|
|
Volume 2: No.
4, October 2005
ORIGINAL RESEARCH
Predictors of Self-rated Health Status Among Texas Residents
Lorraine J. Phillips, MSN, RN, FNP, Renee L. Hammock, RN, MSN, FNP-C, Jimmy
M. Blanton, MPAff
Suggested citation for this article: Phillips LJ, Hammock RL, Blanton JM. Predictors
of self-rated health status among Texas residents. Prev Chronic Dis [serial
online] 2005 Oct [date cited]. Available from: URL: http://www.cdc.gov/pcd/issues/2005/
oct/04_0147.htm.
PEER REVIEWED
Abstract
Introduction
The purpose of this study was to investigate the predictors of self-rated
health status for Texas adults using the current 2003 Behavioral Risk Factor
Surveillance System data. Self-rated health is generally accepted as a valid
measure of health status in population studies, and understanding its
correlates may help public health professionals prioritize health-promotion
and disease-prevention interventions.
Methods
The two research questions addressed by this study involved the predictors of
self-rated health: 1) "Do demographic characteristics, health care coverage,
leisure-time physical activity, and body mass index predict self-rated health status for
Texas residents aged 18 to 64 years?” and 2) “Does choice of interview
language (English vs Spanish) predict self-rated health status for Texas
residents of Hispanic
ethnicity aged 18 to 64 years?” Key analysis variables were identified, and
descriptive statistics were used to describe the major variables and determine
whether the number of respondents for each variable was sufficient for
analysis. Multivariate regression analysis was used to assess the variables.
Results
Multiple logistic regression analysis (controlling for diabetes and
arthritis) of the self-rated health predictors indicated that older age, lack
of health care coverage, lack of a college education, being Hispanic, having a
lower income, obesity, and not exercising explained 19.4% of the variance of
fair and poor self-rated health. The interview language (English or Spanish), age, sex,
education, income, obesity, health insurance coverage, and physical activity
(controlling for chronic illness) explained 22.8% of the variance in fair and
poor self-rated health for Hispanic respondents.
Conclusion
The results of this study suggest that a college education, a
lower body mass index, non-Hispanic ethnicity, and participation in physical activity
are associated with good, very good, or excellent self-rated health status.
The finding that the interview language significantly predicted fair and poor
self-rated health substantiates previous research and emphasizes the
importance of culturally sensitive approaches to health care services.
Back to top
Introduction
The first goal of the U.S. Department of Health and Human Services’ (DHHS’s)
Healthy People 2010 is to help individuals in the United States improve
their quality of life and life expectancy (1). With numerous other federal,
state, and local agencies, DHHS monitors the health of individuals,
communities, and the nation. When a particular health issue is identified,
objectives that focus on strategies to reduce the severity of or eliminate the
problem are developed, and many of these objectives are included in Healthy
People 2010. Healthy People 2010 also includes a model of health
determinants that includes the following components: individual biology and
behavior, individual social and physical environments, policies and
interventions, and access to quality health care. All of these factors may
interact with and affect the health of an individual or a society.
Improving the health of people living in the United States requires an
initial assessment of their health
status. Various instruments exist to measure perceived
health. One such instrument is simply a question that asks
people to rate their health as poor, fair, good, very good, or excellent.
The Centers for Disease Control and Prevention (CDC) uses this self-reported global
assessment of health-related quality of life in the annual Behavioral Risk
Factor Surveillance System (BRFSS).
In population studies, self-rated health is generally accepted by
researchers as a
valid measure of health status. Because it is able to predict
risk of death, self-rated health information measures not only
psychological well-being but also overall health (2,3). Understanding the
correlates of self-rated health may help health care professionals tailor
health-promotion and disease-prevention interventions to the needs of specific
populations. People who rate themselves as being in poor health tend to
lack health care insurance (4-6), be women, be older, be black (7), and report
lower psychological well-being (8). Alternatively, people who
report that they are in good to excellent health tend to report higher
vitality, a more positive mood, less vulnerability to illness (9), more
frequent regular exercise (10,11), more education, and a higher income
(7,12).
Previous studies have found that the relationship between a good or an
excellent health rating and regular physical activity is stronger in men than
women (8). In contrast, sex differences were not found between men and women
in the same age group whose risk of death increased when they
reported a lower level of physical fitness (13). Okosun et
al (2) found that the association between obesity and less than excellent self-rated
health was more pronounced in men than women, although a
significant trend of fewer self-reports of excellent health with increases in
obesity was found in both sexes and all racial/ethnic groups. Because obesity
is associated with an increased risk of developing a chronic disease or
condition, such as type 2 diabetes, high blood pressure, coronary heart
disease, a high blood cholesterol level, osteoarthritis, or gallbladder
disease (14), the lower self-reported health status ratings by obese
individuals support the claim that self-rated health measures can reflect
overall health.
