Health Education Research, Vol. 16, No. 1, 21-31,
February 2001
© 2001 Oxford University Press
Gender and race differences in the predictors of daily health practices among older adults
Department of Health Policy, Management and Behavior, School of Public Health, University at Albany, State University of New York, One University Place, Rensselaer, NY 12144-3456, USA
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Preventive health behaviors are crucial for older adults' well-being. This study examined the factors that influence the practice of positive daily health behaviors over time in a sample of older adults (N = 1266) and investigated whether explanatory factors differ by health behavior, gender or race. Physical activity, weight maintenance, smoking, alcohol consumption and sleep patterns were examined as dependent variables. Independent variables included demographic characteristics, baseline health behavior, health status variables, psychological factors and social network characteristics. Results indicate that age and health status are important predictors of preventive health behaviors. However, the factors that predict preventive health behaviors vary by behavior, gender and race. The independent variables included in this study were most successful in explaining cigarette smoking and weight maintenance, and least successful in explaining amount of sleep. In addition, results suggest that social network variables are particularly influential for women's health behaviors, while health status is more influential among men. Greater education predicts better health behaviors among whites, while formal social integration seems particularly important for the health behaviors of older black women. These results indicate that examining older adults' health behaviors by race and gender leads to a fuller understanding of these behaviors.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Recently, increased attention has been paid to the importance of self-care to the health and well-being of older adults. One important component of self-care includes preventive health practices—those routine, day-to-day behaviors undertaken to promote health and prevent illness. Over the last 30 years, much evidence has accumulated linking these health practices with long-term health outcomes. Among the general population, both physical health status and mortality have been related to the following health behaviors: exercise and physical activity, sleep patterns, maintaining a regular meal schedule, proper nutrition, eating breakfast, cigarette smoking, alcohol consumption, and maintaining appropriate body weight (Belloc and Breslow, 1972
; Wingard et al., 1982; Berkman and Breslow, 1983; McIntosh et al., 1989). This relationship between health practices and health status exists among older adults as well. Breslow and Breslow (Breslow and Breslow, 1993) found that older adults with poor health practices experienced 50% greater disability and mortality over a 10-year period than those with a pattern of good health practices. Similarly, health promotion activities, such as exercise and good nutrition, have been related to older adults' functional health (Duffy and MacDonald, 1990), and impaired function in older women has been linked to obesity, smoking and physical inactivity (Ensrud et al., 1994).
Several studies have attempted to understand the predictors of good health practices among older adults. For example, Brown and McCreedy (Brown and McCreedy, 1986) examined 386 individuals aged 55 and older, and discovered that females tended to have better health behaviors than males. Among females, socioeconomic status was most predictive of health behaviors, while among males, marital status was most predictive. Dean (Dean, 1989), in a study of 465 people over 45, also found that being female predicted better health behaviors. In addition, social network and social support variables were influential in men's health behaviors. Among 2303 Medicare recipients, Potts et al. (Potts et al., 1992) found that women, individuals with stronger social support networks and those endorsing more health-promotive beliefs engaged in more positive health behaviors. Finally, Rakowski et al. (Rakowski et al., 1987) reported that gender and a supportive family environment consistently predicted good health practices.
In addition, similar studies have examined the predictors of preventive health behaviors among middle-aged and younger adults (Langlie, 1977; Hibbard, 1988; Antonucci et al., 1990; Palank, 1991; Stoller and Pollow, 1994). Taken together, these studies suggest that greater education, stronger internal health locus of control, being married, being white, higher socioeconomic status, stronger social support networks and female gender are all predictive of better preventive health behaviors.
Unfortunately, although these studies have been informative, most explain only a small amount of the variance in health behaviors and thus we still know relatively little about what factors influence the performance of these behaviors among older adults. This low predictive power may be due, at least in part, to two factors. First, health habits are usually combined into one summary dependent variable. Although this is intuitively appealing, especially since these health behaviors appear to have an additive effect on health outcomes (Belloc and Breslow, 1972), in reality these behaviors generally do not correlate very highly with one another (Rakowski et al., 1987). Therefore, it is likely that they may be influenced by different predictive factors.
