This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Feeling Better? Trends in General Health Status

¹ Institute of Medicine, National Academies, Washington, D.C.
² Institute for Social Research, University of Michigan, Ann Arbor.
³ University of Medicine and Dentistry of New Jersey, New Brunswick.

Address correspondence to Linda G. Martin, Institute of Medicine, 500 Fifth Street NW, Washington, DC 20001. E-mail: lmartin{at}nas.edu

	Abstract

TOP Abstract Methods Results Discussion References

 
Objectives. We addressed three questions: Have recent improvements in old-age disability been mirrored in changes in self-reported general health status? Are general health status trends similar for younger and older Americans? Have changes in general health status been uniform across demographic and socioeconomic groups?

Methods. Using logistic regression, we analyzed data from the 1982–2003 National Health Interview Surveys (n = 1,445,872 aged 18–69; n = 178,384 aged 70 and older).

Results. The proportion of people aged 70 and older reporting disability declined at 1.38% per year and the proportion 70 and older reporting poor/fair health declined at 1.85% per year. There was less of a decline in reports of poor/fair health at younger ages. Trends for the 18–69 population showed widening health disparities by income but narrowing of the race/ethnicity and education gaps. In the older population, there was no change for those aged 80–84 and 85 and older, the race/ethnicity gap persisted, and both education and income differentials widened over time.

Discussion. Declines in proportions reporting poor/fair health among the older population in recent decades mirror declines in disability. Although the younger population has not experienced such progress, its prevalence of poor/fair health is low throughout the 21-year analysis period. Of concern are the growing socioeconomic disparities in health for both younger and older populations.

AS Americans live longer, researchers and policymakers are asking whether those extra years are spent in good health or bad. Measures of disability have dominated much of the late-life health trends literature, in part because disability is a strong predictor of dependent living and higher medical costs (Trupin, Rice, & Max, 1995), and also because for more than two decades questions regarding disability have been consistently asked of older Americans in a number of nationally representative surveys. For example, drawing on data from the National Long-Term Care Study, Manton and colleagues concluded that disability prevalence among older Americans fell between the early 1980s and late 1990s (Manton, Corder, & Stallard, 1993, 1997; Manton & Gu, 2001). Research using other national data sets and alternative disability measures has largely confirmed that old-age disability prevalence is down (for a review, see Freedman, Martin, & Schoeni, 2002).

Nevertheless, this impressive body of evidence regarding disability trends does not necessarily mean that the health of older people has improved. The most influential conceptual work on disability recognizes that disability is a function of both an individual's underlying health and the social and physical environments in which the individual functions (see, for example, Crimmins, 1996; Pope & Tarlov, 1991; Verbrugge & Jette, 1994); thus, disability is not a pure health measure.

Researchers commonly use self-reported general health assessments (e.g., How would you rate your health? Excellent, very good, good, fair, or poor?) in socioeconomic as well as health surveys, and many studies have found these measures highly predictive of mortality (for reviews of the extensive literature, see Benyamini & Idler, 1999; Idler & Benyamini, 1997), functioning (Idler & Benyamini, 1997), and medical care use (Mutran & Ferraro, 1988). Such assessments, instead of focusing on particular activities of daily living as do disability measures, may represent a more holistic view of health, incorporating a broad array of aspects of physical and mental well-being. But, surprisingly, researchers have not commonly used data on general health status to track trends in health over time. One exception is a study by Waidmann, Bound, and Schoenbaum (1995), which, based on data from the National Health Interview Survey (NHIS), found little change in the 1970s in general health status among the older population but some improvement in the 1980s. Zack, Moriarty, Stroup, Ford, and Mokdad (2004), using data from the Behavioral Risk Factor Surveillance System, found an increase from 1993 to 2001 in the proportion of people aged 18–44 and 45–64 years who reported only poor or fair health, but they found a decline for the 65 and older population. They also noted relatively large increases in reports of poor/fair for the Hispanic population, the least educated, and those with the lowest incomes, but they did not further stratify by age or assess these disparities in multivariate models. Finally, using NHIS data, Goesling (2005) found an increasing gap in poor/fair health by education from 1982 to 2003 for individuals aged 50–69 years and 70 and older.

The goal of this article is to build on the strengths of these studies by using data on general health status spanning the past two decades to examine patterns of change by various age, demographic, and socioeconomic groups. Specifically, the objectives are the following:

1. To determine for the older population whether improvements in old-age disability prevalence over the past two decades have been mirrored in changes in self-reported general health status.
   