A meta-analysis by Idler and Benyamini (15) showed that in 23 of 27
studies, self-ratings of health (independent of known health risk factors)
reliably predicted survival, or life span, in the populations surveyed. The
parsimonious global self-rating of health provides an invaluable and a unique
assessment of health status. When respondents answer the question, “How in
general would you rate your health?” the answer includes perceptions of
their physical, mental, and social constitution. Whether self-rated health
reveals unknown conditions, such as an undiagnosed disease, or is the most
inclusive summary of all other influences on health (e.g., financial and
personal resources, health behaviors, familial risk factors) is less relevant than
its power to predict death (15). Because the correlation
between self-rated health and mortality is well established, Idler and Benyamini also propose that future research on self-rated health status should
focus on measures of morbidity, particularly those that increase mortality,
such as new cases of heart disease, cancer, stroke, or diabetes.
The purpose of this study was to investigate the predictors of self-rated
health status for Texas adults using 2003 BRFSS data. In 2003, Texas ranked
fifth in the United States for the percentage of people who rated their health
as fair or poor in the BRFSS; West Virginia ranked first, Mississippi
second, Kentucky third, and Alabama fourth (16). In addition, Texas
was second only to West Virginia in the percentage of people who reported having
less than a high school education, and Texas residents reported less
leisure-time physical activity than residents of 42 other states and the
District of Columbia. Income levels tended to be lower in Texas, with only
seven other states reporting higher percentages of households earning less
than $25,000. Finally, in 2003, Texas had the highest percentage of uninsured
people in the nation, with 26.6% reporting a lack of health care coverage (17).
The Hispanic/Latino population in Texas comprises 32% of the
total population (18), increasing the chance that the typical categorical
responses of all Texans on a self-rated health status scale will have
cross-cultural differences. In other words, the adjectives associated with
normal health may differ between Hispanics and non-Hispanics. An analysis of
the Hispanic Health and Nutrition Examination Survey (HHANES) revealed
that the language used by the interviewer has a significant effect on
self-ratings of health (19). Angel and Guarnaccia (19) reported that
respondents who were interviewed in Spanish were much less likely to report
excellent health (15%) or good health (48%) than people who were interviewed
in English. Of respondents interviewed in Spanish, almost half of the people who rated their health as fair or poor were
rated by a physician as having very good or excellent health,
suggesting that level of acculturation (measured by language of interview)
significantly affects self-ratings of health. Given
the prevalence of potential predictors of fair or poor self-rated health in
Texas, identification of these factors may help guide the direction of future
research and health-promotion interventions.
Based on findings in the available research, we developed the following questions for this study:
- While controlling for chronic illness, do demographic characteristics,
health care coverage, leisure-time physical activity, and body mass index
(BMI) predict self-rated health status for Texas residents aged 18 to 64
years? (Arthritis and diabetes were chosen to represent the chronic disease
state.)
- While controlling for chronic illness, does choice of interview
language (English vs Spanish) predict self-rated health status for Texas
residents of Hispanic
ethnicity aged 18 to 64 years?
Back to top
Methods
Our study was an analysis of the 2003 BRFSS data. The BRFSS — a
state-based, ongoing telephone survey of persons aged 18 years and older —
links behavior risk factors to chronic illness in the adult population. State
health departments conduct the survey in conjunction with the CDC.
Participants are selected using a random-digit–dialing method to gather a
representative sample of noninstitutionalized adults. The data were weighted
and poststratified to adjust for demographic differences between the sample
and known Texas demographics. The sample of interest for this study
was adults aged 18 to 64 years who were residing in Texas
(N = 4091).
The BRFSS has three sections:
- Core questions, which are asked in every state in the same order, using the same
precise instructions. In 2003, the BRFSS core
had 20 question modules including topics such as respondent demographics, health
status, access to care, and exercise frequency.
- Optional question modules, which are modules that are
supported by the CDC and are optional for each state. The optional modules are
typically used to gather in-depth information about a specific subject such as
asthma, diabetes, or tobacco use.
- Additional questions,
which are developed and added by each state.
