Subgroup-specific effects of questionnaire wording on population-based estimates of mammography prevalence.

OBJECTIVES: This study investigated whether an apparent downturn in prevalence rates of mammography use reported in the 1992 Behavioral Risk Factor Surveillance System (BRFSS) questionnaire resulted from a change in questionnaire wording. METHODS: In a pretest-posttest design (1990-1991 vs 1992), piecewise linear regression analyses were based on monthly prevalence estimates of mammography use among female BRFSS respondents 40 years or older. RESULTS: Self-reported mammography use was lower by 3.5 percentage points (95% confidence interval [CI] = 1.5, 5.5) overall--and lower by 13.6 percentage points (95% CI = 2.6, 24.6) among Black women with less than a high school education--when predicted from 1992 data than when predicted from 1990-1991 data. CONCLUSIONS: A change in questionnaire wording in the BRFSS caused demographic-specific effects in population-based estimates of mammography use.

Assessing alcohol consumption: beverage-specific versus grouped-beverage questions.

OBJECTIVE: The quantity-frequency method is commonly used to measure alcohol intake in large surveys. Because of time and space constraints, questionnaires are often shortened by combining questions on all types of alcohol into a single question. We investigated the effect of this practice using data from the Behavioral Risk Factor Surveillance System. METHOD: We examined data collected from 213,842 respondents to surveys conducted by 32 states and the District of Columbia participating in the years 1987, 1988, 1989 and 1990. The 1987 and 1988 surveys asked questions about respondents' frequency and level of intake of specific alcohol-containing beverages. The 1989 and 1990 surveys asked about the frequency and quantity of intake of alcohol-containing beverages by combining all beverages into a single group. RESULTS: Among drinkers, the mean number of drinks per month was higher for those who were asked beverage-specific questions than for those who were asked grouped-beverage questions (men: 37.0 vs 29.6; women: 17.0 vs 13.9). CONCLUSION: Caution must be used in comparing level of alcohol intake from surveys in which beverages are not grouped identically.

Patterns of health risk behaviors for chronic disease: a comparison between adolescent and adult American Indians living on or near reservations in Montana.

PURPOSE: To compare the chronic disease health risk behavior patterns of adolescents and adults among American Indians living on or near reservations in Montana. METHODS: We analyzed data from the 1993 Youth Risk Behavior Survey of American Indians in Grades 9-12 living on or near Montana reservations. Risk factors included tobacco use, low physical activity, attempted weight loss, and low consumption of fruits, vegetables, and green salad. Similar data were analyzed from a 1994 Behavioral Risk Factor Survey of American Indian adults living on or near reservations in Montana. RESULTS: The prevalence of most adolescent health risk behaviors was high, especially cigarette smoking (45% for males, 57% for females), smokeless tobacco use (44% for males, 30% for females), and infrequent consumption of salad or vegetables (59-76%). With the exception of daily cigarette smoking and inadequate fruit consumption among adolescents of both genders and physical inactivity among adolescent males, the prevalence of chronic disease health risk behaviors among adolescents was similar to or higher than the prevalence of the same risk behaviors among adults. CONCLUSIONS: Many health risk behaviors for chronic diseases are common by the time this group of American Indians in Montana has reached adolescence. Possible reasons may include modeling of familial behaviors, peer pressure, advertising, or age cohort effects. If these risk behavior patterns continue into adulthood, morbidity and mortality from chronic diseases are likely to remain high. Substantial efforts are needed to prevent or reduce health risk behaviors among adolescents and adults in this population.

Prevalence of tobacco use among Alaska Natives: a review.

