Dimensions of Health Maintenance Activities and Opinions

Jacob Hornik, University of Illinois Chicago Circle
Mary Jane Schlinger, University of Illinois Chicago Circle, University of Illinois Medical Center
ABSTRACT - The purposes of this study are to identify dimensions of health maintenance activities and opinions and to look at correlations between selected demographic characteristics and those dimensions. The study utilizes data from a national sample of 738 adult females and 683 males who filled out lengthy life style questionnaires. Factor analyses of items relevant to health maintenance produced 17 male and 16 female factors representing several broad health dimensions.
[ to cite ]:
Jacob Hornik and Mary Jane Schlinger (1980) ,"Dimensions of Health Maintenance Activities and Opinions", in NA - Advances in Consumer Research Volume 07, eds. Jerry C. Olson, Ann Abor, MI : Association for Consumer Research, Pages: 627-632.

Advances in Consumer Research Volume 7, 1980     Pages 627-632


Jacob Hornik, University of Illinois Chicago Circle

Mary Jane Schlinger, University of Illinois Chicago Circle, University of Illinois Medical Center

[The investigators acknowledge and appreciate the support of Leo Burnett Company advertising agency. The study also was partially supported by grant number 3-01-17-45-3-06 from the University of Illinois, Chicago Circle.]


The purposes of this study are to identify dimensions of health maintenance activities and opinions and to look at correlations between selected demographic characteristics and those dimensions. The study utilizes data from a national sample of 738 adult females and 683 males who filled out lengthy life style questionnaires. Factor analyses of items relevant to health maintenance produced 17 male and 16 female factors representing several broad health dimensions.

The United States spent $163 billion for health care in fiscal year 1977--an average of $737 for each man, woman and child (American Hospital Association, 1978). Per capita real expenditures for health have tripled since 1940, and life expectancy has increased nearly 15% (Keyfitz, 1978).

A significant limitation on further health progress in the U.S.A. is the detrimental influence of factors found in the physical and social environment. As acute communicable diseases have been brought under control, the significance of chronic degenerative diseases, such as cancer, stroke and heart disease has increased. Concomitantly, social and psychological processes are replacing infectious agents as the causes of illness, and changes in life style have become crucial to the improvement of overall health status in America (Stone, Cohen and Adler, 1979).

As a consequence, many government officials and health care professionals are urging that the traditional model of health care delivery be broadened to include disease prevention as well as disease cure. This increased emphasis on health maintenance via adoption of healthful life styles is the logical result of several convergent trends. First, as suggested above, the leading causes of premature death among Americans are the degenerative and "social diseases" that go hand in hand with affluence and the "good life": heart, cancer, cirrhosis of the liver, stroke, motor vehicle accidents, etc. (Preston, 1977). Second, new research is forcing increased recognition that there are multiple rather than single causes of many diseases, including both heart and cancer. Third, prominent among the risk factors that are associated with the major causes of death are life style characteristics: cigarette smoking, lack of exercise, high cholesterol diets, excessive alcohol consumption, obesity, and a hard driving personality (Cairns, 1975; Belloc, 1973). Fourth, a relatively new and significant phenomenon in the health care industry is consumerism; health consumers are taking increased responsibility for shaping programs and policies (Wortzel, 1976). For some people, increased responsibility for health care is characterized by active personal participation in their own health maintenance.


The research reported here has two purposes. First, to identify dimensions of health maintenance activities and opinions. And second, to look at the correlation between selected demographic characteristics and those health-related dimensions.

In a commonly accepted paradigm, Kasl and Cobb (1966) distinguishes three types of health related behavior: (1) health or what is here referred to as health maintenance behavior--activities undertaken by healthy individuals to prevent disease or detect disease in the asymptomatic state; (2) illness behavior--undertaken in the presence of symptoms for the purpose of diagnosis and discovering a remedy; and (3) sick role behavior--undertaken for the purpose of getting well.

A major problem in studying health maintenance is in defining what is "healthy" and "unhealthy." It is rare to find public or professional consensus about major health issues. People argue, for example, over the importance of regular exercise, the usefulness of annual check-ups, safety of vitamins and food additives, the effects of smoking, and even the value of immunizations. As marketing researchers, we do not presume to address these controversial issues. Therefore, the definition of "healthy" used in this research is those attitudes and activities that some people--not necessarily a majority of professionals or laypersons--claim to be healthy or unhealthy. Generally the attitudes and activities examined here have gained visibility in the mass media even though their relationship to health or illness may be unproven.


This study utilizes first wave data from a three wave study conducted by Leo Burnett Co. advertising agency as part of their ongoing marketing-psychographics research (Plummer, 1976).