In 2003, the state-added questions used by Texas were related to vitamin use,
physical activity, and weight loss.
The data for our analysis involved only questions from the core module. The
dependent variable in the study — self-rated health — was measured by the
question, “Would you say that in general your health is excellent, very
good, good, fair, or poor?”
Key analysis variables identified for the study included the following: 1)
age (18 to 44 years and 45 to 64 years); 2) health care coverage (yes or no);
3) education (less than high school, high school graduate or some college,
and college graduate); 4) sex (male or female); 5) race/ethnicity
(white, black, Hispanic, or other); 6) household income (<$25,000, $25,000
to $74,999, or ≥$75,000); 7) BMI, calculated from weight and
height (BMI = kg/m2) — not obese (BMI <30) or obese (BMI ≥30); 8) whether physical
activity or exercise other than that involved in a regular job had been
performed in the past month (yes or no); 9) interview language for people of
Hispanic ethnicity (English or Spanish); and 10) self-rated health status
(excellent, very good, good, fair, or poor).
To accommodate the complex sampling design of the BRFSS, data analysis was
performed using SPSS, version 12.0 (SPSS Inc, Chicago, Ill) in
conjunction with SUDAAN (Research Triangle Institute, Research Triangle Park,
NC). Descriptive statistics were used to describe the major variables and
determine whether the number of respondents for each variable was sufficient
for analysis. To address the first research question (“Do demographic
characteristics, health care coverage, leisure-time physical activity, and BMI
predict self-rated health status for Texas residents aged 18 to 64 years?”),
a multivariate logistic regression analysis was used to assess self-rated
health while controlling for chronic illness. As predictor variables, the
analysis included the variables that significantly correlated with the
dependent variable (self-rated health). Household income, education,
exercise, and BMI were found to correlate strongly with self-rated health, as
were race/ethnicity, health care coverage, and age. Because marital
status did not have a statistically significant correlation with self-rated
health, the variable was not included in the final model. We controlled for
the confounding influence of chronic illness on the explanatory power of the
logistic model by including arthritis and diabetes that had been diagnosed by
a physician. In our study, the dependent variable was dichotomized into two
categories: 1) fair/poor health and 2) good/very good/excellent health. The
second research question (“Does choice of interview language [English vs Spanish] predict self-rated health status
for
Texas residents of Hispanic
ethnicity aged 18 to 64 years?”) was also addressed by multivariate logistic
regression analysis (while controlling for chronic disease), with descriptive statistics included for reference. Statistical significance for
both analyses was set at P <.001.
Back to top
Results
Table 1 presents the distribution
of the sample among the categories of the dependent variable, self-rated
health. Overall, older respondents, women, respondents from households with an
income of less than $25,000 per year, obese individuals, and respondents who
participated in no exercise other than that required to perform their job
rated their health as poor. Respondents who rated their health as excellent
were younger, had health care coverage, had a college degree, were white, had
a household income of greater than $75,000 per year, were of a normal weight,
and reported participating in physical activity or exercise other than that
required in their regular job. Most of the respondents classified themselves as white, had a
high school
education, and had health care coverage. Although the majority
of respondents reported an annual household income greater than $25,000, a
separate analysis of income by ethnicity revealed that 24.8% of Hispanic
respondents reported an annual income of less than $15,000 in 2003.
Table 2 is a summary of the multiple logistic regression analysis results
for the predictors of fair/poor self-rated health. The analysis shows that (when
controlling for diabetes and arthritis) older age, lack of health care
coverage, having less than a college education, having a Hispanic ethnicity,
having a lower income, being obese, and not exercising explained 19.4% of the
variance of fair/poor self-rated health (R2 = 0.1942). Sex and a
race/ethnicity designation of black or “other” were not significantly
associated with fair/poor health.
For an additional test of the impact of culture and interview language on
self-rated health, a multiple logistic regression was used to analyze
respondents of Hispanic ethnicity
(Table 3).
The following independent variables were
included: choice of interview language, age, sex, education, income, BMI,
health insurance coverage, and physical activity. The final model controlled
for chronic illness, did not include the sex variable, and explained 22.8% of
the variance in fair/poor self-rated health (R2 = 0.2281).
The participants who chose to be interviewed in Spanish were significantly
more likely to rate their health as fair/poor than were participants who chose
English.
Back to top
Discussion
The results of this study suggest that higher education, a lower BMI,
non-Hispanic ethnicity,
and participation in physical activity are consistently associated with
good, very good, or excellent health status. Because education, BMI, and physical
activity are modifiable, these findings underscore the importance of
including physical activity and nutrition education in public health programs.
