| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Jason T. Newsom, PhD, Institute on Aging, Portland State University, PO Box 751, Portland, OR 97207-0751. E-mail: newsomj{at}pdx.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Health behaviors • Preventive care • Public health
Broadly defined, health behaviors involve any action by an individual that has potential consequences for physical or psychological functioning (Leventhal, Rabin, Leventhal, & Burns, 2001). Health behaviors typically include lifestyle behaviors, such as smoking or exercise, as well as preventive care, such as regular checkups and screening tests (e.g., the Papanicolaou, or Pap, test), known to have serious health consequences. Other health-related choices such as eye exams, sunscreen use, and flu shots receive less attention, but they often have potentially serious health consequences. For example, regular eye exams can help prevent serious health events such as injuries from falls (Felson et al., 1989), sunscreen is considered an important primary preventive action against skin cancer (Feightner, 1994), and annual flu shots can forestall mortality (Liddle & Jennings, 2001).
Despite the importance of health behaviors and preventive care among older adults, recent information about the prevalence of healthy and unhealthy behaviors, preventive care, and attitudes about behavior change is needed for a variety of reasons. Priorities for public health goals are derived in part from prevalence estimates of health behaviors, preventive visits, and screening compliance. Without accurate information about the frequency of various health practices among older adults, it is difficult to decide which practices are the most important to target. Quality information about health-behavior prevalences is currently lacking. As Gochman has argued, "convenience sampling, although less appropriate, is more often the norm in health-behavior research than is population sampling" (1997, p. 393). This is especially true for the older adult population. Descriptive data on the individuals most likely to engage in unhealthy behaviors also can be extremely useful but are often not emphasized. Examining the demographic characteristics of older adults who are most likely to engage in unhealthy behaviors or neglect preventive visits and screening tests can aid in targeting groups who are most at risk for later health conditions.
A better understanding of attitudes about health improvement is needed, but subjective beliefs are infrequently studied in large, population-based surveys. Data regarding health priorities, motivation to change, and perceived barriers are crucial for targeting older adults for interventions. Prior behavior, behavioral intentions, and perceived barriers are at the heart of a number of the most studied health-behavior models (e.g., Leventhal, Zimmerman, & Gutmann, 1984; Rosenstock, 1974; Rosenstock, Strecher, & Becker, 1988), yet there is little information about how often individuals in the general population attempt to improve their health through healthier behavior or through motivations for changing their behavior. Further knowledge about the areas most in need of change will aid in the development of new interventions and the improvement of existing interventions designed to modify health behaviors. To date, many of these interventions have met with only moderate success (e.g., Muehrer, 2000; Pinto, Eakin, & Maruyama, 2000). Whereas a number of large health studies typically provide information on preventive-care and health behaviors, few have information on attempts at behavior change, motivations to improve health, or perceived barriers to changing health (e.g., Adler, 1994; Benson & Marano, 1998; Gentry et al., 1985; National Center for Health Statistics, 1987; Powell-Griner, Anderson, & Murphy, 1997). For instance, little is known about how many older adults believe they should change their behavior to improve their health. Financial barriers to preventive care frequently have been studied, but few large health surveys examine perceived motivational, health, or social barriers to health improvement.
In our present investigation we seek to provide population estimates of health behaviors, preventive care, self-reported behavior change, motivations to change, and perceived barriers to change among older adults in Canada. Although our focus is on a Canadian population, we expect that our results will be useful to researchers, policy makers, and practitioners in the United States and elsewhere. For older adults, the Canadian and U.S. health-coverage systems are similar in many respects. All Canadians receive the same basic health-care coverage, regardless of age or socioeconomic status, under a public insurance plan called Medicare. The Canadian federal government requires that provincial health insurance plans cover all medically necessary physician and hospital services to qualify for full cash transfers. Canadian provinces are responsible for managing Medicare within their jurisdictions. The federal government, in contrast, develops policies, enforces health regulations, promotes disease prevention, enforces health regulations, and provides health services to native communities. Although there are some important variations from province to province in amounts paid to physicians, cost controls, and prescription-drug coverage, the provincial programs, in general, appear to be highly comparable (Lipset, 1990). Perhaps the most significant difference between the Canadian health-care system and U.S. Medicare for older adults is that the Canadian system covers some prescription medications, although exclusions and copayments vary across provinces.