The study questionnaires were distributed in-home during October and November 1978, by Opinion Research Corporation interviewers to a probability sample of the United States designed by ORC in conjunction with Marketmath Inc. In total 738 female and 683 male respondents mailed back completed and usable questionnaires, for a first wave response rate of 36%. Separate questionnaires were employed for the two sexes.

The male and female questionnaires each included over 350 attitude, interest and opinion items (AIO). From these the investigators subjectively selected for analysis 75 male and 82 female items which the literature suggested as related to health maintenance (Cooper, Kehoe and Murphy, 1978; Stone, Cohen and Adler, 1979).

These items were analyzed in the following procedures (Wind & Green, 1974):

1. First, correlations were computed for all pairs of the health life style statements and were factor analyzed by principle component solution.

2. Retained principle components were rotated by the Varimax procedure to produce orthogonal factors.

3. Based on the content of health items with absolute high loadings on each rotated factor, each factor was given a label, e.g., "heavy smoker," "diet conscious," etc.

4. Factors were related to four selected background variables--occupation, age, education and city size--in order to see how the demographics correlate with the factor dimensions.


Psychographic data analysis became a full-fledged activity in marketing research (Mehrota and Wells, 1977). Moreover, life style segmentation has become a fairly standard segmentation system (Bearden, Teel and Durand, 1978; Plummer, 1974). The basic premise of research on health life styles is that the more we know about people's health AIO's, the more effectively we can communicate with them and educate them for a healthier way of life.

The concept of life style patterns and its relationship to marketing and consumer analysis was introduced in 1963 by William Lazer. Lazer defined life style as "a systems concept. It refers to a distinctive mode of living an its aggregate and broadest sense--it embodies the patterns that develop and emerge from the dynamics of living in a society" (Lazer, 1963). What emerged has been studies called "life style," "psychographics" and "activity and attitude" research.

Life style, with its roots in depth or motivation research (Wells and Tigert, 1971), has given many consumer researchers a new construct for thinking about the consumer beyond simple demographic characteristics. Life style as a construct permits one to think about the consumer as a total person and how consumption patterns, health behavior and other activities fit into his or her daily living patterns.

The advantage of healthy-unhealthy attitude and behavior studies based on psychographics are manifold. Not only does a study such as this impose a structure on the researched items, but the groupings that emerge are meaningful and have implications for health policy. Each factor represents a different pattern of needs and health behavior.


Using an eigenvalue of one as the criterion, factor analysis of the male and female health items produced 24 and 26 factors respectively, accounting for 59.9 and 61.8 percent of the total variance. Of these, 17 male and 16 female factors were retained, based on the criteria of meaningfulness, relevance to health maintenance, and (with one exception) representation by two or more items with loadings of .40 or greater. These retained factors, which account for 39 and 38 percent of the variance, are shown in Tables 1 and 2.

The large number of factor dimensions derived from the data suggests the possibility that there is a good deal of fragmentation in health maintenance activities and opinions. It may well be that few people adopt an overall, multidimensional health life style. Instead, most may tend to accept a few healthy opinions and actions, while rejecting others.

Furthermore, it appears that people may engage in behavior that is deemed unhealthy by the medical profession, yet still remain relatively unconcerned about their health status. For example, the male data correlations between two broad health assessment items, "I consider myself to be a very healthy person" and "I worry a lot about my health," and two items relevant to more specific health behavior, "I am overweight" and "I drink more than I should," tend to be rather low:

                        Healthy person        Worry about health

Overweight              -.06                             .15

Drink too much       -.04                             .06

For the most part, the factors shown in Tables 1 and 2 seem straightforward and self-explanatory. Since the verbatim items representing each factor are listed, there is no need to review the contents of each. Thus our comments will be limited to a few observations.

Several types of health-related themes are reflected in the factor structures: (1) social and recreational activities; (2) eating, smoking and drinking; (3) opinions about public policy, e.g. about nuclear power, TV violence, etc.; (4) overall assessment of health status; and, (5) tension and stress. These seem to form the dimensions of health maintenance opinions and activities.

Although the male and female items are not identical, ten common dimensions emerged from the separate analysis. These are labeled Active Sports, Personal Hygiene, Exercise, Conservative Values, Like Smoking, Plan to Quit Smoking, Heavy Smoking, Heavy Drinking, Heavy Eating, and Health Anxiety. The remaining female factors tend to focus on anxiety and nutritional matters, and the remaining male factors reflect outdoor activities and work.

It is interesting that the frequency-of-activity items, marked by asterisks in Tables 1 and 2, tend to factor with the disagree-to-agree opinion and interest statements. Since both the activities and opinions are self-reported, this does not prove the validity of the life style data. However, it does indicate that the data show internal coherence and meaningfulness.

