Summary measures of population health: methods for calculating healthy life expectancy.

This report is one of several appearing as Healthy People Statistical Notes that evaluate methodological issues pertaining to summary measures. Summary measures of population health are statistics that combine mortality and morbidity to represent overall population health in a single number--in this report, health expectancy measures. This report presents a comprehensive discussion of the methods for calculation and methodologic issues related to the interpretation of healthy life expectancy. These measures combine both mortality and morbidity using an abridged life-table procedure. Data from the National Center for Health Statistics and other sources will be used to illustrate the calculation of the statistics and the associated statistical tests.

Summary measures of population health: addressing the first goal of healthy people 2010, improving health expectancy.

This report is one of several Healthy People Statistical Notes that evaluate methodological issues pertaining to summary measures - statistics that combine mortality and morbidity data to represent overall population health in a single number. This report evaluates the consequences of changes in the components of health expectancy measures (i.e., mortality or morbidity) on the overall measure. Any activity limitation is used as a morbidity measure. Simulations are used to evaluate the impacts of reducing 1995 age-specific mortality or activity limitation rates by 5, 10, 25, and 50 percent at all ages. Then it is limited to ages under 25 years, 25 -64 years, and over 64 years. The impact of completely eliminating mortality or activity limitation for the younger age groups is also examined. In general, reducing morbidity rates results in greater changes than the same percent reduction in death rates. The same proportional reduction in age-specific rates for either mortality or morbidity has a greater impact if it occurs at older ages. Reducing mortality results in a greater change in life expectancy than in health expectancy and a decline in the proportion of life lived in healthy states. Reducing morbidity increases both health expectancy and the proportion of life lived in healthy states. Simultaneous reductions in mortality and morbidity have additive effects on health expectancy.

Methodological issues for vital rates and population estimates: 1997 OMB standards for data on race and ethnicity.

A workshop in July 2000 explored research needed to address methodological challenges for population estimates and vital rates arising from the revision of the standards for Federal data on race and ethnicity; the use of different standards for Census 2000 and vital records during the implementation of the revised standards; and underlying differences in the collection of race and ethnicity data in censuses and surveys and in vital records. Matching studies were proposed to compare race and ethnicity reports in vital records and in Census 2000 or survey responses. Work on vital records might include exploring enhanced collection of race and ethnicity data and documenting State vital statistics reporting mandates and practices. Key work on sources of error includes identifying, quantifying, and reporting on bias and random errors related to race and ethnicity in population estimates and vital rates. Also needed are comparisons of Census Bureau and State population estimates and more frequent tests of the accuracy of population estimates and projections. Studies on racial and ethnic identity were proposed to examine changes in reporting over time and to explore origins of racial and ethnic identities. More information on these issues should be developed for technical and nontechnical audiences.

Body fat distribution, obesity, overweight and stroke incidence in women and men--the NHANES I Epidemiologic Follow-up Study.

OBJECTIVE: To test the hypothesis that an elevated ratio of subscapular to triceps skinfold thickness (SFR), a measure of truncal obesity, is associated with increased incidence of stroke independent of overweight. DESIGN: Data from the NHANES I Epidemiologic Follow-up Study were analyzed. SUBJECTS: A cohort of 3652 women and 3284 men with complete data who had no history of stroke at baseline in 1971-1975. MEASUREMENTS: Incidence of stroke diagnosed at hospital discharge or death during the follow-up period through 1992; triceps and subscapular skinfold thickness (SSF) and body mass index (BMI) at baseline. RESULTS: In a complex relationship, higher SFR was associated with a mildly but significantly increased incidence of stroke only in white male former smokers. In white men, SSF showed a U-shaped association with stroke risk. In white men, stroke risk was elevated in the top quartile of BMI only in never smokers. In black women, stroke risk was significantly elevated in the bottom compared to the top quartile of BMI. No significant associations were seen in white women or black men. CONCLUSIONS: In white men, SSF showed a U-shaped association with stroke risk, which was elevated in the top quartile of BMI only in never smokers. Surprisingly, stroke risk was elevated in black women with the lowest BMI. More studies of these associations are needed, especially in black women.