2. To assess whether trends in general health status are similar for younger and older adults.
   
3. To investigate whether the changes among younger and older adults have been uniform across a broad range of demographic and socioeconomic groups.
   

	METHODS

TOP Abstract Methods Results Discussion References

 
Data
We based the analysis on data from the NHIS, which was conducted by the National Center for Health Statistics. A repeated cross-sectional survey of the noninstitutionalized population in the United States, the NHIS each year includes roughly 8,000 adults aged 70 and older. The analysis used data from 1982 to 2003 for a total of 1,624,256 people aged 18 and older. This resulted in a sample of 1,445,872 people aged 18–69 and 178,384 people aged 70 and older (excluding 720 70-year-olds in 1982 for whom disability data were not available). The sampling plan followed a multistage area probability design that permitted the representative sampling of households. In all estimations, we used the final basic weights, which were poststratified to represent the civilian noninstitutional population.

Measures of Health
Since 1982, the NHIS has asked a question about general health status of people of all ages residing in the selected households or their proxies, namely: "Would you say your health in general is excellent, very good, good, fair, or poor?" Since that year, NHIS also has employed two key disability questions for people aged 71 and older in 1982 and aged 70 and older from 1983 on. The first asks about limitations regarding activities of daily living: "Because of any impairment or health problem, does ___ need help of other persons with personal care needs, such as eating, bathing, dressing, or getting around this home?" If a respondent answered no to this question, interviewers then asked him or her about limitations regarding instrumental activities of daily living: "Because of any impairment or health problem, does ___ need help of other persons in handling routine needs, such as everyday household chores, doing necessary business, shopping, or getting around for other purposes?" For this analysis, we combined answers to the two questions in order to yield prevalence of any disability. From 1997 on, interviewers asked everyone aged 70 or older the instrumental activities of daily living question, no matter what their answers to the activities of daily living question. Also survey designers modified slightly the preambles to the disability questions in 1997 and dropped the requirement that the limitation last at least 3 months. After matching the pre-1997 skip logic, chronicity requirement, and editing strategy, we found that the survey changes did not create an obvious discontinuity in estimates of any disability. But to be prudent and to control for other possibly unmeasured aspects of the NHIS 1997 redesign, we included an indicator of post-1996 interview in all our analyses of disability and self-reported health. We assessed the sensitivity of the latter to this inclusion.

Demographic, Survey, and Socioeconomic Measures
Age
For graphs comparing younger and older adults, we used single-year data; for models of the younger population (aged 18–69), the age groups were 18–29 and 10-year groups up to age 69; for models of the entire population (aged 18 and older), the age groups were the same as for the younger population, and we used one group for people aged 70 and older; and for models of the older population (aged 70 and older), the age groups were 70–74, 75–79, 80–84, and 85 and older. No age detail beyond age 85 is available in the NHIS. In general, we expected to find poorer health as age increased, although researchers have found that older respondents in some surveys give disproportionately positive health assessments (Idler, 1993).

Gender
We controlled for gender in all models. We expected that, at least through middle age, women would be more likely than men to report poor or fair health (Case & Paxson, 2005; McCullough & Laurenceau, 2004).

Proxy
We included in all models an indicator of whether a proxy had provided a response, because proxy- and self-reports may differ systematically (Magaziner, Zimmerman, Gruber-Baldini, Hebel, & Fox, 1997). We also interacted the proxy indicator with an indicator of whether the interview was conducted in 1997 or later. Prior to 1997, NHIS provided a variable indicating if someone other than the individual responded in part or fully for that individual. From 1997 on, the proxy variable indicated whether the individual of interest was the respondent for the family.

In the analysis of trends in disparities in general health status, we included variables indicating marital status, region, race/ethnicity, education, and income.

Marital status
A dichotomous variable indicated whether the respondent was married. Being married is associated with lower mortality, at least through reproductive ages (Goldman, 1993), and researchers generally have found that it confers health benefits (Waite & Gallagher, 2000). However, the positive association of general health status and marriage may be ameliorated among women and people experiencing marital transitions (Williams & Umberson, 2004).

Region
We classified respondents' locations at time of interview into four broad regions—Northeast, Midwest, South, and West. Other studies have shown a particular health disadvantage among those in the South (Lin & Zimmer, 2002; Pickle, Mungiole, Jones, & White, 1996; Porell & Miltiades, 2002).

