Surveillance for Five Health Risks Among Older Adults -- United States, 1993-1997

Laurie A. Kamimoto, M.D., M.P.H. ^1,2
Alyssa N. Easton, Ph.D., M.P.H. ^1,3
Emmanuel Maurice, M.S., M.A.³
Corinne G. Husten, M.D., M.P.H.³
Carol A. Macera, Ph.D.⁴

Problem/Condition: Overweight, drinking and driving, inadequate fruit and vegetable consumption, physical inactivity, and smoking are associated with the leading causes of morbidity and mortality among older adults (i.e., persons aged greater than or equal to 65 years) in the United States.

Reporting Period: This report presents data from the Behavioral Risk Factor Surveillance System (BRFSS) for 1994-1997 and from the National Health Interview Survey (NHIS) for 1993-1995.

Description of Systems: BRFSS and NHIS are maintained by CDC and have been used for surveillance purposes. Each survey is administered annually and includes questions about health risks and health behaviors from a representative sample of the U.S. civilian, noninstitutionalized population. The NHIS is designed to provide national estimates and the BRFSS, state estimates.

Results: Prevalences of overweight, drinking and driving, inadequate fruit and vegetable consumption, and smoking decreased with increasing age among older adults in the United States; physical inactivity was the only health risk that increased with increasing age. Sex and race were differentially associated with all five health risks.

Interpretation: Specific subgroups of older adults are at risk for being overweight, drinking and driving, inadequate fruit and vegetable consumption, physical inactivity, and smoking. These health risks varied by age, race, residential state, and socio-economic status and highlight the heterogeneous nature of older adults.

Public Health Action: Surveillance for health risks among older adults provides information to help identify effective interventions for the growing population of older adults in the United States.

In the United States, 25% of the population is aged greater than or equal to 55 years, and 13% is aged greater than or equal to 65 years (1). Persons aged greater than or equal to 65 years represent the fastest growing segment of the U.S. population (2). During 1989-2030, the number of persons aged greater than or equal to 65 years is expected to double, and these persons are projected to account for 22% of the population in 2030 (3). Because aging has been associated with increased prevalence of chronic disease, disability, and death, a public health goal for older adults is to maintain health, independence, and function.

Poor health behaviors are leading contributors to morbidity and premature mortality in the United States (4). Excess bodyweight, drinking and driving, inadequate fruit and vegetable consumption, physical inactivity, and smoking are associated with the leading causes of death (5). In addition, chronic disease or disability (which persons can live with for many years and which can reduce quality of life) is associated with these health risk behaviors. Although the prevalence of health risk behaviors is usually lower among older adults (i.e., persons aged greater than or equal to 65 years) than among middle-aged adults, those behaviors still contribute to morbidity and mortality among older adults. However, increasing evidence indicates that behavior change even late in life is beneficial and can result in improved disease control and enhanced quality of life (6-12).

In this report, prevalence estimates among noninstitutionalized older adults in the United States are presented for five health risks: a) overweight, b) drinking and driving, c) inadequate fruit and vegetable consumption, d) physical inactivity, and e) smoking. In addition, prevalence estimates among adults aged 55-64 years are presented for comparison.

Two data sets were used for the analyses in this report: the Behavioral Risk Factor Surveillance System (BRFSS) and the National Health Interview Survey (NHIS). BRFSS data were used to assess the prevalences of overweight, drinking and driving, consumption of fruits and vegetables greater than or equal to 5 times per day, physical inactivity, and smoking among adults aged greater than or equal to 55 years. The BRFSS is an ongoing, state-based, random-digit-dialed telephone survey of U.S. civilian, noninstitutionalized persons aged greater than or equal to 18 years (13). This survey collects self-reported information regarding risks and behaviors related to health status (with the understanding that self-reports can overestimate or underestimate the prevalence of certain behaviors) (13). The BRFSS excludes households without telephones, but only approximately 2.5% of older adults do not have a phone (14). Institutionalized persons, who are also not included, most likely account for approximately 5% of adults aged greater than or equal to 65 years (15). Each year's survey consists of questions asked annually (e.g., overweight and smoking) and biannually* (e.g., alcohol use, physical activity, and fruit and vegetable consumption).

In this report, multiple years of BRFSS data were aggregated to increase the precision of prevalence estimates; data for 1994-1997 (the most recent years for which data were available) were used. Each year during 1994-1997, a total of 52 reporting areas (i.e., 50 states, the District of Columbia, and Puerto Rico) participated in BRFSS; data from Puerto Rico were not included in these analyses. Data were stratified by age group (55-64, 65-74, and greater than or equal to 75 years), then weighted to both the respondent's probability of selection and the distribution of each state's population by age, sex, and race**, according to current census or intercensal estimates (16,17). Data also were stratified by region***. Software for Survey Data Analysis (SUDAAN) was used to calculate all prevalence estimates and 95% confidence intervals (CIs) (18). Median prevalence estimates were based on combined data from the 50 states and the District of Columbia.**** Missing data or data coded as "don't know" or "refused" were excluded from analyses. State- and sex-specific estimates were considered statistically significant if the chi-square statistic corresponded to p less than 0.05. When multiple comparisons were made, the Bonferroni method was used (p less than 0.05/number of comparisons). Data were not reported when the standard error was greater than or equal to 30% of the prevalence estimate.

CDC's NHIS has collected information regarding tobacco use since 1965. For this report, NHIS data were used to assess national estimates of smoking prevalence, smokers' desire to quit smoking, and smoking cessation attempts among adults aged greater than or equal to 55 years. The NHIS is an ongoing, annual, cross-sectional household survey of the U.S. civilian, noninstitutionalized population. NHIS data are obtained through personal household interview. All household members present at the time of interview are invited to participate. For children and for adults who are unavailable or unable to participate because of illness or impairment, proxy responses are permitted for portions of the NHIS. The smoking questions are asked of one sample adult in each family, and self-response is required for these questions (19-21).

For this report, NHIS data for 1993-1995 (the most recent years for which data were available) were aggregated to increase the precision of the prevalence estimates. Data were stratified by age (55-64, 65-74, and greater than or equal to 75 years), race/ethnicity (non-Hispanic white, non-Hispanic black, and Hispanic*****), sex, education (less than or equal to 8, 9-11, 12, 13-15, and greater than or equal to 16 years), and region. Data were not reported when the standard error was greater than or equal to 30% of the prevalence estimate. SUDAAN was used to calculate prevalence estimates and 95% CIs. Estimates were considered statistically significant if the 95% CIs did not overlap.

Body mass index (BMI) was calculated by using self-reported height and weight measurements obtained from the combined 1996 and 1997 BRFSS results of 81,137 respondents aged greater than or equal to 55 years. For both years combined, sample sizes ranged from 556 (Alaska) to 2,635 (Florida). Weight in kilograms was divided by height in meters squared (weight [kg]/height squared [m²]) to obtain BMI. The National Institute of Health (NIH) defines overweight as a BMI of 25.0-29.9 kg/m² and obesity as a BMI greater than or equal to 30.0 kg/m² (22).****** For this report, these two categories were combined, and persons were classified as overweight if they met the definition for overweight or obesity (BMI greater than or equal to 25 kg/m²).