Fish consumption and hypertension incidence in African Americans and whites: the NHANES I Epidemiologic Follow-up Study.

We sought to test the hypothesis that increased consumption of fish is associated with decreased incidence of essential hypertension. Data on fish consumption and incidence of hypertension from a national cohort of 5,394 blacks and whites normotensive at baseline and followed 10 years in the NHANES I Epidemiologic Follow-up Study (NHEFS) were analyzed. Our results showed that whites aged 25-74 years had no significant association of fish consumption with incidence of hypertension. In black women, after adjusting for multiple risk factors, those who increased their fish intake from <1 time/week to > or = 1 time/week had RR = 0.42, 95% CI 0.22-0.81, p = 0.009. However, those with high intake both times had adjusted RR = 0.75, 95% CI 0.45-1.26, p = 0.28. No consistent significant associations of fish consumption with hypertension incidence were found, perhaps because fish consumption in this population was low. Further studies are needed in blacks.

Race/ethnicity and the 2000 census: implications for public health.

OBJECTIVES: This article addresses the potential impact of the revised standards for race and ethnicity on data from the 2000 census and public health data sources, policies, and programs. METHODS: The authors examine the relationship between race/ethnicity and health in selected measures, identify the factors that influence race/ethnicity identification, consider past experience in race/ethnicity reporting, and explore the challenges in understanding and managing the effects of new racial/ethnic categories in various data sets. RESULTS: The multiple-race group seems to compose only a small percentage of the US population and may have little impact on data for single-race groups. Actual effects will vary according to a number of factors, including the size, composition, and geographic distribution of the group. CONCLUSIONS: More research is needed to support a thorough understanding of the reporting of multirace data and the development of techniques for analyzing these data. Given the importance of understanding the relationship between race/ethnicity and health, the ability to produce useful, comparable, and meaningful data is essential.

Serum folate and cardiovascular disease mortality among US men and women.

BACKGROUND: Folate has been linked to cardiovascular disease (CVD) through its role in homocysteine metabolism. OBJECTIVE: To assess the relationship between serum folate and CVD mortality. DESIGN: In this prospective study, serum folate concentrations were measured on a subset of adults during the Second National Health and Nutrition Examination Survey (1976-1980) and vital status ascertained after 12 to 16 years. SETTING AND PATIENTS: A national probability sample consisting of 689 adults who were 30 to 75 years of age and did not have a history of CVD at baseline. MAIN OUTCOME MEASURE: Vital status was determined by searching national databases that contained information about US decedents. RESULTS: The associations between serum folate and CVD and all-cause mortality differed by diabetes status (P =.04 and P =.03, respectively). Participants without diabetes in the lowest compared with the highest serum folate tertile had more than twice the risk of CVD mortality after adjustment for age and sex (relative risk [RR], 2.64; 95% confidence interval [CI], 1.15-6.09). This increased risk for participants in the lowest tertile was attenuated after adjustment for CVD risk factors (RR, 2.28; 95% CI, 0.96-5.40). Serum folate tertiles were not significantly associated with total mortality, although the age- and sex-adjusted risk was increased for participants in the lowest compared with highest tertile (RR, 1.74; 95% CI, 0.96-3.15). Risk estimates for participants with diabetes were unstable because of the small sample size (n = 52). CONCLUSION: These data suggest that low serum folate concentrations are associated with an increased risk of CVD mortality among adults who do not have diabetes. Arch Intern Med. 2000;160:3258-3262.

Diabetes mellitus, coronary heart disease incidence, and death from all causes in African American and European American women: The NHANES I epidemiologic follow-up study.