Race/ethnicity
We compared the non-Hispanic White population to all other racial/ethnic groups combined with the expectation of poorer health for the latter (Hummer, Benjamins, & Rogers, 2004). We also explored the sensitivity of results to using a Black versus non-Black categorization for race/ethnicity.

Education
We expected that individuals of higher socioeconomic status, as measured by either education or income, would report better health (e.g., House et al., 1994). We classified education into six groups: 0–8 years, 9–11 years , 12 years, 13–15 years, 16 or more years, and don't know. The tables do not report estimates for this last category, which included less than 3% each of the 18–69 and of the 70 and older populations from 1982 to 2003.

Income
We used a relative measure of income, normed annually and separately for the 18–69 and 70 and older groups. We ranked respondents in each of the two broad age groups by their family income and divided them into quartiles for each year. In survey years 1982–1996 (1997–2003), respondents reported family income as being in one of 26 (11) categories. To estimate income quartiles, we used a three-step procedure to calculate for each respondent a continuous income amount within the category reported by the respondent. First, for each year 1982 to 2003, we used the comparably aged population from the March Current Population Survey (the U.S. Census Bureau's source for official estimates of income and poverty) to estimate family income as a function of demographic and socioeconomic variables and the family income categories appearing in the NHIS. Second, we used estimates from this model to calculate an exact family income within the category reported in the NHIS for each respondent. Finally, we grouped individuals in the NHIS into income quartiles for each of the two broad age groups. We evaluated the procedure by comparing the March Current Population Survey and the calculated NHIS income distributions and trends, and we found that they were substantially similar.

Statistical Analyses
To facilitate the comparison of trends in general health status at older ages to trends in disability, we show graphically unadjusted estimates of proportions reporting any disability and of proportions reporting poor/fair or excellent health, as well as mean general health status (calculated by assigning scores ranging from 5 for poor to 1 for excellent) for the older population in each year from 1982 to 2003. We display a third-order polynomial that we fit to each time series. Subsequent analyses fit logistic models for any disability and for poor/fair health (with the reference category of excellent/very good/good) that included as control variables age, gender, proxy response, post-1996 interview, and the interaction of the last two. The key explanatory factor was a linear trend variable that took the value of 0 in 1982 and increased by 1 in each subsequent year, with a maximum value of 21 in 2003. We adopted a more parsimonious linear specification because the second- and third-order polynomial terms (shown graphically) were not statistically significant. For all of our models, we used SUDAAN software to adjust statistical tests for the complex survey design.

To compare trends in general health status at younger and older ages, we reproduced for the younger population the earlier graph for the older population of trends in self-reports of general health status. Next, we showed graphically proportions reporting poor/fair health in the 18 and older population by single years of age for 1982, 1992, and 2002, and by single years of age and selected birth cohorts. We also fit a logistic regression model for the 18 and older population, contrasting poor/fair with excellent/very good/good health. Explanatory variables included age, gender, proxy, post-1996 interview, the interaction of proxy and post-1996 interview, as well as interactions of the linear trend variable with age.

To explore demographic and socioeconomic disparities in trends in general health status, we estimated a set of logistic regressions for poor/fair health (vs excellent/very good/good) for the 18–69 and 70 and older populations. We conducted the analyses for these two populations separately, because preliminary work indicated that some of the findings differed substantially between them. These regressions serially allowed interaction of the linear trend variable with age, gender, marital status, region, race/ethnicity, education, and income. All models controlled for simple effects of all other variables. Besides standard tests of differences from zero, we also present tests for statistically significant differences in the estimated trends across categories within a variable.

Finally, we assessed the sensitivity of the general health status trend results for the 70 and older population to the exclusion of the institutionalized population from the NHIS. The National Nursing Home Surveys of 1985, 1995, 1997, and 1999 provided estimates of the number of nursing home residents by age. To estimate an upper boundary of poor/fair health, we assumed that all residents would report poor or fair health and added them to the numbers by age from the NHIS. Of course, it could be that living outside a nursing home has a positive influence on self-perception of health and thus could contribute to improvements at the population level, but it is not possible to quantify such an effect, however large or small.

The institutionalized younger population is more likely to be found in the military or correctional facilities. Over the study period, there was a decrease in the former (U.S. Census Bureau, 2006), who may be healthier on average, and an increase in the latter (U.S. Bureau of Justice Statistics, 2006), who may be less healthy. However, it is not reasonable to make such global assumptions, and thus we did not conduct sensitivity analyses for the younger population.