Data for this investigation are from the Canadian National Population Health Survey (NPHS). The NPHS is a large, general-health survey designed to study the health of the Canadian population with the goals of aiding the development of health-related policy, studying health correlates, and increasing the understanding of the relation between health-care utilization and health status (Tambay & Catlin, 1995). The NPHS has several important strengths, including its size, recency, and breadth, that will add substantially to our knowledge about health behaviors and attitudes toward change among older adults. Part of an ongoing longitudinal study that began in 1994–1995, the NPHS interviews respondents at 2-year intervals. The 1996–1997 NPHS survey, on which we focus in this report, includes a large, supplemental, random sample, including over 17,000 adults aged 60 years and older, and an extensive collection of questions on health behaviors, preventive care, and attitudes toward health. In this investigation, we focus on healthy and unhealthy behaviors (physical activity, alcohol use, use of a designated driver, smoking, and sunscreen use), preventive health care (blood pressure checks, dental visits, eye exams, flu shots, physical examinations, breast exams, mammograms, and Pap tests), and attitudes about behavior change (areas of health improvement, dietary changes, perceived need for improvements, and intentions to improve health).
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Measures
The NPHS data set provides an extensive list of questions regarding health behavior, preventive-care visits, reports of behavior change, beliefs about change, and intentions to change. Our goal is to present detailed categorical information about each item in order to provide an enduring reference for policy makers and researchers.
Demographics
Gender, age, and education were examined as correlates of health-behavior variables. For descriptive purposes, age was classified into two groups, 60 to 74 years and 75 and older. Education was divided into those with a high school diploma (or equivalent) or less and those with higher education. Tests of age and education differences, however, were based on continuous versions of these variables.
Health behaviors
As a way to measure physical activity, respondents were asked if they participated in any of a list of 20 different physical activities over the past month, including walking for exercise, gardening, swimming, and bicycling. The frequency of participation in the activity and the duration were then recorded. An individual was classified as "sedentary" if he or she did not participate in 1 or more physical activities of 15-min duration per month, "infrequent" if he or she participated in any activity for 15 min or more from 1 to 11 times per month, and "frequent" if participation for 15 min or more was 12 or more times per month. This classification was used in the 1990 Canadian Health Promotion Survey (O'Brien Cousins, 1998). Alcohol consumption was derived from two questions. The first question concerned regular drinking patterns on a monthly basis over the previous 12 months. Those indicating no drinks in the previous 12 months were classified as abstainers. Respondents were also asked about the number of drinks consumed on each day of the previous week (defined as one beer, glass of wine, or liquor drink). Individuals reporting a few drinks over the past year or an average of less than one drink per day in the past week were considered infrequent drinkers. Those consuming an average of one to two drinks per day were considered moderate drinkers, and those consuming more than three drinks per day were considered heavy drinkers. To assess whether a designated driver had been used while drinking, respondents were first asked, "Do you ever go out with friends or family to a place where you will be consuming alcohol?" Those indicating they had gone out to consume alcohol were then asked the question, "When you go out with friends, do you arrange to have a designated driver?" (yes–no). Questions concerning cigarette smoking were used to determine whether the individual currently smoked (defined as occasional or daily) or not (nonsmoker or former smoker). Sunscreen use for protection against skin cancer was reported in response to this question: "Now I would like to know about your use of precautions against exposure to the sun during the months of June, July, and August. How often do you use sunscreen?" Responses to this item were made according to a 5-point scale ranging from "never" to "always."