Another definition of overweight that has been used since the 1970s is based on the National Health and Nutrition Examination Survey II (NHANES II). Although the new NIH criteria was adopted in this country in 1998, the NHANES-derived definition has been included for comparison with older, existing data. For NHANES II, overweight was defined for men as BMI greater than or equal to 27.8 kg/m² and for women as BMI greater than or equal to 27.3 kg/m², which corresponds to the approximate 85th percentile of BMI for men and women aged 20-29 years (24).

The prevalence of driving while under the influence of alcohol was calculated by using the combined 1995******* and 1997 BRFSS results of 76,368 respondents aged greater than or equal to 55 years. Both years combined, sample sizes ranged from 409 (District of Columbia) to 2,837 (Maryland). Drinking and driving was assessed by participants' responses to two questions: a) "During the past month, have you had at least one drink of any alcoholic beverage such as beer, wine, wine coolers, or liquor?" and b) "During the past month, how many times have you driven when you've had perhaps too much to drink?" Respondents who reported greater than or equal to 1 time of drinking and driving during the previous month were classified as drinking and driving. Respondents who answered negatively to the first question or reported no instances of drinking and driving were classified as not drinking and driving. Drinking and driving data were collected by the 50 states and District of Columbia but could not be reported separately because the number of affected respondents was small, and therefore, estimates were unstable.

Frequency of fruit and vegetable intake per day was obtained from the combined 1994 and 1996 BRFSS results of 71,517 respondents aged greater than or equal to 55 years. Both years, sample sizes ranged from 548 (Alaska) to 2,683 (Florida). Consumption of fruits and vegetables was based on the following series of questions:

• "How often do you drink fruit juices such as orange, grapefruit, or tomato?"
• "Not counting juice, how often do you eat fruit?"
• "How often do you eat green salad?"
• "How often do you eat potatoes not including french fries, fried potatoes, or potato chips?"
• "How often do you eat carrots?"
• "Not counting carrots, potatoes, or salad, how many servings of vegetables do you usually eat?"

For each question, respondents indicated never or the number of times per day, week, month, or year that they ate the fruit or vegetable in question. From these questions, the average number of times per day that fruits and vegetables were eaten was calculated, then the prevalence of reported fruit and vegetable consumption of greater than or equal to 5 times per day was calculated.

The prevalence of physical inactivity was calculated by using the combined 1994 and 1996 BRFSS results of 71,517 respondents aged greater than or equal to 55 years. Both years combined, sample sizes ranged from 548 (Alaska) to 2,683 (Florida). Leisure-time physical inactivity was defined as a negative response to the question, "During the past month, did you participate in any physical activities or exercises such as running, calisthenics, golf, gardening, or walking for exercise?"

The prevalence of cigarette smoking was calculated by using the combined 1995-1997 BRFSS results of 116,690 respondents aged greater than or equal to 55 years. During this period, sample sizes ranged from 748 (District of Columbia) to 4,181 (Maryland). Before 1996, respondents were asked, "Have you smoked at least 100 cigarettes in your entire life?", and "Do you smoke cigarettes now?" to assess self-reported smoking status. Current smokers were defined as persons who reported having smoked greater than or equal to 100 cigarettes during their lifetime and who currently smoked. In 1996, the definition of current smoking changed. Respondents were asked, "Have you smoked at least 100 cigarettes in your entire life?", and "Do you now smoke cigarettes every day, some days, or not at all?" Current smokers were defined as persons who reported having smoked greater than or equal to 100 cigarettes during their lifetime and who currently smoked every day or some days. This change in the definition of current smoking resulted in an increase in prevalence rates of approximately 1% (25).

This report also uses data from NHIS to assess the prevalence of cigarette smoking. Data from NHIS were combined for 1993-1995; the total sample size was 17,754. Participants were asked, "Have you smoked at least 100 cigarettes in your entire life?", and "Do you now smoke cigarettes every day, some days, or not at all?" Current smokers were defined as persons who reported smoking greater than or equal to 100 cigarettes during their lifetime and who currently smoke every day or some days. Former smokers were defined as persons who reported having smoked greater than or equal to 100 cigarettes during their lifetime but who do not smoke currently. Persons who reported never smoking were defined as persons who reported smoking no cigarettes or smoking less than 100 cigarettes during their lifetime.

For the 50 states and District of Columbia, the prevalence of overweight (BMI greater than or equal to 25 kg/m²) decreased with increasing age among persons aged greater than or equal to 55 years (Table 1).******** For all age groups, the prevalence of overweight was similar among the four regions of the United States (Table 2). Although the prevalence of overweight varied among blacks and whites in most U.S. regions, in the South the prevalence was significantly greater among blacks than among whites in all age groups.

State-specific prevalence estimates of overweight ranged from 53.3% (95% CI=47.2%-59.4%) in Arizona to 71.7% (95% CI=63.3%-80.1%) in Alaska for persons aged 55-64 years (Table 1). For persons aged 65-74 years, prevalence estimates ranged from 39.6% (95% CI=34.4%-44.8%) in Hawaii to 73.4% (95% CI=64.6%-82.2%) in Alaska. For persons aged greater than or equal to 75 years, the prevalence ranged from 30.5% (95% CI=24.2%-36.8%) in Hawaii to 53.6% (95% CI=48.4%-58.8%) in North Dakota.

Men were significantly more likely than women to be overweight, especially among the younger age groups (Table 3). Among persons aged 55-64 years, men were significantly more likely than women to be overweight in 44 states. Among persons aged 65-74 years, sex-specific differences existed in 30 states and, among persons aged greater than or equal to 75 years, in 16 states.

The prevalence of overweight among whites was significantly different in each age group; the prevalence decreased with increasing age (Table 4). Among blacks, the prevalence of overweight was significantly lower among persons aged greater than or equal to 75 years than among persons in either of the two younger age groups.

Analyses by race, sex, and age indicated that the prevalence of overweight was similar among black men and white men but was higher among black women than white women in all age groups (Table 4). For blacks in the 55-64-year and 65-74-year age groups, a significantly greater percentage of women than men were overweight (BMI greater than or equal to 25 kg/m²); for blacks in the greater than or equal to 75-year age group, the prevalence among men and women was similar. For whites in all age groups, the prevalence of overweight was higher among men than women.

For the combined 50 states and District of Columbia, the prevalence of drinking and then driving was significantly greater among persons aged 55-64 years than among those aged 65-74 years and greater than or equal to 75 years. Of persons aged 55-64 years, 0.8% (95% CI=0.6%-0.9%) reported drinking and driving during the previous 30 days. Drinking and driving was reported by 0.4% (95% CI=0.3%-0.5) of persons aged 65-74 years and 0.2% (95% CI=0.1%-0.3%) of persons aged greater than or equal to 75 years.

