Functional and cognitive criteria produce different rates of mild cognitive impairment and conversion to dementia.

OBJECTIVE: To compare rates of mild cognitive impairment (MCI) and rates of progression to dementia using different MCI diagnostic systems. METHODS: MCI was investigated at baseline in 3063 community dwelling non-demented elderly in the Ginkgo Evaluation of Memory (GEM) study who were evaluated every 6 months to identify the presence of dementia. Overall MCI frequency was determined using (1) a Clinical Dementia Rating (CDR) score of 0.5 and (2) neuropsychological (NP) criteria, defined by impairment on standard cognitive tests. RESULTS: 40.2% of participants met CDR MCI criteria and 28.2% met NP MCI criteria (amnestic MCI = 16.6%). 15.7% were classified as MCI by both criteria and 47.4% as normal by both. Discordant diagnoses were observed in 24.5% who met NP normal/CDR MCI and in 12.4% who met NP MCI/CDR normal. Factors associated with CDR MCI among NP normal included lower education, lower NP scores, more instrumental activities of daily living impairment, greater symptoms of depression and subjective health problems. Individuals meeting NP MCI/CDR normal were significantly more likely to develop dementia over the median follow-up of 6.1 years than those meeting NP normal/CDR MCI. CONCLUSIONS: Different criteria produce different MCI rates and different conversion rates to dementia. Although a higher percentage of MCI was identified by CDR than NP, a higher percentage of NP MCI progressed to dementia. These findings suggest that the CDR is sensitive to subtle changes in cognition not identified by the NP algorithm but is also sensitive to demographic and clinical factors probably leading to a greater number of false positives. These results suggest that identifying all individuals with CDR scores of 0.5 as Alzheimer's disease is not advisable.

Women's Healthy Lifestyle Project: A randomized clinical trial: results at 54 months.

BACKGROUND: The Women's Healthy Lifestyle Project Clinical Trial tested the hypothesis that reducing saturated fat and cholesterol consumption and preventing weight gain by decreased caloric and fat intake and increased physical activity would prevent the rise in LDL cholesterol and weight gain in women during perimenopause to postmenopause. METHODS AND RESULTS: There were 275 premenopausal women randomized into the assessment only group and 260 women into the intervention group. The mean age of participants at baseline was 47 years, and 92% of the women were white. The mean LDL cholesterol was 115 mg/dL at baseline, and mean body mass index was 25 kg/m(2). The follow-up through 54 months was excellent. By 54 months, 35% of the women had become postmenopausal. At the 54-month examination, there was a 3.5-mg/dL increase in LDL cholesterol in the intervention group and an 8.9-mg/dL increase in the assessment-only group (P:=0.009). Weight decreased 0.2 lb in the intervention and increased 5.2 lb in the assessment-only group (P:=0.000). Triglycerides and glucose also increased significantly more in the assessment-only group than in the intervention group. Waist circumference decreased 2.9 cm in the intervention compared with 0.5 cm in the assessment-only group (P:=0.000). CONCLUSIONS: The trial was successful in reducing the rise in LDL cholesterol during perimenopause to postmenopause but could not completely eliminate the rise in LDL cholesterol. The trial was also successful in preventing the increase in weight from premenopause to perimenopause to postmenopause. The difference in LDL cholesterol between the assessment and intervention groups was most pronounced among postmenopausal women and occurred among hormone users and nonusers.

Association between depression and mortality in older adults: the Cardiovascular Health Study.

BACKGROUND: Studies of the association between depressive symptoms and mortality in elderly populations have yielded contradictory findings. To address these discrepancies, we test this association using the most extensive array of sociodemographic and physical health control variables ever studied, to our knowledge, in a large population-based sample of elderly individuals. OBJECTIVE: To examine the relation between baseline depressive symptoms and 6-year all-cause mortality in older persons, systematically controlling for sociodemographic factors, clinical disease, subclinical disease, and health risk factors. METHODS: A total of 5201 men and women aged 65 years and older from 4 US communities participated in the study. Depressive symptoms and 4 categories of covariates were assessed at baseline. The primary outcome measure was 6-year mortality. RESULTS: Of the 5201 participants, 984 (18.9%) died within 6 years. High baseline depressive symptoms were associated with a higher mortality rate (23.9%) than low baseline depression scores (17.7%) (unadjusted relative risk [RR], 1.41; 95% confidence interval [CI], 1.22-1.63). Depression was also an independent predictor of mortality when controlling for sociodemographic factors (RR, 1.43; 95% CI, 1.23-1.66), prevalent clinical disease (RR, 1.25; 95% CI, 1.07-1.45), subclinical disease indicators (RR, 1.35; 95% CI, 1.15-1.58), or biological or behavioral risk factors (RR, 1.42; 95% CI, 1.22-1.65). When the best predictors from all 4 classes of variables were included as covariates, high depressive symptoms remained an independent predictor of mortality (RR, 1.24; 95% CI, 1.06-1.46). CONCLUSIONS: High levels of depressive symptoms are an independent risk factor for mortality in community-residing older adults. Motivational depletion may be a key underlying mechanism for the depression-mortality effect.