Preventive care
A number of questions concerned preventive visits, self-exam, or screening tests. For the most recent blood-pressure check, dental visits, and eye exams to be determined, respondents reported the time of the last exam. Blood-pressure and dental-visit responses were based on seven categories from "less than a year ago" to "never." Eye exams were based on five categories ranging from "less than a year ago" to "3 or more years ago." The most recent flu shot was reported as "less than 1 year ago," "1 to 2 years ago," "2 years ago or more," or "never." To measure physical examinations, respondents were first asked whether they had ever had a checkup when they did not have a health problem. Those reporting never having a checkup without having a health problem were asked if they had ever had an exam because they had a health problem. For those who previously had a routine examination, they were asked how recently. The seven response categories ranged from less than 1 year ago to 5 years or more. Several questions concerned women's health issues. Respondents were asked about their most recent breast exam by a physician, that is, "Other than a mammogram, have you ever had your breasts examined for lumps (tumors or cysts) by a doctor or health professional?," most recent mammogram (or "breast x-ray"), and most recent Pap test, all based on five categories from "less than 6 months ago" to "never."
Reported change, beliefs, intentions, and barriers
A series of questions concerned reports of behavior change to improve health, the respondents' beliefs about whether they should attempt to improve their health, their intentions for changing their health or behavior in the coming year, and perceived barriers to improving their health. Questions took this form: "In the past 12 months, did you do anything to improve your health? For example, lost weight, quit smoking, increased exercise?" Respondents were then asked, "What was the single most important change you made?" Responses were coded into 8 categories (increased exercise, lost weight, changed diet or eating habits, quit smoking, drank less, received medical treatment, took vitamins, or other). Participants were also asked, "What is the most important thing you are doing to follow a healthy diet?" Open-ended responses were coded into 11 categories. To gauge beliefs about whether respondents felt they needed to change their health, they were then asked, "Do you think there is anything you should do to improve your physical health?" Responses were categorized into the following categories: increase exercise, lose weight, improve eating habits, quit smoking, or take vitamins. Behavioral intention was assessed for those indicating they felt they should change their health by this question: "Is there anything you intend to do to improve your physical health in the next year?" Those responding "yes" were asked to specify the type of improvement. Responses were coded into the following categories: start–increase exercise, lose weight, improve eating habits, quit smoking, reduce amount smoked, learn to manage stress, reduce stress level, take vitamins, or other. For those individuals who reported that they should do something to improve their health, they were asked about barriers they perceived: "Is there anything stopping you from making this improvement?" These responses were then classified into the following 8 categories: lack of will power, time, too tired, too difficult, too costly, too stressed, disability–health problem, or other.
Analysis Overview
To examine prevalences of health behaviors, preventive care, reported change, beliefs, and intentions, we generated percentage estimates and significance tests by using SUDAAN release 7.5.4 (Research Triangle Institute, 2001), which adjusts estimates and their standard errors for the sampling design, nonresponse, and poststratification. All percentages reported in the tables and text have been weighted for sampling design. The frequencies (n), however, are unweighted. We grouped weighted percentages by category for gender, age (60–74 or >75 years), and education (
12 years or >12 years) for descriptive purposes. To test for differences in prevalences by gender, age, and education, we conducted a separate ordinary least squares (OLS) regression or logistic regression for each variable based on continuous variables, where possible, in order to maximize power and minimize Type I error (MacCallum, Zhang, Preacher, & Rucker, 2002). We tested gender, age, and education differences for nominal variables with more than two categories by using a chi-square analysis. For each variable, we indicate the type of significance test by superscript in the tables. To control for familywise error caused by a large number of tests, we evaluated significance by using a Bonferroni correction to retain a familywise error rate of.05 (
=.009 per test).