The median percentage of persons who reported eating fruits and vegetables greater than or equal to 5 times per day increased with increasing age (Table 5). Of persons aged 55-64 years in the 50 states and District of Columbia, a median of 26.4% ate fruits and vegetables on average greater than or equal to 5 times daily; of persons aged 65-74 years and greater than or equal to 75 years, a median of 30.4% and 33.6%, respectively, ate fruits and vegetables greater than or equal to 5 times daily.

Compared with other regions of the United States, the prevalence of fruit and vegetable consumption greater than or equal to 5 times daily among persons aged 55-64 years was highest in the West, Northeast, and South and lowest in the Midwest. Among persons aged 65-74 years, the prevalence was highest in the Northeast and West and lowest in the Midwest and South. Among persons aged greater than or equal to 75 years, region-specific prevalences were similar (Table 6). In all three regions with sample sizes that permitted stratification by race (excludes the West), the prevalence of fruit and vegetable consumption was lower for blacks than whites in all age groups, except for adults aged greater than or equal to 75 years in the Midwest.

State-specific estimates of fruit and vegetable consumption greater than or equal to 5 times daily ranged from 16.9% (95% CI=13.2%-20.6%) in Mississippi to 38.1% (95% CI=33.2%-43.0%) in Arkansas for persons aged 55-64 years (Table 5). Among persons aged 65-74 years, prevalence estimates ranged from 17.4% (95% CI=13.5%-21.3%) in Mississippi to 41.9% (95% CI=38.0%-45.8%) in California. Among persons aged greater than or equal to 75 years, consumption of fruits and vegetables greater than or equal to 5 times daily ranged from 20.5% (95% CI=15.0%-26.0%) in Louisiana to 46.9% (95% CI=40.6%-53.2%) in Michigan.

Statistically significant age-specific differences in consumption of fruits and vegetables were observed between men and women (Table 7). Women were more likely than men to have eaten fruits and vegetables greater than or equal to 5 times daily, but this difference decreased with increasing age. For persons aged 55-64 years, significant sex-specific differences in prevalence existed in 32 states; for persons aged 65-74 years, in 23 states; and for persons aged greater than or equal to 75 years, in 10 states.

Compared with blacks, a greater proportion of whites in any age group reported eating fruits and vegetables greater than or equal to 5 times daily (Table 8). The proportion of whites who ate fruits and vegetables greater than or equal to 5 times daily increased with age. However, the age-specific proportion of blacks who reported eating fruits and vegetables greater than or equal to 5 times daily was similar by age group. Analysis of race, sex, and age indicated that women in all age groups reported eating fruits and vegetables greater than or equal to 5 times daily more frequently than men of the same race and age group, except for blacks aged greater than or equal to 75 years (Table 8).

For the combined 50 states and District of Columbia, the prevalence of leisure-time physical inactivity increased with increasing age. The prevalence of physical inactivity was similar for the two youngest age groups (age 55-64 years [median: 33.1%] and age 65-74 years [median: 35.1%]) (Table 9). Of persons aged greater than or equal to 75 years, a median of 46.0% reported no physical activity.

For all age groups, the prevalence of physical inactivity was lowest in the West and among the highest in the South, except for persons aged greater than or equal to 75 years (who had similar prevalences of physical inactivity in all four U.S. regions) (Table 10). In every region and for all age groups (except 65-74- and greater than or equal to 75-year-olds in the Northeast and persons of all ages in the West), the prevalence was higher for blacks than whites.

State-specific prevalence estimates of physical inactivity ranged from 23.8% (95% CI=20.5%-27.1%) in Washington to 57.9% (95% CI=53.6%-62.2%) in Kentucky for persons aged 55-64 years (Table 9). For persons aged 65-74 years, prevalence estimates ranged from 21.3% (95% CI=18.0%-24.6%) in Washington to 58.2% (95% CI=54.1%-62.3%) in Georgia. For persons aged greater than or equal to 75 years, the prevalence ranged from 26.1% (95% CI=21.8%-30.4%) in California to 68.2% (95% CI=63.9%-72.5%) in Kentucky.

The age-specific prevalence of physical inactivity was similar among men and women (Table 11); however, sex-specific differences were observed and occurred more frequently among persons in the oldest age group. For persons aged 55-64 years, sex-specific differences in the prevalence of physical inactivity existed in five states; in three of these states, the prevalence was higher for men than for women. Among persons aged greater than or equal to 75 years, the prevalence of physical inactivity was significantly different for men and women in 18 states; in all 18 states, the prevalence of physical inactivity was higher among women.

The prevalence of physical inactivity was higher among blacks than whites in every age group (Table 12). Among blacks, reported prevalence estimates of physical inactivity were similar for men and women in all age groups. Among whites, the prevalence of physical inactivity was higher for women than men, except for persons aged 55-64 years.

State-specific BRFSS data indicate that for the combined years 1995-1997, the median prevalence of current cigarette smoking was 21.2% among persons aged 55-64 years, 13.3% among persons aged 65-74 years, and 6.8% among persons aged greater than or equal to 75 years (Table 13). State-specific prevalences ranged from 11.8% (95% CI=9.3%-14.3%) in Utah to 29.3% (95% CI=26.3%-32.3%) in Kentucky among persons aged 55-64 years, from 7.7% (95% CI=6.1%-9.3%) in Oklahoma to 26.0% (95% CI=20.4%-31.6%) in Nevada among persons aged 65-74 years, and from 2.6% (95% CI=1.2%-4.0%) in Utah to 14.7% (95% CI=7.9%-21.5%) in Nevada among persons aged greater than or equal to 75 years. Across all age groups, the median prevalence of smoking was greater for men (aged 55-64 years [21.8%], aged 65-74 years [13.9%], and aged greater than or equal to 75 years [8.0%]) than for women (aged 55-64 years [20.1%], aged 65-74 years [12.9%], and aged greater than or equal to 75 years [6.3%]).

Nationwide, for the combined years 1993-1995, NHIS data indicate the prevalence of current cigarette smoking was 23.8% (95% CI=22.6%-25.0%) among persons aged 55-64 years, 15.2% (95% CI=14.3%-16.1%) among persons aged 65-74 years, and 8.4% (95% CI=7.5%-9.3%) among persons aged greater than or equal to 75 years (Table 14). Overall, the prevalence of current smoking was higher among men than women and among non-Hispanic blacks than non-Hispanic whites and Hispanics. Overall, the prevalence of current smoking was highest among persons with less education and among persons living below poverty status.********* The prevalence of current smoking declined with increasing age for all demographic groups.

Excess bodyweight is a health risk for coronary artery disease, some cancers, diabetes, hypertension, and osteoarthritis (26). Although the benefits of intentional weight loss are well documented for adults, few studies have examined weight loss among older adults. A limited number of studies suggest intentional weight loss among older adults might benefit disease management (6,7) and extend life expectancy (27). Although the findings in this report indicate that the prevalence of overweight decreased with increasing age among persons aged greater than or equal to 55 years, this finding might not be generalizable to the total U.S. population of adults aged greater than or equal to 55 years. Because the BRFSS samples only noninstitutionalized older adults (i.e., a healthier subgroup of older adults), the prevalence of overweight might be underestimated. In addition, the results in this report are affected by survival bias because older adults who participated in the study tended to have healthier lifestyles than older adults who had died.