For instance, in the United States in 2003, medical costs attributable to
obesity were estimated as being $75 billion (20); the Texas estimated costs
were $5.34 billion (20). Being overweight significantly increases the risk of developing a chronic illness (21). Women with a BMI greater
than 35 were 17 times more likely to develop diabetes than were women
with a BMI of less than 25, and men with a BMI greater than 35 were 23
times more likely to develop diabetes than were men with a BMI of less than 25 (21). Effective weight control
and reduction programs not only may save billions of U.S. health care
dollars but also may reduce the incidence of chronic disease associated with
obesity.
Brown et al (22) assessed the association between levels of physical
activity and health-related quality of life and found that the relative odds
of having 14 or more unhealthy days (physically or mentally unhealthy) were significantly
lower for people who met recommended levels of physical activity than for
physically inactive adults across all age, racial/ethnic, and sex groups.
Collectively, poor diet and physical inactivity were second only to tobacco
use as the leading cause of death in the United States in 2000 (23). A lifestyle with a poor diet and physical
inactivity not only increases risk of death but also results in
years of lost life, diminished productivity, high rates of disability, and a
decreased quality of life (23). The results of our study concur
with Mokdad et al’s assessment that fair/poor self-rated health
was related to a lack of exercise and obesity.
Higher education and income levels have been linked to better health in
individuals (12). For example, in an 8-year longitudinal study of a Chicago
neighborhood, Browning et al found that when income and
education were included in the health status model, health improved across time in relation
to reported education and income (12); Browning et al did not report a
temporal association between unemployment and poor health-related quality of
life. However, low income (i.e., less than $15,000 per year per household) was
associated with worse health-related quality of life for men and women aged 45
to 64 years (24). Employment status and activity limitation accounted for the
most variability in number of unhealthy days.
The results, which indicated that the interview language significantly
predicted fair/poor self-rated health, substantiate previous research studies.
Angel and Guarnaccia (19) found that level of acculturation, which was
measured by the interview language chosen by the participant, was independently correlated
with the respondent’s subjective assessment of health. One possible
explanation was that the adjectives used to describe normal health for Mexican
Americans and Puerto Ricans differed from those used by people who were not of
Hispanic origin. In addition, lower acculturation was associated with a
tendency to express distress somatically, which was evidenced by higher scores
on standard depressive affect scales in the study (19). The authors highlight
the importance of social and cultural influences on bodily perceptions, which
must be considered when comparing subjective health levels among various
social and cultural groups (19).
We found that the most powerful predictors of self-rated health are the
predictors that are potentially modifiable. Of respondents that were not obese, 86.7% reported being in good to excellent health; in contrast,
74.6% of participants in the obese category reported being in good to excellent health. Of the modifiable
predictors of self-rated health, weight may be the most realistically
changeable factor. Exercise is extremely
important for controlling BMI. Of the respondents who reported being
in excellent health, the highest percentage exercised regularly. Of those
reporting poor health, the highest percentage did not exercise regularly. Many
health care providers are highly respected by their patients — individuals
who may be at risk for developing lifestyle-related chronic diseases.
Health care providers should seize the opportunity to address their patients’ weight issues and sedentary
lifestyles. They should stress the need for exercise and weight control to
increase quality of life.
In addition, the importance of an education — at the very least, a high
school education — should be emphasized to adolescents and their parents as
vitally important to their future health. Culturally sensitive
approaches to health care services and delivery also must be considered when caring
for individuals of various ethnic backgrounds, because as our findings
suggest, health perceptions are influenced not only by medical factors but
also by sociocultural factors.
Back to top
Acknowledgments
The authors thank Shirley Laffrey, PhD, MPH, APRN, BC, and Elizabeth Abel,
PhD, RN, CS, ANP, for their expert comments on previous drafts of this article.
Back to top
Author Information
Corresponding Author: Lorraine J. Phillips, MSN, RN, FNP, The University of Texas at
Austin School of Nursing, 1700 Red River, Austin, TX 78701-1499.
Telephone: 512-248-8641. E-mail: lorrainephillips@yahoo.com.
Author Affiliations: Renee L. Hammock, RN, MSN, FNP-C, The University of Texas at Austin
School of Nursing, Austin, Tex; Jimmy M. Blanton, MPAff, Behavioral Risk Factor
Surveillance System Coordinator,
Epidemiologist, Texas Department of State Health Services, Austin, Tex.