Few data are available on risk for coronary heart disease in African American women with diabetes mellitus, a well-established coronary risk factor in European American women. This study tests the hypothesis that medical history of diabetes predicts coronary heart disease incidence in African American women in a national cohort. Participants in the NHANES I Epidemiologic Follow-up Study in this analysis were 1035 African American and 5732 European American women aged 25-74 years without a history of coronary heart disease. Average follow-up for survivors was 19 years (maximum 22 years). Risk of incident coronary heart disease by baseline diabetes status was estimated. Proportional hazards analyses for African American women aged 25-74 revealed significant associations of coronary heart disease risk with diabetes after adjusting for age (RR = 2.40; 95% CI, 1.58-3.64, P < 0.01). After adjusting for age, smoking, and low education, there was an elevated risk in diabetics age 25-74 (RR = 2. 34; 95% CI, 1.54-3.56, P < 0.01); this association did not differ significantly from that for European American women. Excess coronary incidence in African American compared to European American women aged 25-64 was statistically explained by controlling for diabetes history, age, education, and smoking but only partly explained by age and diabetes history. In African American women aged 25-74, diabetes was also associated with increased coronary heart disease, cardiovascular, and all-cause mortality. The population attributable risk of coronary heart disease incidence associated with a medical history of diabetes was 8.7% in African American women and 6.1% in European American women. Medical history of diabetes was a significant predictor of coronary heart disease incidence and mortality in African American women and explained some of the excess coronary incidence in younger African American compared to European American women.

The relation between fish consumption, death from all causes, and incidence of coronary heart disease. the NHANES I Epidemiologic Follow-up Study.

1 time/week (adjusted relative risk 0.85, 95% CI 0.68-1.06). Similar but nonsignificant trends were seen in white and black women, but not black men. In white men, risk of noncardiovascular death but not cardiovascular death was also significantly reduced in those consuming fish once or more a week. No consistent association of fish consumption and coronary heart disease incidence or mortality was seen. White men consuming fish once a week had significantly lower risk of death over a 22-year follow-up than those never consuming fish. This was mostly attributable to reductions in death from noncardiovascular causes. Similar patterns, though not significant, were seen in women. Further studies are needed to confirm these findings and to elucidate mechanisms for the effect of fish consumption on noncardiovascular mortality.

Risk factors for hip fracture in white men: the NHANES I Epidemiologic Follow-up Study.

This prospective population-based study assessed predictors of hip fracture risk in white men. Participants were members of the Epidemiologic Follow-up Study cohort of the First National Health and Nutrition Examination Survey, a nationally representative sample of noninstitutionalized civilians who were followed for a maximum of 22 years. A cohort of 2879 white men (2249 in the nutrition and weight-loss subsample, 1437 in the bone density subsample) aged 45-74 years at baseline (1971-1975) were observed through 1992. Ninety-four percent of the original cohort were successfully traced. Hospital records and death certificates were used to identify a total of 71 hip fracture cases (61 in the nutrition and weight-loss subsample, 26 in the bone-density subsample). Among the factors evaluated were age at baseline, previous fractures other than hip, body mass index, smoking status, alcohol consumption, nonrecreational physical activity, weight loss from maximum, calcium intake, number of calories, protein consumption, chronic disease prevalence, and phalangeal bone density. The risk adjusted relative risk (RR) of hip fracture was significantly associated with presence of one or more chronic conditions (RR = 1.91, 95% confidence interval [CI] = 1.19-3.06), weight loss from maximum > or = 10% (RR = 2.27, 95% CI 1.13-4.59), and 1 SD change in phalangeal bone density (RR = 1.73, 95% CI 1.11-2.68). No other variables were significantly related to hip fracture risk. Although based on a small number of cases, this is one of the first prospective studies to relate weight loss and bone density to hip fracture risk in men.

Coronary heart disease risk factors and attributable risks in African-American women and men: NHANES I epidemiologic follow-up study.