In this report, the prevalence of overweight varied by state and region. The prevalence of overweight for adults in each of the three age groups was highest in Alaska or North Dakota and lowest in Arizona or Hawaii. Although prevalence estimates were lower in the West for all age groups, the West could not be statistically differentiated from the other three U.S. regions. Some of the high state-specific prevalences might reflect a limitation of the BMI measurement. Because of the inability to distinguish whether excess bodyweight is from fat or muscle, some physically fit persons might be grouped inappropriately with sedentary, overweight persons, who are at risk and who are the focus of this report.

In this analysis of BRFSS data, the prevalence of overweight was generally higher among men than women. Similar results have been reported by applying the new definition of overweight (BMI greater than or equal to 25 kg/m²) to existing NHANES II data (22).

Several sociodemographic characteristics have been associated with being overweight. In this report, a greater proportion of black women than white women was found to be overweight. Sociocultural factors (e.g., education, income, dietary preferences, and attitudes toward overweight) (28) might account for some of the racial difference observed. In addition, among U.S. adults, an inverse relation between overweight and education has been observed (29).

Obesity is a complex topic of study -- especially determining the relation between obesity and its associated mortality risk. In addition to examining potential risk factors for mortality associated with obesity, the distribution of body fat (e.g., wrist-to-hip or waist circumference) (26-33) and the amount of weight gained during adulthood (33) are also being examined. Although weight loss among obese persons has been advocated, a distinction must be made between intentional and unintentional weight loss and the effects of each. Some intentional methods of weight loss (e.g., crash dieting) can be harmful, and weight loss that is unintentional might reflect an underlying pathologic condition that could pose a mortality risk.

The usefulness of BMI in assessing bodyweight among older adults might be limited (33) because BMI is a ratio of weight-to-height that does not distinguish fat from muscle weight. Therefore, rather than focusing on an ideal bodyweight, modification of health behaviors that influence bodyweight -- particularly diet and physical activity -- should be emphasized as part of a comprehensive health assessment (22).

Categorization of overweight in this report was based on self-report of height and bodyweight and is subject to misreporting. Height is generally overreported by older adults (34), and bodyweight is generally underreported by overweight persons, especially by overweight women (34,35). A validation study of BRFSS documented the actual population mean BMI was 1.5 kg/m² greater than the mean BMI calculated from self-reported BRFSS data on height and weight (36). Therefore, the prevalence of overweight in this report might underestimate the actual prevalence of overweight among older adults.

Drinking and driving is a health risk associated with motor vehicle-related injury. In a 1993 study, drivers aged greater than or equal to 65 years increased their risk for a motor-vehicle crash by nearly threefold by drinking and driving (37). How alcohol mediates its effect on older persons -- through cognition, motor function, and memory -- is unknown and subject to ongoing research (38). In addition, prescription medication use, which is more prevalent among older adults, can potentiate the effect of alcohol and impair driving more than alcohol alone (38).

An encouraging finding is that the prevalence of alcohol use decreases with increasing age (30). An estimated 2%-4% of U.S. adults aged greater than or equal to 60 years meet the Diagnostic and Statistical Manual of Mental Disorders III criteria for alcohol abuse or dependence (39). This report examined only one component of alcohol abuse -- drinking and driving. Although the prevalence of alcohol use for the 50 states and District of Columbia combined was low (approximately 1%) for persons aged greater than or equal to 65 years, data were obtained from self-reports and, therefore, are subject to under-reporting.

Since 1991, a national educational program, 5-A-Day, has advocated that persons in the United States eat greater than or equal to 5 servings of fruits and vegetables daily. Diets rich in fruits and vegetables, which contain essential nutrients and vitamins, have been associated with a reduction in cardiovascular disease and some cancers (40-42). Fruits and vegetables also contribute dietary fiber, which has beneficial effects on the gastrointestinal tract. High-fiber diets have been associated with a reduction in diverticular disease and constipation, common problems among older adults (43).

The findings in this report indicate that a small percentage of persons reported eating fruits and vegetables greater than or equal to 5 times daily. However, the percentage increased with increasing age. This finding is consistent with those of other food-intake studies using the same or a different methodology (44-47).

The findings in this report also indicate that the prevalence of consumption of fruits and vegetables greater than or equal to 5 times daily varied by state and region. Mississippi or Louisiana had the lowest prevalence of daily fruit and vegetable consumption for the three age groups (55-64, 65-74, and greater than or equal to 75 years). The states with the highest reported consumption of fruits and vegetables included other southern states, the Midwest, and the West. Regional analyses for adults aged 55-64 and 65-74 years indicated the Midwest to have among the lowest prevalences of fruit and vegetable consumption and the West and Northeast to have among the highest prevalences. These state and regional differences might reflect local availability of fresh produce, socioeconomic conditions that enable the purchase and daily consumption of fruits and vegetables, or cultural patterns of food preparation and dietary preferences and suggest that regional comparisons might not adequately describe fruit and vegetable consumption in the United States.

Racial differences in consumption of fruits and vegetables greater than or equal to 5 times daily were observed in this and other food-intake reports (44,45). For every age group, a greater proportion of whites than blacks reported eating fruits and vegetables greater than or equal to 5 times daily. Race-specific differences might reflect economic status and cultural patterns that influence dietary practices. For example, although educational attainment was not analyzed in this report, it has been positively associated with dietary consumption of fruits and vegetables (44,46). This association might account for some of the race-specific differences observed in this study.

Findings of the sex-specific analysis in this report suggest that women were more likely than men to have eaten fruits and vegetables greater than or equal to 5 times daily. In contrast, findings of the Continuing Surveys of Food Intake by Individuals (CSFII) study, which used 3-day diet records, indicated that men aged greater than or equal to 60 years ate more servings of fruits and vegetables than women the same age (46). One possible explanation for the contrasting sex-specific difference in the findings of these two studies is that, in the CSFII study, the unit of measure was total servings per day; whereas, in the BRFSS questions about food intake, the unit of measure was number of times eaten per day. In addition, in CSFII, mixed dishes that included fruits or vegetables (e.g., stew) were included toward the summed total servings. Furthermore, part of the sex-specific discrepancy in the findings of food-frequency studies and diet-record studies can be accounted for if men actually eat fruits and vegetables less frequently than women but eat greater quantities at each occasion. However, when the number of servings of fruits and vegetables was normalized to 1000 calories, women had eaten a greater number of daily servings than men (47).

This analysis is subject to at least two limitations. First, because BRFSS respondents self-reported frequency of fruit and vegetable consumption for as long as 1 year preceding the interview, the data were subject to recall bias. A validation study of the BRFSS food-frequency methodology documented that BRFSS estimates were lower than those based on diet records (48). Although BRFSS estimates are not appropriate for assessing achievement of national goals (e.g., Health People 2000 objectives), the data serve an important function -- monitoring changes within state-specific populations over time. Second, the BRFSS data in this report assessed consumption of servings of individual fruits and vegetables (i.e., excluded mixed dishes that included fruits and/or vegetables) by older adults. Additional studies are needed to further examine factors affecting fruit and vegetable consumption in this population.