Factors Associated With Hospital Utilization in the Elderly: From the Cardiovascular Health Study.

OBJECTIVE: Analyze clinical, accepted biochemical, physiologic, and socioeconomic risk factors and correlate them with hospital utilization in an elderly population. DESIGN: Prospective, observational study in a defined, randomly recruited population. PARTICIPANTS: 5201 Medicare participants enrolled in the Cardiovascular Health Study (CHS). METHODS: Medicare recipients were randomly assigned to participate in an observational study. Baseline data were compared to hospital admissions and days of hospitalization over four years. DATA ANALYSIS: Data were grouped by type of risk factor and analyzed by Tobit analysis and logistic regression. RESULTS: Baseline variables associated with hospital use (p is less than 0.0001) were history of CHF, stroke, angina, hypertension, ln (timed walk), ln (blocks walked/week), age, gender, and clinic site. Factors not entering the model (p is greater than 0.05) were income, education, smoking, diabetes, weight, dietary fat, marital status, depression, and measures of mental function. CONCLUSIONS: In the elderly, existing health status is the major determinant of hospitalization and overwhelms many classic "risk factors" for morbidity.

Relationship of C-reactive protein to risk of cardiovascular disease in the elderly. Results from the Cardiovascular Health Study and the Rural Health Promotion Project.

Markers of inflammation, such as C-reactive protein (CRP), are related to risk of cardiovascular disease (CVD) events in those with angina, but little is known about individuals without prevalent clinical CVD. We performed a prospective, nested case-control study in the Cardiovascular Health Study (CHS; 5201 healthy elderly men and women). Case subjects (n = 146 men and women with incident CVD events including angina, myocardial infarction, and death) and control subjects (n = 146) were matched on the basis of sex and the presence or absence of significant subclinical CVD at baseline (average follow-up, 2.4 years). In women but not men, the mean CRP level was higher for case subjects than for control subjects (P < or = .05). In general, CRP was higher in those with subclinical disease. Most of the association of CRP with female case subjects versus control subjects was in the subgroup with subclinical disease; 3.33 versus 1.90 mg/L, P < .05, adjusted for age and time of follow-up. Case-control differences were greatest when the time between baseline and the CVD event was shortest. The strongest associations were with myocardial infarction, and there was an overall odds ratio for incident myocardial infarction for men and women with subclinical disease (upper quartile versus lower three quartiles) of 2.67 (confidence interval [CI] = 1.04 to 6.81), with the relationship being stronger in women (4.50 [CI = 0.97 to 20.8]) than in men (1.75 [CI = 0.51 to 5.98]). We performed a similar study in the Rural Health Promotion Project, in which mean values of CRP were higher for female case subjects than for female control subjects, but no differences were apparent for men. Comparing the upper quintile with the lower four, the odds ratio for CVD case subjects was 2.7 (CI = 1.10 to 6.60). In conclusion, CRP was associated with incident events in the elderly, especially in those with subclinical disease at baseline.

Coronary heart disease mortality and adjuvant tamoxifen therapy.