	RESULTS

TOP Abstract Methods Results Discussion References

 
Trends in General Health Status and Disability Among the 70 and Older Population
Figure 1a provides an overview of trends from 1982 to 2003 in any disability and general health status among people aged 70 years and older. The top data series for poor/fair health status and the associated third-order polynomial show a clear downward trend. Over the same period, however, there was a decline in the proportion reporting excellent health. Not shown, but obvious, is that the percentage reporting good or very good health increased. For all years, the proportion reporting poor/fair health (27%–36%) was much greater than the proportion reporting excellent health (11%–16%). The plot of mean general health status indicates little change over time. The graph also shows a decline in any disability over the same period.

View larger version (26K): [in this window] [in a new window]   Figure 1 a. Health Status of Population 70 and Older: 1982–2003. Figure 1b. Health Status of Population 18–69: 1982–2003.

Figure 2 shows poor/fair health by single year of age in selected calendar years, starting with the earliest year and followed by decadal intervals: 1982, 1992, and 2002. For most ages, the data points for 2002 are lower than those for the other years. Given the low percentage reporting poor or fair health at young ages, the change over time in those age groups is difficult to discern. But from age 50 to the late 70s, the proportions reporting poor or fair health clearly declined over time.

View larger version (16K): [in this window] [in a new window]   Figure 2. Percent Poor/Fair by Age: 1982, 1992, and 2002

View larger version (21K): [in this window] [in a new window]   Figure 3. Percent in Poor/Fair Health by Age and Birth Cohort. Panel A. Birth Cohorts Ending in 2. Panel B. Birth Cohorts Ending in 7.

Model 2 in Tables 3 and 4 estimated trends separately by age group within each population while still controlling for all of the other variables. In Table 3, there was a decline in reports of poor/fair health for the 18–29 group. There was no significant trend for the 30–39 and 60–69 groups, and the trend is upward for the 40–49 and 50–59 groups. As shown in Table 4, we found the greatest decline among the older population in the 70–74 group, followed by the 75–79 group. There were not statistically significant changes for the two oldest groups.

Among the younger population, women were more likely than men to report poor/fair health in the base year of 1982, and their relative disadvantage increased with time (Model 3 of Table 3). In the older population there was no gender differential (Model 3 of Table 4), but there were declines in poor/fair health for both women and men that were indistinguishable statistically.

General patterns by marital status differed substantially between the younger and older populations (Models 4). Perhaps surprisingly, in the 70 and older population, those who were married were more likely to report poor/fair health than those who were not, but the declines over time were significantly greater for the married, so the gap narrowed. In the 18–69 population, the baseline differential was reversed, and the relative health advantage of those who were married increased significantly over time.

In both populations, people interviewed in the South were more likely to report poor/fair health than were those interviewed in the West (the reference category). Among the younger population, the Northeast was the healthiest region in 1982, but its trend was upward in comparison to the West, so there was some closing of the regional health gap. In the older population, the regional trends were statistically the same (Model 5).

In both populations, non-Hispanic Whites reported less poor/fair health than the all-others group (Models 6). However, the trend results indicated an increase for the former and a narrowing of the gap among people 18–69, whereas the disparity persisted in the older group. Analysis (not shown) contrasting Blacks and non-Blacks, the latter of whom reported relatively more poor/fair health, yielded essentially the same results with a narrowing of the gap for the younger population and no change for the older.

Models 7 in Tables 3 and 4 highlight trends in poor/fair health by education group. In the younger population, there was some narrowing of the educational gap over time, with the trend downward for the least educated being the greatest and with a relative trend upward for the 12-year and 13–15-year groups. In the older population, there was a widening of the educational gap in health over time, with the most educated group's trend downward larger than that of the least educated group.

Similarly, Models 8 in Tables 3 and 4 show in both populations that the two higher income quartiles (income quartiles 3 and 4) experienced the greatest declines in poor/fair health over time, so there was a widening health disparity by income.