Altering dietary habits requires behavior change and enabling resources. In some instances, enabling resources might be as basic as adequate dentition, availability of fresh fruits and vegetables, and the financial means and transportation to purchase such produce.

The importance of vigorous physical activity in maintaining cardiovascular health is well documented (49). However, during the previous 10 years, several studies and the Surgeon General's Report on Physical Activity and Health concluded that moderate-intensity physical activity (e.g., brisk walking) also is associated with a variety of health benefits for adults of every age (50,51). These benefits include reduced morbidity and mortality associated with coronary heart disease; control of blood pressure, glucose levels, and cholesterol; and improved weight management. As a result of these documented benefits of moderate physical activity, several agencies and professional groups have recommended that all adults obtain a minimum of 30 minutes of moderate-intensity physical activity (equivalent to brisk walking) on most days of the week (50,51). Older adults particularly can benefit from the effects of physical activities that include strength training (52). Although moderate-intensity physical activity is attainable for most adults, less than 30% of the U.S. population attains this level, and older adults are even less active (53,54).

Lack of physical activity is also associated with several musculoskeletal problems that can negatively affect functional ability. Aging, which is associated with lower levels of physical activity, is also associated with loss of muscle mass (sarcopenia) and strength (55); physical inactivity is associated with bone loss and osteoporosis (9,56). Osteoporosis, sarcopenia, and muscle weakness (directly or indirectly) are risk factors for falls and, therefore, fractures, in older adults. As an overall prevention strategy, older adults are encouraged to remain active throughout aging to help preserve functional ability and prevent frailty.

Despite all the benefits of physical activity, approximately one third of all BRFSS respondents aged greater than or equal to 65 years in 1994 and 1996 reported no leisure-time physical activity during the month preceding the interview. Other physical activity studies have reported similar findings among older adults (54,57).

In this report, physical inactivity varied by state and region. The prevalence of physical inactivity for adults aged greater than or equal to 55 years was highest in Kentucky or Georgia and lowest in Washington or California. Regional analyses for persons aged 55-64 and 65-74 years support these state-specific findings, with southern states reporting among the highest levels of physical inactivity and western states reporting among the lowest levels. These differences might be associated with socioeconomic and regional factors that promote and provide opportunities for physical activity.

The finding in this report of higher levels of physical inactivity among blacks than whites is consistent with other physical activity studies (50,53,57). Functional impairment and disability are associated with inactivity and are more prevalent among black older adults than white older adults (58). The relation between race and physical activity might reflect differences in education or income, which are factors positively associated with physical activity (59,60).

Although persons of all ages experience barriers to physical activity, older adults are particularly vulnerable. For example, in a study of perceived neighborhood safety, persons aged greater than or equal to 65 years were less likely than younger persons to report walking as a leisure-time physical activity if they perceived their neighborhood to be unsafe; however, for younger adults, neighborhood safety did not affect leisure-time walking behavior (61). This finding is particularly important because walking is the most common type of physical activity among adults and has been associated with independent living in a longitudinal study (62) and with decreased morbidity (9).

Even persons with chronic diseases (e.g., arthritis) are encouraged to maintain some level of activity (63) or are prescribed physical activity as a part of rehabilitation (32). Nonetheless, a physical activity program should be individually tailored, should consider potential barriers and environmental safety for each person, and can require initial monitoring.

The prevalence estimates of physical inactivity in this report are not comparable with other published studies on physical activity because of methodologic issues. Many physical activity studies estimate the prevalence of those persons who did not meet the frequency and duration of physical activity as defined in the particular study; however, this report estimates the prevalence of no activity during leisure time. In addition, work-related activity, housework, yard work, and childcare activities are excluded from BRFSS and most other physical activity surveys and could vary by socioeconomic status. In addition, because this study is based on self-reports, the estimates are subject to misclassification of activity status.

Despite the lower prevalence of smoking among older adults, older smokers are at greater risks from smoking because they have smoked longer, tend to be heavier smokers, and are more likely to suffer from smoking-related illnesses (11). Older adults who smoke have increased risks for cardiovascular disease, stroke, cancer, and respiratory disease (64-74). Surveillance for smoking among older adults is instrumental in identifying groups at high risk and in planning and evaluating tobacco control interventions.

State-specific and national data regarding the prevalence of smoking among men and women aged 55-64 years indicate that the United States has not met the Healthy People 2000 objectives of reducing cigarette smoking among adults (aged greater than or equal to 18 years) to less than or equal to 15% (objective no. 3.4) (75). Among persons aged greater than or equal to 65 years, the prevalence of current smoking continues to decline with increasing age. This decline is the result of an increased prevalence of quitting and differential mortality among smokers and nonsmokers. Regardless of age group, the prevalence of smoking remains highest among black men.

To reduce the prevalence of smoking among older adults, smoking-cessation programs are needed that take into account ethnic and sex-specific differences in smoking behavior. During 1993-1995, participants of the NHIS were asked, "Would you like to completely stop/quit smoking cigarettes?" Responses for smokers aged greater than or equal to 55 years indicated racial/ethnic and sex-specific differences in the desire to quit. Overall, non-Hispanic blacks (66.2% [95% CI=60.2%-72.2%]) were more likely than non-Hispanic whites (60.6% [95% CI=58.2%-63.0%]) and Hispanics (57.8% [95% CI=47.7%-67.9%]) to indicate the desire to completely stop smoking cigarettes. In addition, non-Hispanic black (68.7% [95% CI=64.8%-72.6%]) and Hispanic (66.5% [95% CI=58.9%-74.1%]) women were more likely than non-Hispanic black (64.3% [95% CI=56.0%-72.6%]) and Hispanic (50.7% [95% CI=37.9%-63.5%]) men to indicate that they would like to completely stop smoking cigarettes (CDC, unpublished data, 1993-1995). However, because of the small sample sizes of Hispanic men and women, this finding should be interpreted with caution.

In addition, during the 1990 NHIS, current smokers were asked, "Have you ever made a serious attempt to stop smoking cigarettes?", "When was the start of your most recent serious quit attempt?", and "How long did you actually stay off cigarettes that time?". For current smokers, response categories used for the analysis were never tried to quit; tried to quit, but not in the past year; and tried to quit in the past year. For former smokers, response categories used for the analysis were abstinent less than 3 months, abstinent 3-11 months, abstinent 1-4 years, abstinent 5-9 years, and abstinent greater than or equal to 10 years. Findings indicated that women were more likely than men to be recent quitters. Among ever smokers aged greater than or equal to 55 years, women were more likely than men to be current smokers who had never tried to quit (15.8% versus 11.0%, respectively); tried to quit, but not in the past year (14.0% versus 10.6%, respectively); and tried to quit in the past year (8.5% versus 6.0%, respectively). Approximately half (46.5%) of ever smokers aged greater than or equal to 55 years were former smokers who had been abstinent for greater than or equal to 10 years. Men were more likely than women to be abstinent for greater than or equal to 10 years (51.6% versus 39.9%, respectively) (CDC, unpublished data, 1990).