Second, little attention has been paid to the possibility of gender and race differences in the predictors of these behaviors. The prevalence of these behaviors has been shown to differ by gender and race, so it is possible that the predictors of these behaviors may be different as well or that the predictors may differ in relative strength. A brief review of the literature concerning gender and race differences in health practices follows.
Several studies have documented better health behaviors among women (Brown and McCreedy, 1986; Rakowski et al., 1987; Dean, 1989; Antonucci et al., 1990; Potts et al., 1992; Stoller and Pollow, 1994). Studies that have examined predictors of health behaviors within genders have found that for women, higher socioeconomic status, older age, greater education, being married, experiencing fewer negative life events, having a more supportive social network and attending church are predictive of better health behaviors (Gottlieb and Green, 1984; Brown and McCreedy, 1986). Among men, higher education, being married, experiencing fewer negative life events, having a more supportive social network and attending church are predictive (Gottlieb and Green, 1984; Brown and McCreedy, 1986; Dean, 1989; Antonucci et al., 1990; Ungemack, 1994).
There is far less research pertaining to the influence of race on health practices, primarily because non-white individuals are not well represented in most study samples. In fact, the lack of knowledge about all types of self-care practices among minority elders has been identified as a serious gap in the self-care literature (Davis and Wykle, 1998). Among the available literature examining older adults' preventive self-care practices, being white is associated with greater preventive health behaviors (Kart and Engler, 1994) and with more positive assessments of capacity for self-care (Kart and Dunkle, 1989).
This study was conducted to examine the following research questions:
- What factors influence the practice of positive daily health behaviors in older adults over time?
- Does this differ according to health behavior?
- Do these explanatory factors differ by gender?
- Do these factors differ by race?
Physical activity, maintaining appropriate body weight, smoking, alcohol consumption and sleep patterns were examined as dependent variables. Independent variables included demographic characteristics, baseline health behavior, health status variables, psychological factors and social network characteristics.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sample
This study uses data from the Americans' Changing Lives (ACL) longitudinal panel survey (House, 1997
). The ACL survey was conducted in 1986 and 1989 with a multistage stratified area probability design. The sample included residents of all US households, age 25 years or older, and excluded residents of Alaska and Hawaii, military bases, group quarters, and institutions. In Wave 1, 3617 persons were interviewed in their homes, with a response rate of 68%. The response rate for Wave 2 was 79%, yielding a sample size of 2867. Older adults and black Americans were oversampled at twice the rate of those under age 60 and whites. The present study examined only that portion of the ACL sample that responded in both Wave 1 and Wave 2, and who were at least 60 years of age at Wave 1. The final sample included 1266 persons, divided into four groups: 112 black males, 284 white males, 251 black females and 619 white females. Mean age was 69 years for black males, 68 years for white males, 69 years for black females and 70 years for white females.
Measures
Independent variables included demographic characteristics, baseline health behavior, health status variables, psychological factors and social network variables. Five dependent variables were examined. These included physical activity, weight maintenance, smoking, alcohol consumption and sleep. All independent variables were measured at Wave 1, while all dependent variables were measured at Wave 2. The variables were measured as follows.
Demographics
Age, measured in years, was computed from the respondent's birth date. Education was determined by asking the participant to indicate the highest degree they obtained. Responses were categorized into one of the following six categories: 0–8, 9–11, 12, 13–15, 16 or 17 or more years. Income represents total household income for the previous 12 months, recoded into 10 categories (<$5K, $5–9K, $10–14K, $15–19K, $20–24K, $25–29K, $30–39K, $40–59K, $60–79K and $80K+). Marital status was dichotomously coded with 1 indicating married and 0 assigned to not married.