BACKGROUND AND PURPOSE: Data from randomized clinical trials in Scotland and Sweden testing the efficacy of tamoxifen therapy in patients with breast cancer have suggested that the drug may also reduce the risk of coronary heart disease. In view of these findings, we examined mortality from coronary heart disease among patients with early stage breast cancer who were enrolled in the National Surgical Adjuvant Breast and Bowel Project B-14 trial of tamoxifen therapy. METHODS: Deaths occurring among women who were randomly assigned to 5 years of either tamoxifen or placebo in the first phase of the B-14 trial were reviewed to determine the cause. Three categories of heart disease-related death were defined: 1) death from a definite fatal myocardial infarction, 2) death from definite fatal coronary heart disease/possible myocardial infarction, and 3) death from possible fatal coronary heart disease. Comparisons of the findings by treatment group were made on the basis of average annual hazard (i.e., death) rates and the corresponding relative hazard of death. RESULTS: The average annual death rate from coronary heart disease was lower for patients who received tamoxifen than for patients who received placebo, but the difference was not statistically significant. There were eight definite heart-related deaths (i.e., definite fatal myocardial infarction or definite fatal coronary heart disease/possible myocardial infarction) among the patients who received tamoxifen, yielding an average annual rate of 0.62 per 1000 patients. There were 12 definite heart-related deaths among the patients who received placebo, yielding an average annual rate of 0.94 per 1000. The corresponding relative hazard of death from definite fatal heart disease (tamoxifen versus placebo) was 0.66 (95% confidence interval = 0.27-1.61). Eleven deaths in the tamoxifen group and 10 deaths in the placebo group were classified as possible cases of fatal coronary heart disease. When these cases and the definite cases were considered together, the average annual death rate for the patients who received tamoxifen was 1.48 per 1000, and the rate for the patients who received placebo was 1.73 per 1000. The corresponding relative hazard of death was 0.85 (95% confidence interval = 0.46-1.58). CONCLUSIONS: The findings from the B-14 trial are consistent with the findings from the Scottish and the Swedish trials, suggesting that tamoxifen treatment reduces coronary heart disease among patients with breast cancer. Continued follow-up of the patients in these trials and in ongoing prevention trials is needed to accumulate enough data so that reliable conclusions can be drawn about the benefits of tamoxifen in preventing heart disease.

Evaluation of a health promotion demonstration program for the rural elderly.

OBJECTIVE: This article evaluates a demonstration program that extended coverage for disease prevention/health promotion services to Medicare beneficiaries. STUDY SETTING/DATA SOURCES: Community-dwelling Medicare beneficiaries who lived in five rural counties in northwest Pennsylvania were recruited between May and December 1989. The demonstration lasted 18 months and beneficiaries were followed for an additional 18 months. Data for the evaluation came from an initial health risk assessment, Medicare administrative records, follow-up surveys, and redeemed vouchers for the waivered services. The waivered services included health screenings, influenza immunization, nutritional counseling, smoking and alcohol cessation, and depression/dementia evaluations. STUDY DESIGN: Medicare beneficiaries were randomized to one of two experimental groups and a control group. One experimental group received the newly waived services from hospitals that received a capitated fee; the other received services from providers who were paid fee-for-service. Eligibility for most waivered services was based on risk. Chi-square tests of association were used to determine if use of health promotion services and use of medical care services varied across groups. Logistic regressions were used to assess the factors associated with participation. Product-limit survival analysis was used to assess whether mortality rates varied across groups. PRINCIPAL FINDINGS: Participation rates in the new programs varied by program and by experimental group, and ranged from 16.8 percent for smoking cessation programs to 58 percent for influenza immunization. The demonstration led to an increase in influenza immunization rates relative to the control group. There were no differences in the use of medical care services or health outcomes between the experimental and control groups. CONCLUSIONS: Older rural Americans will modestly increase their use of disease prevention/ health promotion services if they are covered by Medicare. Use will be higher among those with more education. Further research is needed to assess long-term benefits of such programs.

Mammography and pap smear use by older rural women.

OBJECTIVE: To compare the characteristics of older women who did and did not have screening mammograms and Pap smears during the first two years both services were a Medicare Part B benefit. METHODS: A prospective study was conducted in five rural Pennsylvania counties of 2205 female community-dwelling Medicare Part B beneficiaries who volunteered to participate in a Medicare prevention demonstration project. The baseline health risk appraisal included information on demographics, insurance status, disease history, symptomatology, and functional and cognitive status. These variables were tested for their association with the use of mammography and Pap smear using Medicare utilization claims data from 1991 to 1992. RESULTS: Of 2175 women still alive after three years, 44.6% had had a mammogram and 14.6% had had a Pap smear in either 1991 or 1992. Multivariate logistic regression revealed that women were more likely to have a mammogram if they were younger, were more educated, had supplemental insurance, did not need assistance with activities of daily living, and did not have diabetes or arthritis. Younger, college educated, and non-widowed women were more likely to have Pap smears than women in other categories. CONCLUSIONS: With cost less of a barrier, more aggressive efforts to persuade older women to have mammograms and Pap smears must be developed.

Coronary heart disease mortality and sudden death among the 35-44-year age group in Allegheny County, Pennsylvania.