OBJECTIVES: This study assessed associations of risk factors with coronary heart disease incidence in African Americans. METHODS: The participants in the NHANES I Epidemiologic Follow-Up Study included in this analysis were 1641 Black and 9660 White persons who were aged 25 to 74 years when examined and who did not have a history of coronary heart disease. Average follow-up for survivors was 19 years. RESULTS: Significant, independent risk factors for coronary heart disease were age, systolic blood pressure, and smoking in Black women and age, systolic blood pressure, serum cholesterol, low education, and low family income in Black men. In this cohort, 19% of incident coronary heart disease in Black women and 34% in Black men might be prevented if systolic blood pressure were below 140 mm Hg. In Black men, attributable risk for low education (46%) was even higher than that for elevated blood pressure. CONCLUSIONS: Elevated systolic blood pressure and smoking were predictive of coronary heart disease incidence in African Americans. Estimates of population attributable risk were highest for elevated systolic blood pressure in women and education less than high school in men. Further studies of serum lipids, education, and coronary heart disease in Black women are needed.

Body fat distribution and hypertension incidence in women and men. The NHANES I Epidemiologic Follow-up Study.

OBJECTIVE: To test the hypothesis that an elevated ratio of subscapular to triceps skinfold thickness (SFR), one measure of truncal obesity, is associated with increased incidence of essential hypertension. DESIGN: Data from the NHANES I Epidemiologic Follow-up Study (NHEFS) were analyzed. SUBJECTS: A cohort of 4303 women and 2579 men with complete data who were normotensive at baseline in 1971-1975. MEASUREMENTS: Incidence of hypertension, blood pressure 160/95 mm Hg or greater or on blood pressure medication at follow-up in 1982-1984. RESULTS: There was a statistically significant increase in risk of hypertension over approximately 10 y follow-up in white women aged 25-74 y with SFR in the fifth compared to the first quintile independent of age and body mass index (BMI) (relative risk = 1.52, 95% confidence interval 1.13-2.06, P = 0.006). The association was somewhat diminished after controlling for baseline blood pressure, change in BMI and other risk variables. An even stronger association was seen for subscapular skinfold and hypertension incidence. In white men aged 25-74 y, a significant association of high SFR with age-, BMI-adjusted risk of hypertension was seen (RR = 1.41, 95% CI 1.01-1.96, P = 0.04). Data for black women or black men failed to reveal significant variation in hypertension risk among quintiles of SFR or subscapular skinfold except in black women with low baseline BMI. CONCLUSIONS: Data from NHEFS confirm the association of higher truncal obesity with increased incidence of hypertension in white women. Further studies are needed, especially in larger samples of black women.

Coronary heart disease incidence and survival in African-American women and men. The NHANES I Epidemiologic Follow-up Study.

BACKGROUND: Relatively few data are available on risk for or survival with coronary heart disease in African-American persons. OBJECTIVE: To determine whether incidence of coronary heart disease, rate of survival with the disease, and rate of coronary surgery differ between ethnic groups. DESIGN: Prospective cohort study. SETTING: United States. PARTICIPANTS: Persons who responded to the National Health and Nutrition Examination Survey (NHANES) I Epidemiologic Follow-up Study. Included in this analysis were 11406 white persons and African-American persons aged 25 to 74 years who had no history of coronary heart disease. Average follow-up for survivors was 19 years (maximum, 22 years). MEASUREMENTS: Incident coronary heart disease. RESULTS: Compared with that in white persons, the age-adjusted risk for coronary heart disease was higher in African-American women aged 25 to 54 years (relative risk, 1.76 [95% CI, 1.36 to 2.29]) but was lower in African-American men within each age subgroup. The age-adjusted risk was lower in African-American men for all ages combined (25 to 74 years) (relative risk, 0.78 [CI, 0.65 to 0.93] for coronary heart disease and 0.62 [CI, 0.42 to 0.92] for acute myocardial infarction). The higher rate in African-American women aged 25 to 54 years could be explained statistically by the higher risk factor levels in these women. Ethnic groups did not significantly differ in survival after the first hospitalization for coronary heart disease. However, the incidence of coronary procedures after hospitalization for coronary heart disease was markedly lower in African-American persons than in white persons (age- and sex-adjusted relative risk, 0.40 [CI, 0.16 to 0.99]). CONCLUSIONS: Total incidence of coronary heart disease is higher in African-American women aged 25 to 54 years than in white women of the same ages and is lower in African-American men aged 25 to 74 years than in white men of the same ages.