The assumption that nursing home residents all would report poor or fair health only slightly ameliorated the general health status trend found here for the older population. From 1985 to 1999, the prevalence of poor/fair health declined by 14.8% for the community-based 70 and older population; the decline was 13.8% for the total 70 and older population. For the 70–74 and the 75–79 groups, the results were not sensitive to including the institutional population. For the 80–84 and 85 and older groups in which institutionalization is more common, the addition yielded even greater declines in poor/fair health, as a result of the decline in the percentage institutionalized from 1985 to 1999. For people aged 80–84, there was a 3.2% decline in poor/fair health based on NHIS data but a 5.6% decline based on NHIS plus National Nursing Home Survey data from 1985 to 1999. The declines for the 85 and older population were 0.5% (NHIS only) and 5.9% (NHIS plus National Nursing Home Survey data).

	DISCUSSION

TOP Abstract Methods Results Discussion References

 
For both the younger and older populations, there were overall declines in proportions reporting poor or fair health over the 22-year period. For the older population, this decline is consistent with previously reported declines in the prevalence of disability. The lack of improvement in health in the 80–84 and 85 and older populations found in the multivariate model of Table 4 may reflect the inability to control for the aging of the oldest group in the analysis, as well as the omission of the institutional population. This finding—that older people's broader self-assessment of health is improving—provides some reassurance that trends in self-reports of disability, which may reflect changes in living and working environments as well as social role expectations, are also at least in part indicative of better perceived health. Whether such improvements in perceived health reflect actual improvements in health is less clear, as reports of general health status may capture more than underlying health. For example, they may reflect changing health expectations or changing perceptions of particular diseases and conditions (Idler, 1993; Schnittker, 2005). Nevertheless, the reduction in poor/fair health reports by the older population is good news, given the previously identified relation of general health status to mortality, morbidity, and medical care use.

Our results for trends in general health status for the younger population, which is the older population of the future, are less encouraging. The simple model of Table 2 with age and gender variables did show declines for all age groups in the 18–69 range except the 30–39 group. But the declines were smaller for the younger groups. The upward trends for the 40–49 and 50–59 groups in the sociodemographic model are concerning. Even so, only small proportions of the younger population report poor/fair health in any one year. Moreover, our inclusion in the analysis of poor/fair health of a variable to reflect the 1997 survey redesign may be capturing some of the true declines in poor/fair health and thus may be resulting in conservative trend estimates.

This assessment of trends in general health status for the younger population is somewhat similar to the findings of Zack and colleagues (2004). Their study, which found unadjusted upward trends in poor/fair health for those aged 18–64, relied on data obtained via telephone interviews for the shorter time period of 1993 to 2001. Lakdawalla, Bhattacharya, and Goldman (2004), who analyzed trends in disability for people aged 18–69 from 1984 to 2000, using NHIS data, also raised concerns about the younger population. For those younger than 60, but especially for those aged 30–59 from 1984 to 1996, the authors found increases in disability, albeit from a low initial prevalence.

In the analysis of trends in disparities in health, we found many of the expected gradients in health at baseline (e.g., differentials by age, region, race/ethnicity, education, and income). The differentials by gender and marital status were somewhat surprising. Case and Paxson (2005) found that the poorer self-assessed health of women versus men through middle age reflects differences in chronic diseases that they face. But men experience greater severity (as measured by hospitalization) and mortality in association with cardiovascular disease and some lung disorders (both generally related to smoking). This mortality selection may account in part for the lack of gender differential in poor/fair health at older ages. We also found that the marriage advantage in health at younger ages was reversed in the older population, in which being unmarried, especially for women, becomes more normative with age.

Our analysis of disparities in trends provided both good and bad news. On the positive side for those 18–69 are the narrowing gaps in health by race/ethnicity and by education. For the older population, there was no change in the gap by race/ethnicity and there was a widening gap by education. Both the older and younger populations experienced substantial increases in educational attainment over the two decades considered here, but the older population overall had less education than the younger population. Older people with the least education appear to be increasingly negatively selected for health.

The growing gap in health by income quartile for both the younger and the older populations also is troubling. Our findings for the 70 and older population of growing disparities in general health status by both education and income and persistent disparities by race/ethnicity parallel our recent work on trends in disparities in disability (Schoeni, Martin, Andreski, & Freedman, 2005).