BACKGROUND: Previous reports documented high rates of tobacco use among Alaska Natives (Eskimos, Indians, and Aleuts). In this population, tobacco use is the leading preventable cause of death. Lung cancer is the leading cause of cancer death among Alaska Natives and tobacco is responsible for over one-third of all cancer deaths in this population. Until recently there has been no systematic surveillance of the prevalence of tobacco use in this high-risk population. Data that did exist were not readily available to those primarily responsible for the health care of this population. This is the first time since 1990 that data on Alaska Natives have been collected in one analysis; this permits a more representative evaluation of tobacco use. METHODS: Data on tobacco use were obtained and analyzed from national and state surveys and selected research projects from 1988 to 1993. RESULTS: Alaska Natives have high prevalence of tobacco use, including both cigarettes and smokeless tobacco. Tobacco use prevalence among Alaska Natives exceeds that of Alaska non-Natives, U.S. whites, and American Indians/Alaska Natives in the United States outside of Alaska. Smoking prevalence among Alaska Native women is twice that of non-Native women in Alaska and nearly twice as high among pregnant Alaska Natives than pregnant non-Natives. Overall, prevalence of smokeless tobacco use was four times higher among Alaska Natives than comparative state and national populations. CONCLUSION: Because this population has such high rates of tobacco use, it is important to public health that monitoring and educational programs be in place and that data specific to Alaska Natives be made available.

Prevalence of alcohol-impaired driving. Results from a national self-reported survey of health behaviors.

OBJECTIVE: To estimate how frequently adults in the United States drive while impaired by alcohol. DESIGN: Telephone survey. SETTING: The 49 states (and the District of Columbia) that participated in the Behavioral Risk Factor Surveillance System (BRFSS) in 1993. PARTICIPANTS: A total of 102,263 noninstitutionalized adults aged 18 years or older. MAIN OUTCOME MEASURES: The percentage of respondents who reported alcohol-impaired driving; number of episodes of alcohol-impaired driving per 1000 adult population; and total number of episodes of alcohol-impaired driving-each by age, sex, race, level of education, and state. RESULTS: Overall, 2.5% of adults reported an estimated 123 million episodes of alcohol-impaired driving in 1993. This corresponds to 655 episodes of alcohol-impaired driving for each 1000 adults (range among states per 1000 adults, 165-1550). Alcohol-impaired driving was most frequent among men aged 21 to 34 years (1739 episodes per 1000 adults) and was nearly as frequent among men aged 18 to 20 years (1623 episodes per 1000 adults), despite legislation in all states that prohibited the sale of alcohol to persons younger than age 21 years in 1993. CONCLUSIONS: Alcohol-impaired driving is common even among underage persons. Strict enforcement of laws that discourage alcohol-impaired driving is needed along with community and patient education to reduce the prevalence of alcohol-impaired driving and prevent injuries and deaths from alcohol-related motor vehicle crashes. Data from the BRFSS, an ongoing source of national and state-specific data on the number of episodes of alcohol-impaired driving, are potentially useful for monitoring trends and evaluating the effect of future efforts to reduce alcohol-impaired driving.

Trends in overweight among US adults from 1987 to 1993: a multistate telephone survey.

OBJECTIVES: Using data from the Behavioral Risk Factor Surveillance System, this study describes trends in the prevalence of overweight between 1987 and 1993. METHODS: Data were examined from 33 states participating in an ongoing telephone survey of health behaviors of adults (n = 387,704). Self-reported weights and heights were used to calculate sex-specific prevalence estimates of overweight for each year from 1987 to 1993. Time trends were evaluated with the use of linear regression. RESULTS: Between 1987 and 1993, the age-adjusted prevalence of overweight increased by 0.9% per year for both sexes (from 21.9% to 26.7% among men and from 20.6% to 25.4% among women). The increasing linear trend was observed in all subgroups of the population but was most notable for Black men (1.5% per year) and men living in the Northeast (1.4% per year). Secular changes in smoking and leisure-time physical activity did not entirely account for the increase in overweight. CONCLUSIONS: The prevalence of overweight among American adults increased by 5% between 1987 and 1993. Efforts are needed to explore the causes of this adverse trend and to find effective strategies to prevent obesity.

Trends in smokeless tobacco use among men in four states, 1988 through 1993.

OBJECTIVES: Trends in smokeless tobacco use were examined for men in Indiana, Iowa, Montana, and West Virginia from 1988 through 1993. METHODS: State survey data from the Behavioral Risk Factor Surveillance System (BRFSS) were analyzed. RESULTS: Demographic characteristics associated with smokeless tobacco use included age less than 35 years, a high school education or less, and rural residence. Overall, there was little change in smokeless tobacco use among men in these states (range = -0.4-0.4 percentage points annually); only West Virginia had a significant decline. CONCLUSIONS: Reasons for the overall lack of decline may include increased advertising and promotional expenditures or substitution of smokeless tobacco for cigarettes. Increased prevention and cessation efforts are needed.