Back to top
References
- U.S. Department of Health and Human Services. Healthy people 2010.
2nd ed. Washington (DC): U.S. Government Printing Office; 2000.
- Okosun IS, Choi S, Matamoros T, Dever GE.
Obesity is associated with
reduced self-rated general health status: evidence from a representative
sample of white, black, and Hispanic Americans. Prev Med 2001;32(5):429-36.
- Mossey JM, Shapiro E.
Self-rated health: a predictor of mortality among
the elderly. Am J Public Health 1982;72:800-8.
- Ayanian JZ, Weissman JS, Schneider EC, Ginsburg JA, Zaslavsky AM.
Unmet
health needs of uninsured adults in the United States. JAMA 2000;284(16):2061-9.
-
Centers for Disease Control and Prevention.
Self-assessed health status and selected behavioral risk factors among persons
with and without health-care coverage — United States, 1994-1995. MMWR
1998;47(9):176-80.
- Hsia J, Kemper E, Sofaer S, Bowen D, Kiefe CI, Zapka J, et al.
Is
insurance a more important determinant of healthcare access than perceived
health? Evidence from the Women’s Health Initiative.
J Womens Health Gend Based Med 2000;9(8):881-9.
- Franks P, Gold MR, Fiscella K.
Sociodemographics, self-rated health, and
mortality in the US. Soc Sci Med 2003;56:2505-14.
- Piko B.
Health-related predictors of self-perceived health in a student
population: the importance of physical activity. J Community Health
2000;25(2):125-37.
- Andersen M, Lobel M. Predictors of health self-appraisal: what’s
involved in feeling healthy? Basic & Applied Social Psychology
1995;16(1/2):121-36.
- Friis RH, Nomura WL, Ma CX, Swan JH. Socioepidemiologic and
health-related correlates of walking for exercise among the elderly: results
from the longitudinal study of aging. J Aging Physical Activity
2003;11(1):54-65.
- Okano G, Miyake H, Mori M.
Leisure time physical activity as a
determinant of self-perceived health and fitness in middle-aged male
employees. J Occup Health 2003;45(5):286-92.
- Browning C, Cagney K, Wen M.
Explaining variation in health status
across space and time: implications for racial and ethnic disparities in
self-rated health. Soc Sci Med 2003;57:1221-35.
- Miilunpalo S, Vuori I, Oja P, Pasanen M, Urponen H.
Self-rated health
status as a health measure: the predictive value of self-reported health
status on the use of physician services and on mortality in the working-age
population. J Clin Epidemiol 1997;50(5):517-28.
- Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH.
The
disease burden associated with overweight and obesity. JAMA
1999;282(16):1523-9.
- Idler EL, Benyamini Y.
Self-rated health and mortality: a review of
twenty-seven community studies. J Health Soc Behav 1997;38:2137.
- Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System Prevalence
Data, 2003. Atlanta (GA): Centers for Disease Control and Prevention; 2003.
- Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System Prevalence
Data, 2002. Atlanta (GA): Centers for Disease Control and Prevention; 2002.
- U.S. Census Bureau. State & County QuickFacts. Washington (DC): U.S.
Department of Commerce, U.S. Census Bureau; 2001.
- Angel R, Guarnaccia P.
Mind, body, and culture:
somatization among
Hispanics. Soc Sci Med 1989;28(12):1229-38.
- Finkelstein EA, Fiebelkorn IC, Wang G.
State-level estimates of annual
medical expenditures attributable to obesity. Obes Res 2004;12(1):18-24.
- Field A, Coakley EH, Must A, Spadano J, Laird N, Dietz WH, et al.
Impact of overweight on the risk of developing common chronic diseases during
a 10-year period. Arch Intern Med 2001;161(13):1581-6.
- Brown D, Balluz L, Heath G, Moriarty D, Ford E, Giles W, et al.
Associations between recommended levels of physical activity and
health-related quality of life. Findings from the 2001 Behavioral Risk Factor
Surveillance System (BRFSS) survey. Prev Med 2003;37(5):520-8.
- Mokdad AH, Marks JS, Stroup DF, Gerberding JL.
Actual causes of death
in the United States, 2000. JAMA 2004;291(10):1238-45.
- Centers for Disease Control and Prevention.
Public health and
aging: health-related quality of life among low-income persons aged 45-64
years-United States, 1995-2001. MMWR 2003;52(46):1120-4.
Back to top
|
|