There is much work to be done in explaining trends over time in both health measures. Analysis of time series of cross-sections has not yet provided a full understanding of the causal mechanisms. See, for example, studies of the influence on trends in disability and functional limitations among the older population of trends in education (Freedman & Martin, 1999) and trends in chronic diseases (Crimmins & Saito, 2000; Cutler, Landrum, & Smith, 2006; Freedman & Martin, 2000). The current study and these other works focus on age- and period-specific measures of health, but the alternative perspective of age and cohort may illuminate mechanisms through which disparities in health change with age and may suggest fruitful lines of investigation for understanding time trends. For example, Herd (2006) explored reasons for the declining socioeconomic disparities in functional limitations by age within a single cohort approaching retirement and found that change in social context (such as the availability of Medicare) rather than mortality selection was the more likely mechanism through which disparities declined with age. Moreover, she found that even for the most educated, the onset of health problems was delayed only so long.

Sorting out the roles of trends in socioeconomic factors, medical care, and underlying biological processes, among other possible explanations for trends in health, is a tall order. In the meantime, the development of cost-effective interventions to improve the health of disadvantaged populations, young and old, could help extend into the future the encouraging overall health trends that we found for the older population over the past two decades.

	Acknowledgments

 
This research was supported by National Institute on Aging Grant R01-AG021516. An earlier version was presented at the annual meeting of the Population Association of America, Los Angeles, March 26–April 1, 2006. We thank Susan S. Jack of the National Center for Health Statistics for providing insight into design changes in the National Health Interview Survey. The views expressed here are our own and do not reflect the opinions of our sponsor or institutions.

	Footnotes

 
Decision Editor: Kenneth F. Ferraro, PhD

Received for publication March 14, 2006. Accepted for publication July 21, 2006.

	References

TOP Abstract Methods Results Discussion References

 