Deaths among 35- to 44-year-old black and white men and women residing in Allegheny County, Pennsylvania, were investigated. All coroner-certified nontraumatic deaths and practitioner-certified deaths coded as heart, cerbrovascular, or arterial disease, diabetes mellitus, and sudden or ill-defined causes were studied. Using autopsy, coroner, hospital, physician, and/or informant information about medical history, characteristics, and circumstances of death, physicians validated the deaths as due to coronary heart disease (CHD) or another cause. In 1984 to 1989, 616 deaths were investigated, 384 of which were sudden (within 24 hours of onset). Overall CHD mortality was 35.4/100,000/y for white males, 8.4/100,000/y for white females, 61.3/100,000/y for black males, and 19.5/100,000/y for black females. Although rates varied widely, characteristics, circumstances, and disease history were similar across race-sex groups. CHD mortality was 73% higher in black than white males. Approximately 80% of CHD deaths were sudden.

Nocturnal enuresis in community-dwelling older adults.

OBJECTIVE: To investigate the prevalence and characteristics of nocturnal enuresis in community-dwelling older adults and to identify potential predisposing variables. DESIGN: Interview survey. SETTING: Five rural counties in northwestern Pennsylvania. PARTICIPANTS: Subjects were 3884 community-dwelling older adults aged 65 to 79 years who volunteered for a health promotion demonstration. MEASUREMENTS: The dependent variable was self-reported accidental loss of urine during sleep. Independent variables included demographic variables, self-reported disease history and symptomatology, and standardized screening instruments for depression (CES-D), dementia (MMSE), and functional status (ADLs). MAIN RESULTS: Prevalence of nocturnal enuresis was 2.1%, and was significantly higher among women (2.9%) compared with men (1.0%; P < .0001). Compared with subjects with daytime incontinence only, those with nocturnal enuresis reported greater severity and impact of incontinence on several parameters. Enuretics were more likely to have received treatment; treatment outcome, however, was significantly less successful. In univariate analyses, enuresis was significantly associated with symptoms of congestive heart failure (CHF), impairment in activities of daily living, depression, and use of sleep medications at least once per week. In stepwise logistic regression modeling, two symptoms of congestive heart failure and regular use of sleep medication entered the model. CONCLUSIONS: Nocturnal enuresis appears to be uncommon among older adults, but it may be associated with poorer therapeutic outcomes compared with the more common forms of daytime incontinence. The findings are consistent with the hypothesis that daytime fluid accumulation followed at night by mobilization of excess fluid is a contributor to enuresis in older adults.

Surveillance and ascertainment of cardiovascular events. The Cardiovascular Health Study.

While previous prospective multicenter studies have conducted cardiovascular disease surveillance, few have detailed the techniques relating to the ascertainment of and data collection for events. The Cardiovascular Health Study (CHS) is a population-based study of coronary heart disease and stroke in older adults. This article summarizes the CHS events protocol and describes the methods of surveillance and ascertainment of hospitalized and nonhospitalized events, the use of medical records and other support documents, organizational issues at the field center level, and the classification of events through an adjudication process. We present data on incidence and mortality, the classification of adjudicated events, and the agreement between classification by the Events Subcommittee and the medical records diagnostic codes. The CHS techniques are a successful model for complete ascertainment, investigation, and documentation of events in an older cohort.

Coronary heart disease mortality and sudden death: trends and patterns in 35- to 44-year-old white males, 1970-1990.

Trends in coronary heart disease mortality and sudden death were studied in 35- to 44-year-old white male residents of Allegheny County, Pennsylvania. Deaths coded as any cardiac or vascular disease, diabetes, unexplained sudden death, and other rubrics were eligible for investigation, and the cause of death was validated by physicians examining multiple data sources about the deaths. During 1970-1990, 1,424 white male deaths were investigated, with 903 validated as coronary heart disease. In that time span, white male coronary heart disease mortality fell from 93.4 to 36.7 per 100,000 population per year, a 60% decline. Little proportionate change was seen in characteristics of the deaths, which were predominantly sudden and out-of-hospital. Diabetes mellitus history increased proportionately over time, largely because diabetics' mortality rates, unlike those of all other subgroups, did not fall. These observations support the contention that the decline in coronary heart disease mortality relates to risk factor modification more than to improvements in the treatment of coronary heart disease. Differences in death certification practices must be considered when interpreting and comparing vital statistics data.

Participation in health promotion programs by the rural elderly.