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
2(6) = 23.48, p <.009, more likely to report an improvement in diet, 2(10) = 54.51, p <.001, more likely to indicate that they should do something to improve their health, 2(8) = 35.01, p <.005, and more likely to intend to do something to improve their health in the next year (OR = 1.31, p <.005). For those respondents intending to improve their health, women were more likely to indicate they would increase exercise (OR = 1.183, p <.005) and lose weight (OR = 1.436, p <.001) than men. Women were more likely to mention health or disability as a barrier to improving health (OR = 1.514, p <.005).
Younger respondents were more likely to be physically active (b = –.142, SE =.02, p <.001), consumed more moderate amounts of alcohol (b = –.111, SE =.013, p <.001), were more likely to smoke (OR =.765, p <.001), were more likely to use designated drivers (OR =.839, p <.001), reported more frequent use of sunscreen (b =.108, SE =.028, p <.001), and more frequently engaged in some preventive-care practices, such as more recent dental visits (b =.224, SE =.017, p <.001) and breast self-exams (b = –.042, SE =.015, p <.001). Younger respondents were also more likely to report improvements in diet, 2(10) = 35.03, p <.005, and that they should do something to improve their health, 2(6) = 25.242, p <.005. Older respondents, however, were more likely to have recent blood-pressure checks (b = –.039, SE =.01, p <.005), eye exams (b = –.113, SE =.024, p <.001), or flu shots (b = –.086, SE =.003, p <.001), and they were less likely to intend to improve their health (OR =.840, p <.001). Intentions to improve health through exercise (OR =.933, p <.005) and weight loss (OR =.735, p <.001) were more likely to be mentioned by younger respondents. Younger respondents were also more likely to cite will power (OR =.770, p <.009), lack of time (OR =.758, p <.005), and cost (OR =.821, p <.009) as barriers to improving health. Older respondents, however, were more likely to report health or disability as a barrier (OR = 1.286, p <.001).
Consistent across most health behaviors, those with higher education levels had healthier lifestyles. Those with higher education levels were more likely to be physically active (b = –.052, SE =.009, p <.001), to use alcohol (b = –.015, SE =.004, p <.001), to be nonsmokers (OR = 1.11, p <.001), to use sunscreen (b = –.210, SE =.030, p <.001), to have recent dental exams (b = –.621, SE =.096, p <.001), to have recent eye exams (b = –.036, SE =.009, p <.001), to have a routine exam (b = –.475, SE =.115, p <.001), to have physician breast exams (b = –.180, SE =.030, p <.001), to have recent mammograms (b = –.185, SE =.022, p <.001), to have recent Pap tests (b = –.212, SE =.016, p <.001), and to conduct recent breast self-exams (b = –.084, SE =.030, p <.005). Those with higher levels of education were also more likely to report attempting to improve their health in the previous year, 2(8) = 35.87, p <.005, and that they should do something in the future to improve their health, 2(6) = 48.23, p <.001. Respondents intending to improve their health through exercise (OR =.864, p <.001) or taking vitamins (OR =.709, p <.005) were also more educated. Those with more education were more likely to indicate that lack of time was a barrier to improving health (OR =.780, p <.001).
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
One key advantage of the current study is the assessment of reported behavioral change, attitudes toward change, and perceived barriers to change. The results indicate that most older adults have not made attempts in the previous year to improve their health. This finding is noteworthy because most health studies, including the current report, indicate substantial room for improvement among older adults in areas of diet, physical activity, weight control, and preventive care (e.g., Ahmed, 1992; Simonsick et al., 1993; Wilson & Kannel, 2002). Significantly, efforts to maintain a good diet were described by respondents as eating "balanced meals" rather than as specific efforts to lower fat intake or increase fruits, vegetables, and grains. This finding suggests that further efforts are likely to be needed in dietary health education, given studies that indicate substantial dietary deficiencies in these areas (Ahmed, 1992).