Epidemiology of hysterectomy in the United States: demographic and reproductive factors in a nationally representative sample.

We describe the epidemiology of hysterectomy, overall as well as for specific indications. Data were obtained from the Epidemiologic Follow-up to the First National Health and Nutrition Examination Survey, a nationally representative cohort followed prospectively from the mid-1970s through 1992. Black and white women 25-49 years of age, interviewed during follow-up, were included in the analyses. The probability of undergoing a hysterectomy was estimated by demographic and reproductive factors. Hysterectomy as confirmed by hospital records was our main outcome measure. We found that women who had completed 9-11 years of education were more likely to have undergone a hysterectomy than were women with either more or less education. Women who had completed 9-11 years of education were also more likely to have had a hysterectomy because of menstrual problems. Three or more miscarriages, especially if caused by uterine prolapse, increased the probability of hysterectomy. Having had no live births decreased the probability of hysterectomy for menstrual disorders and uterine prolapse, but women who had their first child before age 20 were at increased risk of hysterectomy because of endometriosis. Hysterectomy appears to be associated with low education, high parity, and a history of multiple miscarriages. The influence of these factors varies depending on the primary indication for the hysterectomy.

Use of postmenopausal hormone replacement therapy: estimates from a nationally representative cohort study.

The objective of this study was to describe trends in the use of hormone replacement therapy (HRT) in the United States by demographic, life-style, and heart disease risk factors. Data were obtained from the Epidemiologic Followup Study to the First National Health and Nutrition Examination Survey, a nationally representative cohort followed from the mid-1970s until 1992. A total of 5,602 women who had become menopausal by their last follow-up interview were included. An estimated 45% of the cohort of menopausal US women 25-74 years of age in the early 1970s used HRT for at least one month and 20% continued use for 5 or more years. Between 1987 and 1992, as the younger members of the cohort became menopausal, the proportion of this cohort who had ever used HRT and used it for 5 or more years increased by 32% and 54%, respectively. A higher probability of HRT use was found among women who were white, who were more highly educated, and who lived in the West, or who had experienced a surgical menopause. Women who were overweight or who abstained from alcohol were less likely to use HRT. These data support the hypothesis that HRT use is associated with sociodemographic factors, and that women tend to discontinue use within several years.

The great efficacy of personal and equipment assistance in reducing disability.

OBJECTIVES: Personal and equipment assistance are common strategies to reduce disability. This study sought to determine how often assistance reduces or even completely resolves health-related difficulties in everyday tasks. METHODS: Data are from the NHANES I Epidemiologic Followup Study. Adults aged 35 to 90 reported difficulty doing 12 everyday tasks on their own without assistance. Those stating that they had much difficulty or were unable were asked if they had personal assistance and/or equipment assistance, and their degree of difficulty with assistance. Use and efficacy of assistance are studied by gender, age, intrinsic (unassisted) degree of difficulty, and type of assistance. RESULTS: Most people use assistance for the 12 tasks; "personal assistance only" is the principal type used for upper-extremity and body transfer tasks; "equipment only" ranks first for lower-extremity tasks. Assistance reduces difficulty for the great majority of persons (75% to 85%) and completely resolves difficulty for about 25%. Equipment only proves to be the most efficacious strategy for reducing and resolving limitations. CONCLUSIONS: Equipment's success may be due to greater perceived gains when people accomplish the assistance by themselves.

Phalangeal bone density and hip fracture risk.