Health status
Self-rated physical health was assessed by asking respondents to characterize their health as poor, fair, good, very good or excellent. Respondents were also asked to indicate whether or not they had any of the following chronic health problems during the last 12 months: arthritis or rheumatism, lung disease, hypertension, heart attack, diabetes, cancer, foot problems, stroke, broken bones, or urine beyond control. These items were summed to create a measure of number of chronic conditions. A functional health index was created with items measuring the amount of difficulty the respondent has in bathing, climbing stairs, walking and doing heavy housework. The resulting index ranged from 1 = worst functional health (i.e. most severe functional impairment) to 4 = best functional health (i.e. no functional impairment).
Psychological factors
Six items were used to create a measure of mastery. These items were `I take a positive attitude toward myself', `I can do just about anything I really set my mind to do', `At times I think I am no good at all', `All in all, I am inclined to feel that I am a failure', `Sometimes I feel that I am being pushed around in life' and `There is really no way I can solve the problems I have'. Response choices for all six items ranged from strongly agree to strongly disagree on a four-point scale. Items were recoded so that higher scores reflected greater levels of competence. These items were summed to create a mastery measure with a possible range of 6–24. The 
reliability for this scale was 0.65.
Respondents were also asked to indicate whether or not they had experienced any of the following events in the past 3 years: death of a spouse, death of a child, death of a parent, death of a close friend or relative, divorce, assault, involuntary job loss, burglary and any other upsetting event. These nine events were dichotomously coded and summed to obtain a measure of stressful life events with a potential range of 0–9.
Depressive symptoms were measured with the Iowa form for the Center for Epidemiologic Studies Depression Scale (CES-D) (Radloff, 1977; Kohout et al., 1993). The Iowa form is an 11-item version of the CES-D that taps the same underlying dimensions as the original 20-item scale (depressed affect, positive affect, somatic complaints and interpersonal problems). Items ask how often respondents experienced each depressive symptom during the past week. Response choices include hardly ever or never, some of the time and much or most of the time. Items were summed to create a measure of depressive symptoms with a range of 0–22. Previous research indicates an reliability of 0.76 for this short version, which is comparable with the original scale ( = 0.80). The reliability in this sample was 0.80.
Social factors
Several indicators of social support were included in this study. The item `How often does someone remind you to do things which will help you stay healthy, such as getting enough sleep or exercise, or taking medications?' was used as an indicator of health behavior-specific support. The original responses of often, sometimes, rarely and never, were recoded into a new dichotomous variable in which 1 = often or sometimes and 0 = rarely or never.
Informal social integration was measured with two items which asked how often the respondent (1) talks on the telephone with friends, neighbors or relatives and (2) gets together with friends, neighbors or relatives. Response categories included more than once a day, once a day, 2–3 times a week, about once a week, less than once a week and never. Responses to these two items were summed. Similarly, formal social integration was measured with two items that asked how often the respondent (1) attended meetings or programs of groups, clubs or organizations that they belong to and (2) attended religious services.
General social support was indicated with a summary measure combining measures of support received from a spouse, children, and friends or relatives. For each source of support, respondents were asked `How much does your _______ make you feel loved and cared for?' and `How much is ______ willing to listen when you need to talk about your worries or problems?'. Response choices for these items ranged from a great deal to not at all on a five-point scale. These six items were summed to reflect general social support. The reliability for this scale was 0.67.
Health behaviors
Five health behaviors were assessed at both Waves 1 and 2. Alcohol consumption was indicated by two items that, in combination, estimated the number of drinks in the last month. These items were `During the last month, on how many days did you drink beer, wine or liquor?' and `On days that you drink, how many cans of beer, glasses of wine or drinks of liquor do you usually have?'. A physical activity index was created by asking respondents how often they engage in the following activities: work in the garden or yard, participate in active sports or exercise and take walks. A four-point response scale ranged from often to never. If the respondent reported smoking cigarettes, they were asked, `On the average, how many cigarettes or packs do you usually smoke in a day (1 pack = 20 cigarettes)?'. Hours of sleep were assessed with the item, `How many hours of sleep do you usually get in a 24-h period, including naps?'. An ordinal variable was created with 1 assigned to less than 6 or more than 9 h of sleep each day, 2 assigned to 6 or 9 h of sleep each day, and 3 assigned to 7 or 8 h of sleep each day. Body mass index (BMI) was used to indicate weight maintenance. Respondents were classified into five gender-specific categories of BMI as follows: underweight (lowest 5% of cases), low normal (next lowest 25%), mid-normal (middle 30%), high normal (next to highest 25%) and overweight (highest 15%). These categories were then combined into a three-category ordinal scale which included underweight or overweight, low normal and high normal, and mid-normal. Corresponding Wave 1 health behavior variables were used as control variables in the analyses.