In addition to relatively few past efforts to improve health, most older adults felt they did not need to make efforts to improve their health. Additional public health efforts may be needed to better inform older adults about the potential benefits of leading healthier lifestyles for increasing both longevity and quality of life. It is also important to note that, among those who believed improvements were warranted and who intended to make changes, efforts seemed to be focused disproportionately on exercise. Although physical activity has a number of primary and secondary benefits, most health recommendations also emphasize a variety of other highly beneficial actions, such as a healthy diet, smoking cessation, multivitamin supplements, weight control, and preventive tests.
The present study is one of only a few that investigate perceived psychological and social barriers to behavior change, factors rarely examined in population-based samples. Almost half of the respondents who were asked about barriers indicated "lack of will power" as a barrier to change. This outcome suggests a need for interventions that target motivation to improve health, an important component of recent models of health behavior (Leventhal et al., 2001), or self-efficacy, another potentially important factor in the success of behavioral change (Bandura, 1992, 2000). A possible approach for interventions is to encourage health-care providers to use strategies for enhancing older patients' motivation. One such approach, motivational interviewing, has been shown to be successful in facilitating behavior change (see, e.g., Miller & Rollnick, 2002). Motivational interviewing techniques allow health-care providers (a) to locate their patients on the stages of change, and (b) to facilitate movement of patients from precontemplation to contemplation and then to the actual processes involved in change itself. This methodology has been applied in numerous medical and public health settings (Resnicow et al., 2002) and could well be used with older adults. Although it is important to stress that interventions and behavior modification can be helpful and should be considered for older adults (Morely & Flaherty, 2002), it is also likely that intervening at earlier ages will be more effective in improving health outcomes and increasing longevity in the long run (Vaillant, 2002).
Our analyses of older smokers suggest some urgent areas for public health improvement. Less than 3% of smokers reported attempting to modify smoking habits. Moreover, only approximately one fifth of smokers expressed a belief that they should quit or reduce smoking. Even among those who believed they should modify their tobacco use, only one quarter reported any intention to quit in the coming year. It is important to recall the considerable evidence supporting the value of smoking cessation for older adults (Khaw, 1997).
The findings indicated a number of gender, age, and educational differences. Female respondents had a somewhat greater tendency to have healthier lifestyles, although male respondents reported more physical activity and were more likely to consume alcohol in moderate amounts. Female respondents were less likely to drink heavily, were more likely to use designated drivers, used sunscreen more often, had more recent eye exams, reported more attempts to improve their health, and were more likely to feel they should improve their health. Younger adults appear to be much better at adhering to screening recommendations, but they also were more likely to be heavy drinkers and to smoke. Education was perhaps the most consistent factor associated with healthier behaviors, including more regular physical activity, less heavy drinking, less smoking, and more recent preventive visits and tests. The relationship of education to health behaviors appears to be independent of a variety of other factors, at least in the case of physical activity (e.g., Kaplan, Newsom, McFarland, & Lu, 2001). At present, it is not fully understood why education plays a role in health behaviors. It is possible that those with more education have greater access to health information (e.g., through Internet use) or have greater ability to evaluate the risks and benefits to health when making lifestyle choices (Preston & Taubman, 1994). The association of education with health behaviors and preventive visits suggests that special public-education efforts may be needed for better informing those with lower education levels of the importance of exercise, smoking cessation, regular physician visits, and screening tests. It is curious that educational differences are still seen for preventive-care practices even when universal health care is available. Physicians may need to more aggressively communicate recommendations and implement screening tests for patients with lower educational or socioeconomic status.
There are several limitations of the present study that require mention. First, as with any large health survey that includes a wide range of questions, detailed information is often not available. For instance, although the NPHS distinguishes between routine physicals and those that occur in relation to a health problem, it is not possible to know the range of tests performed during exams for routine or health-related physician visits. In a similar manner, respondents were asked about dietary changes, but specific changes, such as lowering salt intake or reducing caloric intake, were not thoroughly investigated. Detailed information about the type of alcohol consumed also was lacking, and this may be an important factor in determining what constitutes healthy drinking patterns. A number of studies have suggested that cardiovascular benefits may be greater for moderate consumption of certain types of alcohol such as red wine (Landrault et al., 2001; Puddey, Croft, Abdu-Amsha Caccetta, & Beilin, 1998; van der Gaag, Ubbink, Sillanaukee, Nikkari, & Hendriks, 2000). Thus, it would be useful for future surveys to obtain more specific information on the quantities of each type of alcohol consumed.