• Benyamini, Y., Idler, E. L. (1999). Community studies reporting association between self-rated health and mortality: Additional studies, 1995–1998. Research on Aging, 21,392-401.[Free Full Text]
• Case, A., Paxson, C. (2005). Sex differences in morbidity and mortality. Demography, 42,(2), 189-214.[Medline]
• Crimmins, E. M. (1996). Mixed trends in population health among older adults. Journal of Gerontology: Social Sciences, 51B,S223-S225.[Medline]
• Crimmins, E. M., Saito, Y. (2000). Change in the prevalence of diseases among older Americans: 1984–1994. Demographic Research, 3(9). Retrieved March 10, 2006, from http://www.demographic-research.org/Volumes/Vol3/9/3-9.pdf.
• Cutler, D. M., Landrum, M. B., Smith, K. A. (2006). Intensive medical care and cardiovascular disease disability reductions. NBER Working Paper Series, 12184.
• Freedman, V. A., Martin, L. G. (1999). The role of education in explaining and forecasting trends in functional limitations among older Americans. Demography, 36,461-473.[Medline]
• Freedman, V. A., Martin, L. G. (2000). Contribution of chronic conditions to aggregate changes in old-age functioning. American Journal of Public Health, 90,1755-1760.[Abstract/Free Full Text]
• Freedman, V. A., Martin, L. G., Schoeni, R.F. (2002). Recent trends in disability and functioning among older adults in the United States: A systematic review. Journal of the American Medical Association, 288,3137-3146.[Abstract/Free Full Text]
• Goesling, B. (2005, March 31–April 2). Educational disparities in health, 1982–2003. Paper presented at the annual meeting of the Population Association of America, Philadelphia, PA.
• Goldman, N. (1993). Marriage selection and mortality patterns: Inferences and fallacies. Demography, 30,189-208.[Medline]
• Herd, P. (2006). Do functional health inequalities decrease in old age? Educational status and functional decline among the 1931–1941 birth cohort. Research on Aging, 28,375-392.[Abstract/Free Full Text]
• House, J. S., Lepkowski, J. M., Kinney, A. M., Mero, J. P., Kessler, R. C., Herzog, A. R. (1994). The social stratification of aging and health. Journal of Health and Social Behavior, 35,213-234.[Medline]
• Hummer, R. A., Benjamins, M. R., Rogers, R. G. (2004). Racial and ethnic disparities in health and mortality among the U.S. elderly population. In N. B. Anderson, R. A. Bulatao, & B. Cohen (Eds.), Critical perspectives on racial and ethnic differences in health in late life (pp. 53–94). Washington, DC: National Academies Press.
• Idler, E. L. (1993). Age differences in self-assessments of health: Age changes, cohort differences, or survivorship? Journal of Gerontology: Social Sciences, 48,S289-S300.
• Idler, E. L., Benyamini, Y. (1997). Self-rated health and mortality: A review of twenty-seven community studies. Journal of Health and Social Behavior, 38,(1), 21-37.[Medline]
• Lakdawalla, D., Bhattacharya, J., Goldman, D. (2004). Are the young becoming more disabled? Health Affairs, 23,(1), 168-176.[Abstract/Free Full Text]
• Lin, G., Zimmer, Z. A. (2002). A geographical analysis of spatial differentials in mobility and self-care limitations among older Americans. International Journal of Population Geography, 8,395-408.
• Magaziner, J., Zimmerman, S. I., Gruber-Baldini, A. L., Hebel, J. R., Fox, K. M. (1997). Proxy reporting in five areas of functional status: Comparison with self-reports and observations of performance. American Journal of Epidemiology, 146,418-428.[Abstract/Free Full Text]
• Manton, K. G., Corder, L., Stallard, E. (1993). Estimates of change in chronic disability and institutional incidence and prevalence rates in the U.S. elderly population from the 1982, 1984, and 1989 National Long Term Care Survey. Journal of Gerontology: Social Sciences, 48,S153-S166.
• Manton, K. G., Corder, L., Stallard, E. (1997). Chronic disability trends in elderly United States populations: 1982–1994. Proceedings of the National Academy of Sciences, USA, 94,2593-2598.[Abstract/Free Full Text]
• Manton, K. G., Gu, X. L. (2001). Changes in the prevalence of chronic disability in the United States black and nonblack population above age 65 from 1982 to 1999. Proceedings of the National Academy of Sciences, USA, 98,6354-6359.[Abstract/Free Full Text]
• McCullough, M. E., Laurenceau, J. P. (2004). Gender and the natural history of self-rated health: A 59-year longitudinal study. Health Psychology, 23,651-655.[Medline]
• Mutran, E., Ferraro, K. F. (1988). Medical need and use of services among older men and women. Journal of Gerontology: Social Sciences, 43,S162-S171.
• Pickle, L. W., Mungiole, M., Jones, G., White, A. A. (1996). Atlas of United States mortality. Hyattsville, MD: National Center for Health Statistics.
• Pope, A. M., Tarlov, A. R., (Eds.). (1991). Disability in America: Toward a national agenda for prevention. Washington, DC: National Academy Press.
• Porell, F. W., Miltiades, H. B. (2002). Regional differences in functional status among the aged. Social Science and Medicine, 54,1181-1198.
• Schnittker, J. (2005). When mental health becomes health: Age and the shifting meaning of self-evaluations of general health. Milbank Quarterly, 83,(3), 397-423.[Medline]
• Schoeni, R. F., Martin, L. G., Andreski, P., Freedman, V. A. (2005). Persistent and growing socioeconomic disparities in disability among the elderly: 1982–2002. American Journal of Public Health, 95,2065-2070.[Abstract/Free Full Text]
• Trupin, L., Rice, D. P., Max, W. (1995). Medical expenditures for people with disabilities in the United States, 1987. Washington, DC: U.S. Department of Education, National Institute on Disability and Rehabilitation Research.
• U.S. Bureau of Justice Statistics. (2006). Corrections statistics. Retrieved October 27, 2006, from http://www.ojp.usdoj.gov/bjs/correct.htm.
• U.S. Census Bureau. (2006). Statistical abstract of the United States (125th ed.). Retrieved October 27, 2006, from http://www.census.gov/compendia/statab/.
• Verbrugge, L. M., Jette, A. M. (1994). The disablement process. Social Science and Medicine, 38,(1), 1-14.
• Waidmann, T. A., Bound, J., Schoenbaum, M. (1995). The illusion of failure: Trends in the self-reported health of the U.S. elderly. Milbank Quarterly, 73,(2), 253-287.[Medline]
• Waite, L. J., Gallagher, M. (2000). The case for marriage. New York: Doubleday.
• Williams, K., Umberson, D. (2004). Marital status stability, marital transitions, and health: A gendered life course perspective. Journal of Health and Social Behavior, 45,(1), 81-98.[Medline]
• Zack, M. M., Moriarty, D. G., Stroup, D. F., Ford, E. S., Mokdad, A. H. (2004). Worsening trends in adult health-related quality of life and self-rated health: United States, 1993–2001. Public Health Reports, 119,493-505.[Medline]

This article has been cited by other articles:

				L. G. Martin, V. A. Freedman, R. F. Schoeni, and P. M. Andreski Health and Functioning Among Baby Boomers Approaching 60 J Gerontol B Psychol Sci Soc Sci, May 1, 2009; 64B(3): 369 - 377. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation