis of diabetes. For this analysis, we excluded women who were diagnosed with diabetes only during pregnancy (n = 105) and individuals with type 1 diabetes, defined as those who were diagnosed at <30 years of age and who had continuous insulin use since diagnosis (n = 23). The remaining 1,480 respondents were considered to have type 2 diabetes. Based on previous studies demonstrating that self-reported diagnosis is valid and reliable, we believe this sample accurately represents a national population of adults with type 2 diabetes (16,17). Individuals were considered to be treated with diet alone if they reported taking neither oral hypoglycemic medications nor insulin.
Standing height and weight were obtained during the physical exam for 1,312 individuals and were used to calculate BMI. Individuals were considered obese if their BMI was ≥30 kg/m2 and were considered overweight if their BMI was 25–29.9 kg/m2 (18). HbA1c values were collected during the laboratory exam and were available for 1,253 subjects.
Dietary assessment was based on data from both the food frequency questionnaire and a 24-h food recall (19,20). Food frequency information was collected during the household interview and has been shown to be a valid and reliable method for assessing average consumption (19). Respondents were asked how often over the past month they had eaten the specified food items. Foods were reported as number of items consumed per day, per week, per month, or never. All frequencies of consumption variables were standardized as times per day using the conversion factor of 30.4 days/month. If the frequency of consumption was reported as “never,” the value was recorded as zero. For this analysis, we created a combined variable for daily fruit and vegetable intake that includes all portions from both the fruit and vegetable groups. The fruit group included all fresh, frozen, dried, and canned fruit and fruit juices but excluded fruit drinks. The vegetable group included all raw or cooked fresh, frozen, or canned vegetables and juices (19). Information on the percentage of daily calories from total fat and saturated fat was based on a 24-h food recall interview. We used the cut points of five fruits and vegetables per day, 30% of daily calories from fat, and 10% of daily calories from saturated fat based on recommendations from national guidelines (4,5).
Individuals were classified as inactive if they did not report engaging in any of the following activities during the previous month: walking, jogging, bike riding, swimming, aerobics, dancing, calisthenics, gardening, lifting weights, or other physical activity outside of their occupation. Physical activity was classified as moderate or vigorous intensity based on metabolic equivalent intensity levels (21). Individuals were considered to fulfill national recommendations for physical activity if they reported five or more episodes per week of moderate-intensity physical activity or three or more episodes per week of vigorous-intensity physical activity (6). Those reporting some physical activity during the preceding month but not at the recommended levels were classified as obtaining insufficient physical activity. To assess disability and immobility, all respondents were asked if they had difficulty “walking for a quarter of a mile” and “walking up 10 steps without resting” (13).
Data were weighted to account for the unequal probability of selection that resulted from the survey cluster design, nonresponse, and oversampling of certain target populations (22,23). Sampling weights were used to calculate population estimates, and sampling strata and primary sampling units were accounted for to estimate variances and test for significant differences. Statistical analysis was performed using STATA 6.0 software (24) to take into account the complex sampling design. All results are presented as unweighted counts (n) and weighted percentages and odds ratios.
Bivariate and multivariate logistic regression analyses were used to determine associations between socioeconomic and health characteristics and 1) consumption of fewer than five servings of fruits and vegetables per day, 2) >30% of total calories from fat, 3) >10% of daily calories from saturated fat, and 4) level of physical activity. Pearson’s χ2 test was used to compare bivariate associations between socioeconomic and health characteristics and these nutritional and physical activity variables. Multivariate logistic analyses were used to assess the independent associations of socioeconomic and health characteristics with obtaining no regular physical activity, having a diet high in saturated fats, and reporting low consumption of fruits and vegetables.
Table 1 displays the population characteristics from this nationally representative sample of individuals with type 2 diabetes. The mean age was 61 years, and the majority of the sample was white. Almost one fifth had incomes below the federal poverty level, and 45% had less than a high school education. Eighty-two percent of the sample had a BMI ≥25 kg/m2, with 36% classified as overweight and 46% as obese. Almost one third of the sample reported no regular physical activity in the month before the survey, and another 38% reported an insufficient amount of physical activity. Forty-two percent of respondents reported consumption of 30–40% of their daily calories from fat, and 26% reported intakes of >40% of their daily calories from fat. Sixty-two percent of individuals reported eating fewer than five servings of fruits and vegetables per day, and 61% consumed >10% of their daily calories from saturated fats.
Bivariate associations between socioeconomic characteristics and level of physical activity are displayed in Table 2. Individuals over age 65 years, women, Mexican Americans, African Americans, and those using insulin were more likely to report engaging in no physical activity. Half of individuals with an income below the poverty level and 40% of those with less than a high school education reported no physical activity. Of individuals who reported no difficulty in walking a quarter-mile or in walking 10 steps without resting, 43% reported insufficient physical activity. BMI was not associated with level of physical activity. In a multivariate logistic regression, age ≥ 65 years, female sex, and low income were independently associated with inactivity (Table 4). Physical inactivity was not associated with level of education, BMI, or type of diabetes treatment in multivariate analyses. When ability to walk a quarter-mile was included in this multivariate model, age was no longer independently associated with reporting no physical activity (data not shown).