Radon awareness and testing behavior: findings from the Behavioral Risk Factor Surveillance System, 1989-1992.

Radon is considered an important environmental risk factor for lung cancer. Previous studies have shown that relatively few homes have been tested for radon. Results from the Behavioral Risk Factor Surveillance System from 1989--1992 show significant interstate and demographic variation in levels of radon awareness and home testing. There was evidence that levels of radon awareness and testing homes for radon have increased from 1989 through 1992. Although these trends in the data are encouraging, the data also suggest that continued education and other interventions may be necessary to reach the Public Health Service's testing and mitigating objectives for residential radon.

The epidemiology of walking for exercise: implications for promoting activity among sedentary groups.

The relative contribution of walking to overall leisure-time physical activity participation rates was studied among respondents from the 45 states that participated in the 1990 Behavioral Risk Factor Surveillance System (n = 81,557). The percentages of low income, unemployed, and obese persons who engaged in leisure-time physical activity (range = 51.1% to 57.7%) were substantially lower than the percentage among the total adult population (70.3%). In contrast, the prevalence of walking for exercise among these sedentary groups (range = 32.5% to 35.9%) was similar to that among the total population (35.6%). Walking appears to be an acceptable, accessible exercise activity, especially among population subgroups with a low prevalence of leisure-time physical activity.

Behavioral Risk Factor Surveillance, 1991: monitoring progress toward the nation's year 2000 health objectives.

PROBLEM/CONDITION: Risk reduction is a major focus of the national health objectives for the year 2000. Progress toward several of these objectives can be evaluated by using data from the Behavioral Risk Factor Surveillance System (BRFSS). Year 2000 objective areas measurable by BRFSS data include those for overweight, lack of physical activity, smoking, safety belt use, and medical screening for breast and cervical cancer and elevated blood cholesterol. BRFSS data have been used to guide health promotion/disease prevention programs. REPORTING PERIOD: 1991. DESCRIPTION OF SYSTEM: BRFSS is a state-based random-digit-dialing telephone survey of noninstitutionalized adults (> or = 18 years of age); 47 states and the District of Columbia participated in BRFSS in 1991. RESULTS: Some year 2000 objectives appear to be readily attainable for many states, whereas others do not. For example, among participating states, a median 57.8% (range = 45.6%-82.8%) of women ages > or = 50 years reported having had both a clinical breast examination and a mammogram in the previous 2 years (year 2000 objective: > or = 60%). In contrast, a median 37.3% (range = 22.1%-52.5%) of persons with annual family income < $20,000 reported that they did not engage in leisure-time physical activity--more than twice the year 2000 objective (> or = 17%). INTERPRETATION: BRFSS data demonstrate substantial state-to-state variation in progress toward year 2000 objectives and highlight areas (e.g., lack of leisure-time physical activity) in which substantial progress remains to be made in most states. Action taken: The BRFSS will continue to report data that relate to year 2000 health objectives. BRFSS data will enable states to monitor progress toward these objectives and develop health policies aimed at achieving them.

Behavioral Risk Factor Surveillance System: summary of data for 1991.

PROBLEM/CONDITION: High-risk behaviors, such as smoking cigarettes and driving under the influence of alcohol, contribute heavily to morbidity and mortality from noninfectious disease and injury. Substantial variation exists among states in the prevalences of these behaviors. REPORTING PERIOD: 1991. DESCRIPTION OF SYSTEM: The Behavioral Risk Factor Surveillance System (BRFSS) is a state-based random-digit-dialing telephone survey of noninstitutionalized adults (> or = 18 years of age). In 1991, 47 states and the District of Columbia participated in BRFSS. The system focuses on behaviors that are related to one or more of the 10 leading causes of death. In 1991, BRFSS also began collecting data on self-reported lack of health insurance. RESULTS: As in previous years, BRFSS data for 1991 indicate substantial state-to-state variation in the prevalence of risk factors such as chronic or binge alcohol consumption, sedentary lifestyle, and overweight. In addition to measures reported in previous years, the current report includes state prevalences of high blood cholesterol awareness (range = 13.5%-21.5%; median = 16.9%) and lack of health insurance (range = 7.2%-25.7%; median = 14.5%). INTERPRETATION: Because prevalence estimates vary considerably from state to state, state estimates may be preferable to national ones for use in planning programs. ACTIONS TAKEN: The BRFSS will continue to provide state-specific data about health behaviors to allow states to monitor trends that affect the burden of chronic diseases in the United States.