The Health Care Financing Administration (HCFA) funded a series of demonstration programs to learn about the implications of extending coverage for disease prevention/health promotion services to Medicare beneficiaries. This article examines the use of such services by a rural population under this demonstration program. Individuals enrolled in the demonstration were eligible for specific risk reduction interventions. They were enrolled in one of two groups: (1) a hospital-based group in which hospitals were paid a capitated fee for providing all services and (2) a physician-based group in which physicians were paid fee-for-service for providing each service. Chi-square tests of association as well as logistic regression models were used to assess whether eligibility for services, and use of services by those eligible, varied by group and by sociodemographic characteristics. Forty-one percent were eligible for a nutrition program, 11% for smoking cessation, 2% for alcohol counseling, and 7% for dementia/depression evaluations. Participation in the programs varied across the programs and within programs by gender, education, and group assignment. Older rural Americans will use some disease prevention/health promotion services if they are covered by Medicare. Use will be higher among those with more education. Rural beneficiaries are more likely to use preventive services if encouraged to do so by their doctors rather than by hospital-based programs.

Selection bias and nonresponse to health promotion in older adults.

Some epidemiologic studies have compared the characteristics of individuals who participate, refuse, and are unreachable in population studies, but results have been inconsistent. The Rural Health Promotion Project attempted to recruit all Medicare Part B noninstitutionalized individuals age 65-79 years in a rural community for a trial of preventive health services. Of 962 potential subjects, 360 (37.4%) participated, 253 (26.3%) refused, 176 (18.3%) were ineligible, and 152 (15.8%) were never reached by phone or mail. Approximately 3 years later, we reinterviewed the participants, refusals, and as many of the unreachables as possible. The 3-year mortality was similar for both refusals and participants (approximately 9%) but was much higher for ineligibles (29.0%) and unreachables (23.7%). Participants were more likely to have disease history, to have behavioral risk factors for disease, and to use health screening services. Refusals were the healthiest and possibly chose not to participate because they did not have risk factors targeted by the program. The unreachables had the highest prevalence of disability and health care inpatient reimbursement and may have been ineligible for the demonstration had they volunteered. We conclude that failure to reach potential participants for health promotion services may be a warning of "high risk."

Impact of Medicare reimbursement on influenza vaccination rates in the elderly.

BACKGROUND: Influenza is responsible for significant morbidity, mortality, and medical costs, but immunization rates in the elderly remain low. METHODS: As part of a demonstration project in rural Pennsylvania, 1,989 community-dwelling Medicare beneficiaries, 65-79 years old, completed a health risk appraisal including questions about flu shots. Participants were randomized to two experimental (hospital or physician) or control groups. Experimental groups were offered free flu shots for the 1990-91 flu season. Follow-up interviews to determine vaccination rates were completed about 1 year later. RESULTS: Baseline immunization rates were almost identical for the hospital-based (41.2%), physician-based (41.3%), and control group not offered free immunizations (40.6%). Follow-up rates for the groups offered immunizations rose significantly to 63.6 and 69.1% for hospital and physician groups, respectively, while the control group also increased significantly to 54.1%. Individuals who were more educated, were older, and had greater chronic diseases history (myocardial infarction, hypertension, and pulmonary disease) were more likely to be immunized. Male and married elderly were more likely to be immunized as a result of the demonstration. CONCLUSIONS: Increasing community education and providing immunizations free through Medicare will increase immunization rates among elderly. The elderly are more likely to receive flu shots provided through physicians' offices than through hospital-based clinics.

Treatment seeking for urinary incontinence in older adults.

OBJECTIVE: To examine treatment seeking for urinary incontinence among older adults and to identify characteristics associated with treatment-seeking behavior. DESIGN: Survey. SETTING: Five rural counties in northwestern Pennsylvania. PARTICIPANTS: 1104 community-dwelling ambulatory older adults aged 65 to 79 years with self-reported urinary incontinence. Participants were a subgroup of a large sample (n = 3884) who volunteered for a study of health promotion services. Those who reported urinary incontinence within the past year, during an in-person health risk appraisal, were included in this analysis. MEASUREMENTS: Reporting incontinence to the participant's physician was the main dependent measure. MAIN RESULTS: 37.6% of the participants had told their physician about loss of urine. Reporting incontinence to a physician was strongly associated with severity of incontinence as indicated by eight measures (P < 0.001). Treatment seeking was also related to type of incontinence (P < 0.001), physical disability (P < 0.01), and the pattern of health care utilization (P < 0.01). In multiple logistic regression analyses, younger age, physical disability, and frequency of physical and rectal examinations had significant predictive value independent of severity. Not associated with treatment seeking were gender, marital status, income, employment status, educational level, and distance from health care provider. CONCLUSIONS: The majority of older adults with urinary incontinence do not report the condition to their doctor. Severity of incontinence, physical disability, and a pattern of regular health care utilization appear to be the strongest predictors of treatment-seeking behavior.