A second important limitation of the current study is that all information is derived from self-report, which may overestimate or underestimate some health behaviors. Reports of behavior change, for example, are likely to be overestimates of actual changes in behavior. It is remarkable that, despite this potential bias, a high proportion of respondents do not report any behavior change. Third, for a number of variables we examined, it is impossible to discern whether participants sought care for preventive or routine purposes or they sought care because of a need to treat current problems. For example, older respondents were more likely to get eye exams, probably because of a higher likelihood of visual impairment or prescription changes for eye glasses. Nevertheless, such actions often have preventive effects such as averting falls or other accidents (Felson et al., 1989). Fourth, this study only provides general information about correlates of health behaviors. Other investigations have pursued more in-depth analyses of correlates of specific health behaviors (e.g., Kaplan et al., 2001; Kaplan, Newsom, & McFarland, 2002).
The results presented here are based on a sample of older Canadians and there are several considerations when one is attempting to generalize to the U.S. population. Although older Canadians receive coverage for many prescription drugs, access to health care for preventive visits and tests is likely to be similar in the two populations for those aged 65 and older. In addition, a number of health and health-behavior characteristics seem to be remarkably comparable in the two populations. We compared adults over the age of 65 in the NPHS and in the 1996 Behavioral Risk Factor Surveillance System (BRFSS; Division of Adult and Community Health, 1996) and found them to be similar on several key variables. For example, self-rated health was highly comparable in the two samples (12% rated their health as excellent in both samples, and 6% in the NPHS vs. 8% in the BRFSS rated their health as poor). Smoking rates (15% in the NPHS and 11% in the BRFSS) and alcohol use in the prior month (36% in the NPHS and 35% in the BRFSS) were also similar. In contrast, lifelong universal health care may have important effects on either health or patterns of utilization. Canadians have a longer life expectancy than Americans (United Nations Statistics Division, 2002), and one possible reason may be the cumulative effects of universal health-care access. On the basis of the results presented here, however, we see that the large number of preventive-care practices that clearly require additional improvement demonstrates that public health efforts must go well beyond increasing access to preventive care and will have to extend to educational and motivational efforts.
The findings from this investigation provide important information on the prevalence of health behaviors, preventive practices, behavior change, and motivations from a large population-based sample of older adults. The results point the way toward several possible avenues of health education and intervention. Of perhaps the greatest concern are results suggesting that surprisingly few older adults attempt to improve their health, the majority do not believe they should change their health behaviors, and a large proportion believe they lack the will power to modify their behavior. These findings suggest that the most important areas on which to focus education and intervention efforts include convincing older adults that they can change their behavior, that doing so has important effects on quality of life and longevity, and, for many older adults, that healthier behaviors are greatly needed.
|
Footnotes |
|---|
The analyses presented here are based on Statistics Canada's National Population Health Survey 1996–1997, Household Component, Public Use Microdata Files, which contain anonymized data. All computations on these data were prepared by Portland State University, and the responsibility for the use and interpretation of these data is entirely that of the authors.
1 Institute on Aging, Portland State University, Oregon.
2 Department of Psychiatry, Oregon Health and Science University, Portland.
Decision Editor: Linda S. Noelker, PhD
Received for publication February 11, 2003. Accepted for publication June 27, 2003.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K.-L. Chou The Prevalence and Clustering of Four Major Lifestyle Risk Factors in Hong Kong Chinese Older Adults J Aging Health, October 1, 2008; 20(7): 788 - 803. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