Health-care providers should be informed that smoking-cessation counseling, even brief advice to quit smoking, can be effective in encouraging older persons to quit smoking. According to guidelines from the Agency for Health Care Policy and Research, all health-care practitioners should repeatedly and consistently deliver smoking-cessation interventions to their patients (76). In 1992, the NHIS Cancer Control Supplement documented that 59% of current smokers aged greater than or equal to 55 years examined by a physician during the year preceding the survey reported that they had been advised to quit smoking by their physician. Overall, smokers who reported that a physician advised them to quit smoking during the preceding year were significantly more likely to report planning to quit during the next 6 months than smokers who were not advised to quit (CDC, unpublished data, 1992).

Methods of smoking cessation vary across age groups and should be considered when planning smoking-cessation programs for older adults. In 1998, the American Lung Association, in conjunction with Yankelovich Partners, conducted a survey of adult smokers who had unsuccessfully attempted to quit smoking cigarettes. Methods most frequently used among adult smokers aged greater than or equal to 50 years included quitting "cold turkey" (68.0%), using the nicotine patch (47.0%), quitting gradually (slowly reducing the number of cigarettes smoked) (41.0%), and chewing regular gum (27.0%). Smokers aged greater than or equal to 50 years were significantly less likely than smokers aged 18-34 years to chew regular gum as a method of smoking cessation. Older smokers (aged greater than or equal to 50 years) were significantly more likely than smokers aged 18-34 years to choose the nicotine patch prescribed by a doctor (20.0%) (77).

The prevalence estimates of smoking among older adults in this report are subject to at least two limitations. First, some sample sizes were inadequate to assess differences among the three age stratifications by demographic subgroups. Second, the BRFSS and the NHIS randomly sample the noninstitutionalized adult population, which is in better health compared with older adults who are institutionalized. Approximately 5% of persons aged greater than or equal to 65 years and 20% of persons aged greater than or equal to 85 years live in long-term care institutions (78).

Projected increases in the population of older adults ensure continued medical and economic costs associated with smoking. Public health initiatives and smoking-cessation programs aimed at older adults are necessary to reduce premature morbidity and mortality later in life.

Prevalences of the five health risks discussed in this report are affected by aging processes, including survival bias (i.e., persons with healthier life practices outliving those with less healthy lifestyles). In this report, the prevalences of drinking and driving, overweight, and smoking decreased with increasing age; the prevalence of fruit and vegetable consumption increased with advancing age. These trends might be associated with a combination of changes in behavior as age increases and with survival bias. However, physical inactivity increased with advancing age, and this finding most likely reflects the greater prevalence of chronic diseases and disability that are also associated with increasing age (79).

Older adults in the United States are a heterogeneous group that differ in health risks by age, sex, race, residential state, and socioeconomic status. Similarly, interventions need to be tailored for specific communities or groups to effectively change behavior and reduce health risks.

Overweight, drinking and driving, inadequate fruit and vegetable consumption, physical inactivity, and smoking are associated with the development of many diseases and injuries. Some diseases are strongly associated with a particular health risk (e.g., lung cancer's association with cigarette smoking); other disease etiologies are multifactorial, with each health risk contributing additional risk of disease occurrence (e.g., the combined effects of smoking, physical inactivity, excess bodyweight, and poor diet on cardiovascular disease). This multifactorial nature can result in confounding of one health risk by another because persons who have one health risk are more likely to have another (80). In addition, interventions to modify one health risk can effectively reduce another (e.g., increasing physical activity can result in loss of excess weight). Therefore, caution should be exercised when interpreting the benefits of health risk reduction for a disease because the benefit could be attributable to reduction of another health risk or a combination thereof.

More intervention studies, many of which use a multifactorial approach, are involving older adults. These studies are limited in number but have the encouraging message that altering poor health behaviors and reducing health risks, even late in life, is beneficial. As more of these special population studies are conducted, planning better interventions for older adults will be possible.

As the population of older adults increases and changes, ongoing surveillance is needed to monitor the prevalence of health risks in this population. Surveillance data can be used for public health policy development and as a tool to assess outcomes. However, future surveys might need to be expanded to adequately represent the diversity of older adults in the United States.

Special thanks to Mary Serdula, M.D., Division of Nutrition and Physical Activity, National Center for Chronic Disease Prevention and Health Promotion.