Data analyses
Hierarchical multiple linear regression analyses involving stepwise selection were used to predict the number of drinks last month, physical activity, number of cigarettes smoked each day, BMI and hours of sleep among the four race and sex groups separately. For all analyses, independent variables were entered in blocks, to assess the influence of each set of predictors over and above the influence of previous variables. Thus, respondent demographic characteristics (age, education, marital status and income) were entered first. Block 2 consisted of Wave 1 health behavior. Next, health status variables were entered (self-rated health, number of chronic conditions, and functional health), followed by the psychological factors (stress, depressive symptoms, mastery). Finally, the social support variables were entered. Within each block, stepwise selection was used to determine significant predictors. The 95% significance level was used as a cut-off for statistically significant results.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Table I
presents means and standard deviations for all study variables by population subgroup.
|
Results of the multivariate regression analyses are presented in Tables II–VI
. Differences in both the amount of variance explained and the significant predictors that emerged are evident across health behaviors and across population subgroups.
|
|
|
|
|
With respect to percent variance explained, cigarette smoking is the dependent variable best predicted by the set of independent variables used in these analyses, with the R2 ranging from 0.37 for black males to 0.63 for white females. For all groups, number of cigarettes smoked per day was best predicted by cigarette smoking at Wave 1, which accounted for the vast majority of variance explained. Also significant across all age groups was age, with older individuals smoking fewer cigarettes. Results indicate that health status influences smoking among males while among females, social network characteristics influence smoking behavior. White males with more chronic conditions smoke fewer cigarettes and black males with better self-rated health smoke more. White females who have someone who reminds them to take care of their health tend to smoke more, while black females who report greater amounts of informal social integration smoke fewer cigarettes.
BMI, used as an indicator of weight maintenance behavior, was also fairly well explained by the current set of predictors, with percent variance explained ranging from 39 to 51%. As Table III
illustrates, for males, only BMI at Wave 1 is predictive of BMI at Wave 2. For females, a few other factors emerge as predictors of BMI. White females who have greater education and who have fewer depressive symptoms are more likely to have a BMI in the ideal range. Older black females and those who have fewer chronic conditions are more likely to have a BMI in the ideal range.
The percent variance explained in physical activity ranged from 29 to 51% (see Table IV) and the greatest number of significant predictors emerged for this health behavior. For both black and white females, older age and lower income predicted less physical activity, while greater education predicted more physical activity for white males and females. Again, for all population subgroups, physical activity at Wave 1 was a strong influence on Wave 2 health behavior. Health status also emerged as a significant influence on physical activity, particularly for black males. Among both black males and white females, better functional health predicted greater physical activity; for black males this variable accounted for an additional 7% explained variance. White males and females who reported better self-rated health had greater levels of physical activity, as did black females who reported fewer chronic conditions. Finally, social network variables emerged as significant, but weak, predictors of physical activity for all groups except black males. White males who had someone to remind them to take care of their health were less likely to be physically active, while white females with greater levels of general social support and black females who reported greater formal social integration were more likely to be physically active.
Alcohol consumption was only moderately well explained by this model, with R2 values ranging from 0.21 to 0.30 (see Table V). As with all the other dependent variables, alcohol consumption at Wave 1 was the strongest independent variable across all population subgroups. Among whites, greater education predicted greater alcohol consumption, while among males, being married predicted lower alcohol consumption. In addition, older white females were less likely to consume alcohol. For black males, self-rated health influenced alcohol consumption, with those individuals reporting better health more likely to consume more alcohol. On the other hand, psychological factors emerged as weak, but significant predictors for both white males and females. White males with greater mastery scores and white females with more stressful life events had greater alcohol consumption. Among black females, formal social integration again emerged as a significant predictor of health behavior, with those individuals who were more socially integrated reporting less alcohol consumption.