OBJECTIVE: To assess the long-term predictive usefulness of radiographic absorptiometry measurements of phalangeal bone density for hip fracture risk. METHODS: Participants were members of the First National Health and Nutrition Examination Survey Epidemiologic Follow Up Study cohort. Subjects were followed up for a maximum of 16 years. The First National Health and Nutrition Examination Survey data were obtained from a nationally representative sample of non-institutionalized civilians. A cohort of 3481 white and black subjects (1559 white women) aged 45 through 74 years at baseline (1971-1975) were observed through 1987. Ninety-eight percent of the original cohort completed the study. Hospital records and death certificates were used to identify a total of 72 hip fracture cases. Phalangeal bone density at baseline was measured using photodensitometry (PD), and later reanalyzed by radiographic absorptiometry (RA), a newer, more sophisticated technique. RESULTS: Results were evaluated to determine the relative risk for hip fracture per 1-SD decrease in bone density, after controlling for age at baseline, race, gender, weight, and previous fractures. Both RA and PD measurements showed a significant inverse relationship to hip fracture risk, with RA density measurements showing a slightly higher adjusted relative risk per 1-SD density decrease than PD measurements. For RA bone density, the relative risk for all subjects was 1.81 (95% confidence interval, 1.34-2.44) compared with 1.57 (95% confidence interval, 1.19-2.07) for PD bone density after adjusting for age at baseline, race, gender, weight, and previous fractures. Results for white women were essentially the same as those for all subjects for RA bone density and PD bone density. CONCLUSIONS: Phalangeal bone density determined from standard hand x-ray films is a significant predictor of future hip fracture risk. Availability of a valid method to assess fracture risk using conventional radiographs will expand the ability to identify individuals with osteoporosis.

Plan and operation of the NHANES I Epidemiologic Followup Study, 1992.

OBJECTIVES: The NHANES I Epidemiologic Followup Study (NHEFS) is a longitudinal study that uses as its baseline those adult persons 25-74 years of age who were examined in the first National Health and Nutrition Examination Survey (NHANES I). NHEFS was designed to investigate the association between factors measured at baseline and the development of specific health conditions. The three major objectives of NHEFS are to study morbidity and mortality associated with suspected risk factors, changes over time in participants' characteristics, and the natural history of chronic disease and functional impairments. METHODS: Tracing and data collection in the 1992 Followup were undertaken for the 11,195 subjects who were not known to be deceased in the previous surveys. No additional information was collected in the 1992 NHEFS for the 3,212 subjects who were known to be deceased before the 1992 NHEFS data collection period. RESULTS: By the end of the 1992 NHEFS survey period, 90.0 percent of the 11,195 subjects in the 1992 Followup cohort had been successfully traced. Interviews were conducted for 9,281 subjects. An interview was conducted for 8,151 of the 8,687 surviving subjects; 551 interviews were administered to a proxy respondent because the subject was incapacitated. A proxy interview was conducted for 1,130 of the 1,392 decedents identified in the 1992 NHEFS. In addition, 10,535 facility stay records were collected for 4,162 subjects reporting overnight facility stays. Death certificates were obtained for 1,374 of the 1,392 subjects who were identified as deceased since last contact. Approximately 32 percent of the NHEFS cohort is known to be deceased with a death certificate available for 98 percent of the 4,604 NHEFS decedents.

The relationship between fish consumption and stroke incidence. The NHANES I Epidemiologic Follow-up Study (National Health and Nutrition Examination Survey).

OBJECTIVE: To assess the level of fish consumption as a risk factor fo r stroke. METHODS: Participants were members of the National Health and Nutrition Examination Survey I Epidemiologic Follow-up Study, a longitudinal cohort study of a national sample. Included in this analysis were white and black women and men aged 45 to 74 years when examined in 1971 through 1975 who did not report a history of stroke at that time. Average follow-up for survivors was 12 years (maximum, 16 years). The main outcome measure was incident stroke (fatal and nonfatal). Fish consumption at baseline was obtained from a 3-month food frequency questionnaire. RESULTS: White women aged 45 to 74 years who consumed fish more than once a week had an age-adjusted risk of stroke incidence only about half that of women who never consumed fish. This effect persisted after controlling for multiple stroke risk variables (relative risk, 0.55;95% confidence interval [CI], 0.32 to 0.93). Fish consumption more than once a week compared with never was not associated with age-adjusted stroke risk in white men aged 45 to 74 years (relative risk, 0.85;95%CI,0.49 to 1.46). In black women and men combined aged 45 to 74 years, any fish consumption compared with never was significantly associated with reduced adjusted stroke risk (relative risk, 0.51;95%CI,0.30 to 0.88).