References

1. US Bureau of the Census. Statistical abstract of the United States, 1996. 116th ed. Washington, DC: Department of Commerce, Economics and Statistics Administration, 1996.
2. US Bureau of the Census. Statistical abstract of the United States, 1997. 117th ed. Washington, DC: Department of Commerce, Economics and Statistics Administration, 1997.
3. US Senate Special Committee on Aging/American Association of Retired Persons/Federal Council on the Aging/US Administration on Aging. Aging America: trends and projections, 1991 edition. Washington, DC: US Department of Health and Human Services, 1991; DHHS publication no. (FCoA)91-28001.
4. McGinnis JM, Foege WH. Actual causes of death in the United States. JAMA 1993;270:2207-12.
5. Desai MM, Zang P, Hennessy CH. Surveillance for morbidity and mortality among older adults -- United States, 1995-1996. In: Surveillance for selected public health indicators affecting older adults, CDC surveillance summaries. MMWR 1999;48(no. SS-8):7-25.
6. Hankey CR, Wallace AM, Lean MEJ. Plasma lipids, dehydroepiandosterone sulphate and insulin concentrations in elderly overweight angina patients, and effect of weight loss. Int J Obes Relat Metab Disord 1997;21:72-7.
7. Reaven GM, and Staff of the Palo Alto GRECC Aging Study Unit. Beneficial effect of moderate weight loss in older patients with non-insulin-dependent diabetes mellitus poorly controlled with insulin. J Am Geriatr Soc 1985;33:93-5.
8. Whelton PK, Appel LJ, Espeland MA, et al. Sodium reduction and weight loss in the treatment of hypertension in older persons: a randomized controlled Trial of Nonpharmacologic Interventions in the Elderly (TONE). JAMA 1998;279:839-46.
9. Elward K, Larson EB. Benefits of exercise for older adults: a review of existing evidence and current recommendations for the general population. Clin Geriatr Med 1992;8:35-50.
10. McCarron DA, Oparil S, Chait A, et al. Nutritional management of cardiovascular risk factors: a randomized clinical trial. Arch Intern Med 1997;157:169-77.
11. Rimer BK, Orleans CT, Keintz MK, Cristinzio S, Fleisher L. The older smoker: status, challenges and opportunities for intervention. Chest 1990;97:547-53.
12. Hirdes JP, Maxwell CJ. Smoking cessation and quality of life outcomes among older adults in the Campbell's survey on well-being. Can J Public Health 1994;85:99-102.
13. Powell-Griner E, Anderson JE, Murphy W. State- and sex-specific prevalence of selected characteristics--Behavioral Risk Factor Surveillance System, 1994 and 1995. In: CDC surveillance summaries (August 1). MMWR 1997;46(No. SS-3):1-31.
14. US Bureau of the Census. Phoneless in America [Statistical brief]. Washington, DC: Department of Commerce, Economics and Statistics Administration, 1994; publication no. SB/94-16.
15. US Bureau of the Census. National population estimates. Available on the Internet at http://www.census.gov/population/www/estimates/uspop.html#monthly. Accessed July 12, 1999.
16. Frazier EL, Franks AL, Sanderson LM. Behavioral risk factor data. In: Using chronic disease data: a handbook for public health practitioners. Atlanta, GA: US Department of Health and Human Services, Public Health Service, CDC;1992:4-1-4-17.
17. Gentry EM, Kalsbeek WD, Hogelin GC, et al. The behavioral risk factor surveys: II. Design, methods, and estimates from combined state data. Am J Prev Med 1985;1:9-14.
18. Shah BV. Software for survey data analysis (SUDAAN). Version 7.0. [Software documentation]. Research Triangle Park, NC: Research Triangle Institute, 1989.
19. National Center for Health Statistics. Health interview survey procedures, 1957-74. Hyattsville, MD: US Department of Health, Education, and Welfare, Public Health Service, Health Resources Administration, 1975; DHEW publication no. (HRA)75-1311. (Vital and health statistics; series 1, no. 11).
20. National Center for Health Statistics, Kovar MG, Poe GS. The National Health Interview Survey design, 1973-84, and procedures, 1975-83. Hyattsville, MD: US Department of Health and Human Services, Public Health Service, 1985; DHHS publication no. (PHS)85-1320. (Vital and health statistics; series 1, no. 18).
21. National Center for Health Statistics, Massey JT, Moore TF, Parsons VL, Tadros W. Design and estimation for the National Health Interview Survey, 1985-94. Hyattsville, MD: US Department of Health and Human Services, Public Health Service, CDC, 1989; DHHS publication no. (PHS)89-1384. (Vital and health statistics; series 2, no. 110).
22. National Heart, Lung, and Blood Institute. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: the evidence report. Bethesda, MD: US Department of Health and Human Services, National Institutes of Health, 1998.
23. World Health Organization. Obesity: preventing and managing the global epidemic -- report of a WHO consultation on obesity. Geneva: World Health Organization, 1998; publication no. WHO/NUT/NCD/98.1.
24. Najjar MF, Rowland M. Anthropometric reference data and prevalence of overweight, United States, 1976-80. Hyattsville, MD: US Department of Health and Human Services, Public Health Service, CDC, 1987; DHHS publication no. (PHS)87-1688. (Vital and health statistics; series 11, no. 238).
25. CDC. Cigarette smoking among adults -- United States, 1992, and changes in the definition of current cigarette smoking. MMWR 1994;43:342-6.
26. Pi-Sunyer FX. Medical hazards of obesity. Ann Intern Med 1993;119:655-60.
27. Lean MEJ, Powrie JK, Anderson AS, Garthwaite PH. Obesity, weight loss and prognosis in type 2 diabetes. Diabet Med 1990;7:228-33.
28. Flegal KM, Carroll MD, Kuczmarski RJ, Johnson CL. Overweight and obesity in the United States: prevalence and trends, 1960-1994. Int J Obes Relat Metab Disord 1998;22:39-47.
29. Kuczmarski RJ, Flegal KM, Campbell SM, Johnson CL. Increasing prevalence of overweight among US adults: the National Health and Nutrition Examination Surveys, 1960 to 1991. JAMA 1994;272:205-11.
30. Myers HF, Kagawa-Singer M, Kumanyika SK, Lex BW, Markides KS. Panel III: behavioral risk factors related to chronic disease in ethnic minorities. Health Psychol 1995;14:613-21.
31. Pamuk E, Makuc D, Heck K, Reuben C, Lochner K. Health, United States, 1998, with socioeconomic status and health chartbook. Hyattsville, MD: US Department of Health and Human Services, CDC, 1998:114-5; DHHS publication no. (PHS)98-1232.
32. Fletcher GF. How to implement physical activity in primary and secondary prevention: a statement for healthcare professionals from the Task Force on Risk Reduction, American Heart Association. Circulation 1997;96:355-7.
33. Solomon CG, Manson JE. Obesity and mortality: a review of the epidemiologic data. Am J Clin Nutr 1997:66(suppl):1044S-1050S.
34. Rowland ML. Reporting bias in height and weight data. Stat Bulletin (April-June) 1989:2-11.
35. Jeffery RW. Bias in reported body weight as a function of education, occupation, health and weight concern. Addict Behav 1996;21:217-22.
36. Jackson C, Jatulis DE, Fortmann SP. The Behavioral Risk Factor Surveillance Survey and the Stanford Five-City Project Survey: a comparison of cardiovascular risk behavior estimates. Am J Public Health 1992;82:412-6.
37. Thompson DC, Rivara FP, Thompson RS, Salzberg PM, Wolf ME, Pearson DC. Use of behavioral risk factor surveys to predict alcohol-related motor vehicle events. Am J Prev Med 1993;9:224-30.
38. Dufour MC, Archer L, Gordis E. Alcohol and the elderly. Clin Geriatr Med 1992;8:127-41.
39. Adams WL, Cox NS. Epidemiology of problem drinking among elderly people. International Journal of Addiction 1995;30:1693-716.
40. Gazino JM, Manson JE, Buring JE, Hennekens CH. Dietary and anti-oxidants and cardiovascular disease. Ann N Y Acad Sci 1992;669:249-59.
41. Block G, Patterson BH, Subar AF. Fruit, vegetables and cancer prevention: a review of the epidemiologic evidence. Nutr Cancer 1992;18:1-29.
42. Steinmetz KA, Potter JD. Vegetables, fruit, and cancer, I: epidemiology. Cancer Causes Control 1992;2:325-57.
43. Schaefer DC, Cheskin LJ. Constipation in the elderly. Am Fam Physician 1998;58:907-14.