Sleep was the health behavior that was least well predicted by the present set of independent variables, with percent variance explained ranging from 4 to 21% (see Table VI). Again, across all groups, hours of sleep at Wave 1 was the best predictor of sleep at Wave 2. Demographic characteristics predicted sleep among white males and females, with younger males and females with greater income reporting better sleep habits. Health status variables predicted hours of sleep for all subgroups except black males. White males with more chronic conditions were more likely to report less than ideal sleep habits, while for females, both black and white, better functional health predicted more ideal sleep. Psychological factors and social network characteristics did not emerge as significant predictors of hours of sleep for any of the subgroups.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The prediction of daily preventive health practices
As these results indicate, the predictive model tested in these analyses, which included demographics, baseline health behavior, health status variables, psychological factors and social network characteristics, was moderately successful in predicting the practice of daily preventive health behaviors. The overall success of the model, as well as which individual predictors attained significance, varied markedly by health behavior and by race/gender subgroup. This lends support to the notion that to more fully understand older adults' health practices, we need to examine them individually as well as within race and gender groups.
For each of the five health behaviors examined here, the corresponding health behavior at Wave 1 was, not surprisingly, the strongest predictor of behavior at follow-up, explaining from 2 to 56% of the variance at Wave 2. While this reflects a fair amount of stability in these behaviors over time, the large amount of variance left unexplained by baseline behavior in most of these analyses indicates a significant potential for change in these routine behaviors among the older adult population.
The strength of baseline health behavior in predicting follow-up behavior varied significantly across behaviors. Cigarette smoking and BMI had the most variance accounted for by baseline behavior, indicating that cigarette smoking and weight maintenance are relatively stable behaviors over time. This finding is not surprising in light of the fact that these behaviors are among the most difficult to change. Hours of sleep, on the other hand, was least well predicted by baseline sleep and was least well predicted overall.
Other than baseline health behavior, no other variables emerged as strong predictors of all five health practices, but looking across all analyses, some predictors emerged as more influential than others. Age attained significance in at least one group for every behavior examined, reflecting that older individuals exhibit better health practices. This supports findings found in some previous literature (Gottlieb and Green, 1984; Prohaska et al., 1985; Hibbard, 1988). The conclusions to be drawn from this finding are uncertain, however. It may be that as individuals age, they make positive changes in their health practices. Equally likely is a selection effect. That is, individuals with good health practices are more likely to survive to older ages and, thus, good health practices are better represented among older individuals. Health status variables also emerged as significant predictors, especially for physical activity and sleep. This makes intuitive sense, as many health problems common to older adults may directly interfere with one's ability to be physically active and to get a good night's sleep. However, inasmuch as being active and well-rested contribute to other positive health outcomes, these findings indicate that older adults with impaired functional health or with a greater number of chronic conditions, for example, represent an at-risk population with a special need for intervention.
Interestingly, physical activity emerged as the behavior with the most number of different significant predictors and sleep the least. This underscores the fact that physical activity is a complex behavior that is responsive to many influential factors, thus making it a challenge for intervention.
Differences by gender
Interesting gender differences emerged in these analyses. For two behaviors, cigarette smoking and weight maintenance, the predictive model employed in these analyses explained more variance for females than males. In addition, cigarette smoking was significantly influenced by social network characteristics for women only and by health status variables for men only. In fact, throughout these analyses, social network variables emerged as more important for women than men in general. This indicates that women may perform many of these health behaviors within a social context, and this has important implications for the nature and location of intervention efforts. Formal social integration demonstrated a special importance for black women, particularly for smoking, physical activity and alcohol consumption. Formal social settings, such as religious groups, clubs and other organizations, may be especially appropriate intervention settings for this segment of the population. Finally, although marital status has been previously shown to be influential in men's health behaviors (Brown and McCreedy, 1986), in this study being married was predictive only of less alcohol consumption for men.
