The Behavioral Risk Factor Surveillance System (BRFSS) is a state-based, random-digit–dialed telephone survey of the US noninstitutionalized, civilian population. We analyzed the self-reported data from 427,269 adults aged 18 years or older in 2007 from the 50 states, the District of Columbia, Guam, Puerto Rico, and the US Virgin Islands. The median response rate among geographic units, based on the Council of American Survey and Research Organizations guidelines, was 47.8% (range, 26.9% in New Jersey to 79.9% in Guam). This rate reflects both telephone sampling efficiency and the degree of participation among eligible respondents contacted. The median cooperation rate for the 2007 BRFSS survey was 73.3% (range, 49.6% in New Jersey to 95.0% in Guam) and reflects the proportion of eligible people contacted who completed an interview. Additional details on the survey can be found at www.cdc.gov/brfss
The HRQOL module has been used in BRFSS since 1993 and allows the assessment of general health, recent physical or mental health or both, and activity limitations (2
). Participants provide subjective ratings of general health ("Would you say that in general your health is excellent, very good, good, fair, or poor?"), recent physical health ("Now thinking about your physical health, which includes physical illness and injury, for how many days during the past 30 days was your physical health not good?"), recent mental health ("Now thinking about your mental health, which includes stress, depression, and problems with emotions, for how many days . . . mental health not good?"), and activity limitations ("For how many days did poor physical or mental health keep you from doing your usual activities, such as self-care, work, or recreation?"). The questions have been validated with the medical outcomes short study form (12
We analyzed 5 unfavorable HRQOL measurements among people with self-reported CHD, which we refer to as "unhealthy days measurements" when discussing them as a group. General health status was dichotomized as good/excellent (respondents reporting excellent, very good, or good health) or fair/poor. The number of days in the past 30 days in which a person reported constraints related to physical, mental, total (physical and mental), and activity-limited days was calculated as 14 or more days compared with less than 14 days. These unhealthy days measurements are traditionally used with BRFSS data, have been associated with chronic disease, and indicate a substantial level of impairment (13
). In this study, we defined people with CHD as those who reported ever being told by a doctor or other health professional that they had had a "heart attack, also called a myocardial infarction," or "angina or coronary heart disease" during their lifetime.
Differences in the prevalence of each unhealthy days measure were assessed by age group (18-34, 35-49, 50-64, and ≥65 y), sex, race/ethnicity, and other socioeconomic indicators (education, health insurance coverage, annual household income, and household size). Self-identified race/ethnicity was either non-Hispanic white, non-Hispanic black, Hispanic (any race), Asian, Native American, or other. Native American was used for respondents who self-identified as being of American Indian or Alaska Native race. The "other" race category included respondents who self-identified as being of Native Hawaiian, Pacific Islander, or other, and those who indicated more than 1 race. Education levels were based on highest grade or year of school completed and categorized as not completing high school (<12 y), completing high school or its equivalent (12 y), some college course work, or college graduate or more. Respondents were considered to have health insurance coverage if they reported any type of health insurance. Annual household income was categorized as less than $20,000, $20,000 to $34,999, $35,000 to $49,999, or $50,000 or more, and as unknown/refused. Household size was categorized as living alone, with 1 other person, or with 2 or more people.
We excluded from our analysis observations with missing data on any of the unhealthy days measurements (4.2%) or CHD status (<0.1%), and we excluded pregnant women (0.8%), resulting in a sample size of 405,641; we focused the analysis on the 35,378 participants with self-reported CHD. Prevalence and 95% confidence intervals (CIs) of these unhealthy days measurements were determined for selected socioeconomic characteristics. Prevalence estimates were age-standardized to the 2000 US standard population except for those associated with specific age groups. Multivariate logistic regression models were developed for each of the 5 unhealthy days measurements; age group, sex, race/ethnicity, education, income, household size, and health insurance coverage were covariates. All covariates were entered into each of the 5 models to allow for comparison between the models. Data were weighted to reflect each state's noninstitutionalized, adult population.
Significant differences for estimates by characteristics for the 5 unhealthy days measurements were assessed by pairwise comparison tests with a reference group we selected for comparison. For the multivariate logistic regression model, reported P values for the t test of the beta coefficients are reported. A P value of <.05 was considered significant for the estimates by characteristics and in the multivariate logistic regression models. SAS version 9.2 (SAS Institute, Inc, Cary, North Carolina) and SUDAAN version 10.0 (RTI International, Research Triangle Park, North Carolina) statistical software were used to account for the complex sampling design so that accurate variance estimates could be calculated.