Stroke mortality and its socioeconomic, racial, and behavioral correlates in Florida.

Stroke mortality is associated both with being black and with having low socioeconomic status. However, it is uncertain to what extent that increased risk is related to rates of behavior-related risk factors, such as hypertension, cigarette smoking, obesity, or alcohol consumption. The investigators performed an ecologic analysis to estimate the contributions of behavioral risks, socioeconomic status, and black race to regional variations in stroke mortality rates among persons 55-84 years of age in Florida. They used data from the 1980 census and from the Behavioral Risk Factor Surveillance System (BRFSS) for 1986 through 1988. Weighted multiple linear regression models indicated that regions in Florida with high stroke mortality rates were characterized by high prevalences of poverty, obesity, and hypertension. Although limited by its ecologic design, this study suggests that socioeconomic status and prevalence of behavioral risks contribute independently to interregional disparities in stroke mortality rates in Florida. BRFSS data, now available for more than 45 States, can be used to help clarify the relative contributions of behavioral and other risks to population-based mortality rates.

North Florida is part of the stroke belt.

Florida is the only State in the southeastern United States that is not part of the "stroke belt." The National Heart, Lung, and Blood Institute (NHLBI), National Institutes of Health, defines the stroke belt as those States with an age-adjusted stroke mortality rate for the years 1979 to 1981 that is more than 10 percent above the national annual rate of 40.3 per 100,000 population. By reproducing at the county level the methods that were used at the State level to describe the stroke belt, the authors identified a group of 34 contiguous northern Florida counties (population exceeds 2 million) with an age-adjusted stroke mortality rate of 47.2 per 100,000--higher than 3 of the 11 stroke belt States. They concluded that north Florida is part of the stroke belt and should be included as a priority region for stroke prevention programs. County-level analyses that are methodologically comparable with those used by NHLBI to describe the stroke belt may be a useful technique for identifying high stroke-rate regions within States which might be missed when needs assessment is based only on State-level data.

Behavioral risk factor surveillance, 1986-1990.

Since 1984, an increasing number of states (including the District of Columbia) have participated in the Behavioral Risk Factor Surveillance System (BRFSS). This report provides state-specific estimates of the prevalence of selected health-risk behaviors for the years 1986 through 1990. Apparent trends and progress toward several of the year 2000 national health objectives are discussed, both for the entire adult population (persons ages greater than or equal to 18 years) and selected high-risk demographic subgroups. Now that BRFSS includes 45 states and covers over 90% of the nation's adult population, it can be used both as a measure of state-specific risk factor prevalence and an indicator of national trends.

Symptoms and disease associations in idiopathic intracranial hypertension (pseudotumor cerebri): a case-control study.

To identify the symptoms and coexisting medical conditions associated with idiopathic intracranial hypertension (IIH), we administered an 83-item questionnaire at the time of diagnosis to 50 IIH patients and 100 aged-matched controls. Ninety percent of the IIH patients were women; the mean age was 33. Obesity and recent weight gain were much more common among patients than controls. Symptoms most commonly reported by IIH patients were headache (94%), transient visual obscurations (TVO) (68%), and intracranial noises (ICN) (58%). Daily occurrence of these symptoms was much more common among patients than controls. Controls also reported these and other IIH symptoms, but at lower frequencies. Several conditions previously associated with IIH were no more common in patients than controls including iron deficiency anemia, thyroid disease, pregnancy, antibiotic intake, and use of oral contraceptives. We conclude that previous studies of IIH, mostly uncontrolled and retrospective, have underestimated the frequency of symptoms in IIH patients and reported chance and spurious associations with common medical conditions and medications. The profile of a young obese woman with headaches and either TVO or ICN should alert the clinician to the diagnosis of IIH, especially when the symptoms occur daily.