Differences by race
Fewer conclusions can be drawn in terms of differences by race. Overall, the model explained more variation for whites than blacks, although this may be due to sample size effects. The model was particularly poor in explaining black males' smoking and weight maintenance behavior, indicating an area for further research. The only other race difference that emerged here was that greater education predicted more physical activity and less alcohol consumption among whites, as compared to blacks.
The present study has several limitations. The analyses reported here were limited by the constraints on the choice of independent and dependent variables that are inherent in secondary data analysis. The preventive health behaviors were all represented by single-item measures, when in reality multiple-item indicators may better reflect these complex behaviors. Preventive health behaviors may be better explained if the selection of predictors could be based on more theoretical considerations. The Health Belief Model and Social Cognitive Theory seem particularly well suited to serve as guides for the prediction of these behaviors. For example, it would be instructive to include measures of perceived benefits and barriers, self-efficacy and outcome expectations as predictors in future research. In addition, the 3-year time frame between waves of data collection in this study may not represent the most meaningful time lag for the investigation of influences on preventive health behaviors among older adults. Finally, it must be recognized that the data used in these analyses are a decade old, and the important influences on preventive practices may be different for current and future cohorts of older adults. Despite these limitations, however, the differences that emerged among the four race–gender subgroups provide an intriguing indication that the influential forces on preventive health behaviors may vary to a considerable degree for men and women, and for blacks and whites. At the very least, intervention efforts that target older adults' health behaviors may need to be tailored to the racial and gender makeup of the target population.
The overall low amount of variance explained in these analyses reaffirms the notion that we still have significant progress to make in terms of understanding older adults' preventive health behaviors. In general, efforts to understand these types of health behaviors have concentrated on the role of individual factors in determining behavior. Future research would do well to also take a broader focus that includes a look upstream for broader social and environmental influences that shape older adults' behavior by enhancing or restricting opportunities to behave in health promotive ways.
|
Acknowledgments |
|---|
We thank Michelle van Ryn for her helpful comments on an earlier draft of this manuscript. This research was supported by a Faculty Research Awards Program grant from the University at Albany to M. P. G.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Antonucci, T. C., Akiyama, H. and Adelmann, P. K. (1990) Health behaviors and social roles among mature men and women. Journal of Aging and Health, 2, 3–14.
Belloc, N. B. and Breslow, L. (1972) Relationship of physical health status and health practices. Preventive Medicine, 1, 409–421.[Medline]
Berkman, L. F. and Breslow, L. (1983) Health and Ways of Living: The Alameda County Study. Oxford University Press, New York.
Breslow, L. and Breslow, N. (1993) Health practices and disability: some evidence from Alameda County. Preventive Medicine, 22, 86–95.[Web of Science][Medline]
Brown, J. S. and McCreedy, M. (1986) The hale elderly: health behavior and its correlates. Research in Nursing and Health, 9, 317–329.
Davis, L. and Wykle, M. L. (1998) Self-care in minority and ethnic populations: the experience of older black Americans. In Ory, M. G. and DeFriese, G. H. (eds), Self-Care in Later Life. Springer, New York, pp. 170–179.
Dean, K. (1989) Self-care components of lifestyles: the importance of gender, attitudes and the social situation. Social Science and Medicine, 29, 137–152.
Duffy, M. E. and MacDonald, E. (1990) Determinants of functional health of older persons. The Gerontologist, 30, 503–509.[Abstract]
Ensrud, K. E., Nevitt, M. C., Yunis, C., Cauley, J. A., Seeley, D. G., Fox, K. M. and Cummings, S. R. (1994) Correlates of impaired function in older women. Journal of the American Geriatric Society, 42, 481–489.
Gottlieb, N. H. and Green, L. W. (1984) Life events, social network, life-style, and health: an analysis of the 1979 National Survey of Personal Health Practices and Consequences. Health Education Quarterly, 11, 91–105.[Web of Science][Medline]
Hibbard, J. H. (1988) Age, social ties and health behaviors: an exploratory study. Health Education Research, 3, 131–139.