Use and outcomes of a cholesterol-lowering intervention for rural elderly subjects.

Few studies have evaluated the efficacy of cholesterol-lowering interventions in a community setting and have included a control or comparison group. As part of a preventive health demonstration project in rural Pennsylvania, Medicare beneficiaries underwent cholesterol screening to identify high-risk individuals with serum cholesterol levels > or = 240 mg/dL. These high-risk individuals were randomized to a cholesterol-lowering intervention through either local hospitals or physicians' offices or to a control group. Baseline and follow-up serum cholesterol levels collected two to three years later were compared according to service location (hospital versus physician's office), intervention attendance, degree of participation, baseline heart disease history, and cholesterol-lowering medication use at follow-up. Serum cholesterol levels decreased between 5.7% and 6.6% in the hospital-based and physician-based groups, as well as in a control group not offered the intervention. Participation rates did not differ between treatment groups, nor did participation affect serum cholesterol levels. Attendance level and heart disease history were not associated with a greater decrease in serum cholesterol levels. Individuals reporting cholesterol-lowering drug use at follow-up had significantly higher baseline serum cholesterol levels and a significantly greater decrease in total serum cholesterol (P < .0001) than those not on medication. Both nonpharmacological (diet) and pharmacological (drug) interventions will reduce serum cholesterol levels and heart attack risk. The study results suggest that, at least for older individuals, the impact of nonpharmacological interventions on the community is minimal. We conclude that only very aggressive treatment will significantly loser serum cholesterol levels in older individuals at risk for heart attack.

Morbidity and mortality in rural community-dwelling elderly with low total serum cholesterol.

BACKGROUND: Low serum cholesterol has been associated with morbidity and mortality in the elderly. This study compared the health, functional status, and two-year mortality rates of community-dwelling rural elderly with serum cholesterol < 150 mg/dl to age- and sex-matched controls with serum cholesterol 200-240 mg/dl. METHODS: Self-reported disease history, disability, health habits, and cognitive function data were collected at a health risk appraisal interview. A single blood sample was also collected and analyzed for total serum cholesterol at a central lab. RESULTS: Of the 3,874 participants, 109 (2.8%) had total cholesterol levels < 150 mg/dl. Seventy-five percent of the low cholesterol group were male compared to 44% in the main study population. The low cholesterol group had significantly greater smoking history, current cigarettes smoked, diabetes history, angina and COPD symptoms, and assistance needed for heavy and light work. Men in the low cholesterol group had significantly lower blood pressure. After two years, 14 (12.8%) of the low cholesterol group had died vs 16 (7.3%) in the control group. There was no relationship to specific causes of death and cholesterol level. CONCLUSION: A very low cholesterol level in older individuals should be evaluated carefully to determine whether it is due to genetic or life-style factors such as diet or, more likely, is a marker of disease.

Comparison of recruitment strategies and associated disease prevalence for health promotion in rural elderly.

BACKGROUND: Although interest in health promotion for the elderly is increasing, the issues of recruitment into such programs and self-selection have not been well explored. While clinical studies require high participation levels and expensive recruitment, community efforts are satisfied with recruiting small numbers of volunteers from poorly defined populations. These small samples may not be representative of the populations at risk. METHODS: As part of the Rural Health Promotion Project, a Medicare demonstration, community-based recruitment methods were evaluated and participant characteristics were compared. A total of 3,884 individuals ages 65-79 were recruited in northwestern Pennsylvania, using four sequential recruitment strategies, varying in aggressiveness. The methods were: (A) mail only, (B) mail with phone recruitment follow-up, (C) mail with phone recruitment and scheduling, and (D) mail with aggressive phone recruitment and scheduling. RESULTS: Recruitment yields were Method A, 13.5%; B, 21.1%; and C, 31.6%. The most aggressive Method (D) yielded 37.0% participation. More aggressive methods (C and D) recruited more educated individuals. No other demographic or health status differences were noted. CONCLUSION: These data show that large numbers of the elderly can be recruited into a health promotion program using aggressive methods and professional interviewers.