Bivariate associations between nutritional intake and socioeconomic and health variables are displayed in Table 3. Individuals ≥65 years were less likely to consume a high-fat diet and more likely to report eating more than five servings of fruits and vegetables per day. There were no associations between the three nutrition intake variables and sex, education, or type of diabetes treatment. In multivariate analysis, individuals ≥65 years and Mexican Americans were less likely to consume >10% of their daily calories from saturated fats and eat fewer than five servings of fruits and vegetables per day (Table 4).
The majority of adults with type 2 diabetes in this nationally representative sample ate a diet high in saturated fat and consumed fewer than the minimum daily recommended servings of fruits and vegetables. In addition, almost one third reported no regular physical activity, and another 38% reported insufficient levels of physical activity. Higher consumption of fruits and vegetables and diets lower in fat were more common among individuals over the age of 65 years and among Mexican Americans. Older individuals, women, and those with low incomes were more likely to report no regular exercise. The majority of this sample was overweight or obese, consistent with recent reports about the increasing prevalence of both diabetes and obesity in the U.S. (25).
The main limitations to this data are the potential biases introduced by the self-report of food frequency data, the 24-h food recall, and level of physical activity. In addition, we do not know the actual duration of physical activity. Previous studies have shown that white-collar workers were more active during leisure time than blue-collar workers, those who were retired, or homemakers (26). Because we do not have information on nonleisure physical activity, total activity levels may be underestimated. Despite these limitations, our study is consistent with other national data from the general U.S. population regarding poor dietary quality (27–29) and low levels of physical activity (30). The most recent Behavioral Risk Factor Surveillance System (BRFSS) survey reported that almost one third of adults did not engage in any physical activity and another third were not regularly active (30). In the National Health Interview Survey from 1990, only one third of people with diabetes reported exercising regularly (13).
Although poor nutrient intake among individuals with diabetes has been reported outside of the U.S. (9–12), we are not aware of any data on the dietary intake from a nationally representative sample of individuals with type 2 diabetes in the U.S. Our results are consistent with studies of dietary intake in the general population suggesting that a majority of Americans are consuming an inadequate quantity of fruits and vegetables (25,31). Only one quarter of the participants interviewed in the most recent BRFSS consumed the recommended five servings of fruits and vegetables (25). In addition, our data regarding healthier diets among Mexican Americans are consistent with previous reports suggesting that on average, Latinos have healthier diets than whites or African Americans (27,32,33).
Previous studies have reported low levels of physical activity (13) and higher levels of disability among individuals with diabetes than in the general population (34). Disability among individuals with diabetes is primarily from coronary heart disease and obesity, followed by poor vision (34). For individuals with significant disability, an individualized program for physical activity is essential. In our study, individuals who reported the inability to walk a quarter-mile were less likely to report regular exercise. However, >40% of those who reported that they could walk a quarter-mile engaged in insufficient physical activity and could be an important target population for interventions to increase exercise activity.
Healthy diets and regular physical activity can improve glycemic control among individuals with type 2 diabetes (35–38). Intervention studies have shown that physical activity plays an important role in glucose tolerance and insulin sensitivity (35). A recent meta-analysis found that, on average, moderate-intensity physical activity can reduce HbA1c by 0.6% among individuals with type 2 diabetes (39). According to data from the U.K. Prospective Diabetes Study, this is a level sufficient to reduce the risk of microvascular complications by 22% (40). Lifestyle modification involving nutrition and physical activity also has the potential to improve cardiovascular risk factors such as blood pressure and lipid levels (39,41). Because physical activity and suboptimal dietary behaviors are associated (42), targeting both exercise and diet may be needed for the optimal treatment of individuals with type 2 diabetes.
A recent systematic review of diabetes self-management studies suggests that tailored interventions actively involving patient participation can change lifestyle behaviors, including diet and exercise (43). Increasing physical activity levels can be achieved by a variety of means, including public health programs and community-based interventions (14). Primary care-based physical activity counseling is also moderately effective (44), although few studies have been specifically designed for individuals with type 2 diabetes. The current challenge is translating research findings into routine clinical practice and public health efforts to improve health outcomes for all individuals with type 2 diabetes (45).
We would like to thank Gail Harrison, PhD, for her assistance in analyzing the nutrition variables in NHANES III.
This paper was presented at the American Diabetes Association meeting in San Francisco, California, on June 17, 2002.
Address correspondence and reprint requests to Karin Nelson, MD, MSHS, VA Puget Sound Health Care System, 1660 South Columbian Way (S-111-GIMC), Seattle, WA 98108-1597. E-mail:.
Received for publication 15 April 2002 and accepted in revised form 10 July 2002.
A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.
- DIABETES CARE