House, J. S. (1997) Americans' Changing Lives: Waves I and II, 1986 and 1989 [Computer file]. ICPSR Version. University of Michigan, Survey Research Center, Ann Arbor, MI [Producer]. Inter-university Consortium for Political and Social Research, Ann Arbor, MI [Distributor].
Kart, C. S. and Dunkle, R. E. (1989) Assessing capacity for self-care among the aged. Journal of Aging and Health, 1, 430–450.
Kart, C. S. and Engler, C. A. (1994) Predisposition to self-health care: who does what for themselves and why? Journal of Gerontology, 49, S301–308.
Langlie, J. K. (1977) Social networks, health beliefs, and preventive health behavior. Journal of Health and Social Behavior, 18, 244–260.[Web of Science][Medline]
McIntosh, W. A., Shifflet, P. A. and Picou, J. S. (1989) Social support, stressful events, strain, dietary intake, and the elderly. Medical Care, 27, 140–153.[Web of Science][Medline]
Palank, C. L. (1991) Determinants of health-promotive behavior. Nursing Clinics of North America, 26, 815–832.
Potts, M. K., Hurwicz, M.-L., Goldstein, M. S. and Berkanovic, E. (1992) Social support, health-promotive beliefs, and preventive health behaviors among the elderly. Journal of Applied Gerontology, 11, 425–440.
Prohaska, T. R., Leventhal, E. A., Leventhal, H. and Keller, M. L. (1985) Health practices and illness cognition in young, middle aged, and elderly adults. Journal of Gerontology, 40, 569–578.
Rakowski, W., Julius, M., Hickey, T. and Halter, J. B. (1987) Correlates of preventive health behavior in late life. Research on Aging, 9, 331–355.[Abstract]
Stoller, E. P. and Pollow, R. (1994) Factors affecting the frequency of health enhancing behaviors by the elderly. Public Health Reports, 109, 377–387.[Web of Science][Medline]
Ungemack, J. A. (1994) Patterns of personal health practice: men and women in the United States. American Journal of Preventive Medicine, 10, 38–44.[Web of Science][Medline]
Wingard, D. L., Berkman, L. F. and Brand, R. J. (1982) A multivariate analysis of health-related practices: a nine-year mortality follow-up of the Alameda County Study. American Journal of Epidemiology, 116, 765–775.
Received on January 31, 2000; accepted on April 17, 2000
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. L. Sellers, V. Bonham, H. W. Neighbors, and J. W. Amell Effects of Racial Discrimination and Health Behaviors on Mental and Physical Health of Middle-Class African American Men Health Educ Behav, February 1, 2009; 36(1): 31 - 44. [Abstract] [PDF]
|
|
|
|
 |
|
 L. Levy-Storms and J. E. Lubben Network Composition and Health Behaviors Among Older Samoan Women J Aging Health, December 1, 2006; 18(6): 814 - 836. [Abstract] [PDF]
|
|
|
|
 |
|
 J. J. Peterson, J. B. Lowe, N. A. Peterson, and K. F. Janz The relationship between active living and health-related quality of life: income as a moderator Health Educ. Res., February 1, 2006; 21(1): 146 - 156. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Y. Kinney, L. E. Bloor, C. Martin, and R. S. Sandler Social Ties and Colorectal Cancer Screening among Blacks and Whites in North Carolina Cancer Epidemiol. Biomarkers Prev., January 1, 2005; 14(1): 182 - 189. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Thrasher, M. K. Campbell, and V. Oates Behavior-Specific Social Support for Healthy Behaviors Among African American Church Members: Applying Optimal Matching Theory Health Educ Behav, April 1, 2004; 31(2): 193 - 205. [Abstract] [PDF]
|
|
|
|
 |
|
 C. M. Baldwin, K. Long, K. Kroesen, A. J. Brooks, and I. R. Bell A Profile of Military Veterans in the Southwestern United States Who Use Complementary and Alternative Medicine: Implications for Integrated Care Arch Intern Med, August 12, 2002; 162(15): 1697 - 1704. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||