44. Serdula MK, Coates RJ, Byers T, Simoes E, Mokdad AH, Subar AF. Fruit and vegetable intake among adults in 16 states: results of a brief telephone survey. Am J Public Health 1995;85:236-9.
45. Patterson BH, Block G, Rosenberger WF, Pee D, Kahle LL. Fruit and vegetables in the American diet: data from the NHANES II Survey. Am J Public Health 1990;80:1443-9.
46. McClelland JW, Demark-Wahnefried W, Mustian RD, Cowan AT, Campbell MK. Fruit and vegetable consumption of rural African Americans: baseline survey results of the Black Churches United for Better Health 5-A-Day Project. Nutr Cancer 1998;30:148-57.
47. Krebs-Smith SM, Cook DA, Subar AF, Cleveland L, Friday J. US adults' fruit and vegetable intakes, 1989 to 1991: a revised baseline for the Healthy People 2000 objective. Am J Public Health 1995;85:1623-9.
48. Smith-Warner SA, Elmer PJ, Fosdick L, Tharp TM, Randall B. Reliability and comparability of three dietary assessment methods for estimating fruit and vegetable intakes. Epidemiology 1997;8:196-201.
49. Powell KE, Thompson PD, Caspersen CJ, Kendrick JS. Physical activity and the incidence of coronary heart disease. Annu Rev Public Health 1987;8:253-87.
50. US Department of Health and Human Services. Physical activity and health: a report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Public Health Service, CDC, 1996.
51. Pate RR, Pratt M, Blair SN, et al. Physical activity and public health: a recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. JAMA 1995;273:402-7.
52. American College of Sports Medicine. Exercise and physical activity for older adults: position stand. Med Sci Sports Exerc 1998;30:992-1008
53. Jones DA, Ainsworth BE, Croft JB, Macera CA, Lloyd EE, Yusuf HR. Moderate leisure-time physical activity: Who is meeting the public health recommendations? Arch Family Med 1998;7:285-9.
54. Yusuf HR, Croft JB, Giles WH, et al. Leisure-time physical activity among older adults, United States, 1990. Arch Intern Med 1996;156:1321-6.
55. Evans WJ. What is sarcopenia? J Gerontol A Biol Sci Med Sci 1995;50A:5-8.
56. Wagner EH, LaCroix AZ, Buchner DM, Larson EB. Effects of physical activity on health status in older adults. I. observational studies. Annu Rev Public Health 1992;13:451-68.
57. Crespo CJ, Keteyian SJ, Heath GW, Sempos CT. Leisure-time physical activity among US adults: results from the third National Health and Nutrition Examination Survey. Arch Intern Med 1996;156:93-8.
58. US Senate Special Committee on Aging/American Association of Retired Persons/Federal Council on the Aging/US Administration on Aging. Aging America: trends and projections, 1991 edition. Washington, DC: US Department of Health and Human Services, 1991:154; DHHS publication no. (FCoA)91-28001.
59. Washburn RA, Kline G, Lackland DT, Wheeler FC. Leisure time physical activity: are there black/white differences? Prev Med 1992;21:127-35.
60. Clark DO. Racial and educational differences in physical activity among older adults. Gerontologist 1995;35:472-80.
61. CDC. Neighborhood safety and the prevalence of physical inactivity -- selected states, 1996. MMWR 1999;48:143-6.
62. Strawbridge WJ, Cohen RD, Shema SJ, Kaplan GA. Successful aging: predictors and associated activities. Am J Epidemiol 1996;144:135-41.
63. Arthritis Foundation. Arthritis information: exercise and your arthritis, 1998. Available at http://www.arthritis.org. Accessed February 22, 1999.
64. LaCroiz AZ, Omenn GS. Older adults and smoking. Clin Geriatr Med 1992;8:69-87.
65. Agner E. Smoking and health in old age: a ten-year follow-up study. Acta Med Scand 1985;218:311-6.
66. Kaplan GA, Seeman TE, Cohen RD, Knudsen LP, Guralnik J. Mortality among the elderly in the Alameda County Study: behavioral and demographic risk factors. Am J Public Health 1987;77:307-12.
67. LaCroiz AZ. Smoking and mortality among older men and women in three communities. N Engl J Med 1991;324:1619-25.
68. US Department of Health and Human Services. The health consequences of smoking: chronic obstructive lung diseases -- report of the Surgeon General. Rockville, MD: US Department of Health and Human Services, Public Health Service, Office on Smoking and Health, 1984; DHHS publication no. (PHS)84-50205.
69. Jajich CL, Ostfeld AM, Freeman DH. Smoking and coronary heart disease mortality in the elderly. JAMA 1984;252:2831-4.
70. Khaw KT, Barrett-Connor E, Suarez L, Criqui MH. Predictors of stroke-associated mortality in the elderly. Stroke 1984;15:244-8.
71. Shinton R, Beevers G. Meta-analysis of relation between cigarette smoking and stroke. BMJ 1989;298:789-94.
72. Harris T, Cook EF, Kannel WB, Goldman L. Proportional hazards analysis of risk factors for coronary heart disease in individuals aged 65 or older: the Framingham Heart Study. J Am Geriatr Soc 1988;36:1023-8.
73. US Department of Health and Human Services. Reducing the health consequences of smoking: 25 years of progress -- a report of the Surgeon General, 1989, executive summary. Rockville, MD: US Department of Health and Human Services, Public Health Service, CDC, Office on Smoking and Health, 1989; DHHS publication no. (CDC)89-8411.
74. Husten CG, Shelton DM, Chrismon JH, Lin YW, Mowery P, Powell F. Cigarette smoking and smoking cessation among older adults: United States, 1965-94. Tobacco Control 1997;6:175-80.
75. US Department of Health and Human Services. Healthy people 2000 review, 1998-99. Hyattsville, MD: US Department of Health and Human Services, CDC, National Center for Health Statistics, 1999; DHHS publication no. (PHS)99-1256.
76. Fiore MC, Bailey WC, Cohen SJ, et al. Smoking cessation (Clinical Practice Guideline no. 18]. Rockville, MD: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, 1996; AHCPR publication no. 96-0692.
77. American Lung Association. Smoking Cessation Study. Available at http://www.lungusa.org/index2.html. Accessed November 2, 1998.
78. Hickey T, Speers M, Prohaska TR, eds. Public health and aging. Baltimore, MD: The Johns Hopkins University Press, 1997:81.
79. CDC. Comorbidity of chronic conditions and disability among older persons -- United States, 1984. MMWR 1989;38:788-91.
80. Serdula MK, Byers T, Mokdad AH, Simoes E, Mendlein JM, Coates RJ. The association between fruit and vegetable intake and chronic disease risk factors. Epidemiology 1996;7:161-5.

* Questions about alcohol use have been included in BRFSS only in odd-numbered years; questions about physical activity and fruit and vegetable consumption have been included only in even-numbered years.

** Race-specific data are presented only for blacks and whites because sample sizes for other racial groups were too small for meaningful analysis.

*** Northeast=Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont; Midwest=Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin; South=Alabama, Arkansas, Delaware, District of Columbia, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia; and West=Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming.

**** The median is the midpoint of a set of measurements arranged in order of magnitude and, unlike the mean, is less influenced by extreme high or low measurements.

***** Persons of Hispanic origin can be of any race.

****** The NIH definition is the same as the World Health Organization's international definition of overweight and obesity (23).

******* The District of Columbia did not administer the 1995 BRFSS.

******** Data presented in Tables 1-4 use a definition of overweight that combines the NIH definitions of overweight and obesity. For comparison of older, existing NHANES II data with the findings in this report, Tables A-D in the Exhibit at the end of this report present data using the NHANES II definition of overweight and obesity.

********* Poverty statistics are based on definitions developed by the Social Security Administration in 1964 (which were subsequently modified by federal interagency committees in 1969 and 1980) and prescribed by the Office of Management and Budget as the standard to be used by federal agencies for statistical purposes.