Some further points:

The female "Like Smoking" and "Heavy Eating" factors both reflect some psychological imbalance or dissatisfaction, i.e., smokers often consider quitting and heavy eaters realize that they are overweight and eat too much. With regard to smoking, this is also true of males.

Both men and women who want to quit smoking (Plan-to-Quit Smoking factor) appear to consider switching to low tar cigarettes as a possible alternative action.

The male and female Heavy Smoking dimensions reflect: (1) social influence on health, i.e., children nagging the parent to quit; (2) denial, i.e., "I don't believe smoking is...harmful .... ;" and, (3) addiction, i.e., "I could (not) stop smoking any time .... "

Reported incidence of self breast examination seems to correlate negatively with worry, tension and dissension within the family, suggesting that stress may interfere with positive health actions.

Women who are "Weight Watching" are trying to cut both food and alcohol calories.

In the male data, there is a separation between team sports activities (Active Sports) and individual exercise (Exercise).

In the male data, worry about health (Health Anxiety) is specifically associated with taking pills and tranquilizers, whereas in the female data, pills, tranquilizers and meditation form a separate factor (Pills and Tranquilizers) that seems to reflect response to general stress.






The life style questionnaires yielded data on a total of 22 demographic variables. Of these, four were selected to relate to the health maintenance R factors. These four--occupation of head of household, age, education and city size--are frequently used both in health and marketing segmentation studies.

A factor loadings for the four selected demographic variables on the health maintenance factors are shown in Table 3. These in effect are the correlations between the demographics and the health related dimensions. It should be noted that the male and female health items were factored both with and without the demographic variables, and the factor structure remained essentially the same. It is data from the factor analysis that included demographic data that are shown in this paper.

The correlations in Table 3 indicate that age and occupation of the head of household account for more variance in health related factors than education or city size. Since age is closely related to physical condition and objective medical need, one would expect a strong correlation between that demographic and health prevention ideas and actions.

The demographic variables are coded as follows: occupation of the head of household is coded from executives (scored 1) to never employed (scored 8), so that the more prestigious the occupation, the lower the score; age refers to the respondents' reported years of age; education goes from less than eighth grade (1) to graduate work (9); and city size runs from 1 million or more (1) to nonmetro area (5).

Table 3 indicates that the Active Sports (M, F) and Workaholic (M) factors tend to describe younger, relatively upscale respondents. Women who score high on the Pills and Tranquilizers dimension tend to be older and less well educated. And there are marked differences between heavy drinkers and heavy smokers. Heavy drinking is associated with relatively prestigious occupations, older age, and big city living. (This seems to be the demographic profile of the "three martini lunch.") Heavy smokers, on the other hand, tend to be in low prestige occupations and, in the case of women, less educated.

Among the factors that show little relationship to the four selected demographics are Pro Nuclear Power (M), High Tar Smokers (M, F), Heavy Eating (M, F) and Tension (F).




"Most of the bad things that happen to people are at present beyond the reach of medicine" (Wildavsky, 1977). Alcohol consumption, cigarette-smoking, stress, violence, pollution, etc., are an integral part of the modern pattern of life in industrial societies. Changes in the primary sources of morbidity and mortality during the twentieth century illustrate the etiological significance of environmental and life style factors. Thus, efforts directed toward disease prevention and health promotion must extend beyond the traditional medical care system to include institutional measures to promote health maintenance, encourage advances in technology and the quality of the environment, and advocate modification of individual life styles.

Research studies to date have not provided much evidence concerning the influence of life style on health behavior (Ng, Davis and Mendorsheid, 1978). Our view of the role of life style variables is that they serve to condition both the individual's perceptions and his attitudes and subsequent behavior with regard to preventative health actions. This view modifies the more narrow approach taken in Rosenstock's well known Health Belief Model-HBM (Rosenstock, 1974). The HBM focuses on the individual's preventative health actions as motivated by his perceived susceptibility to illness, the perceived severity of illness, his belief about the efficacy of alternative actions, and psychological barriers to actions--all mediated by the individual's demographic characteristics. The model falls short in dealing with problems of environmental health and personal life style. It has been suggested that life style must at some point be considered a conditional variable (Stone, Cohen, et. al., 1979), which ultimately should provide viable input relevant to decision making in the area of public health policy.


American Hospital Association. "National Health Expenditures, Fiscal Year 1977." May, 1978.

Bearden, W. O., Teel, J. E. and Durand, R. M. "Media Usage, Psychographics, and Demographic Dimensions of Retail Shoppers." Journal of Retailing, Spring, 1978, 65-74.

Belloc, N.B. "Relationship of Health Practices and Mortality.'' Preventative Medicine, Vol. 2, 1973, 67-81. Cairns, J. "The Cancer Problem." Scientific American, 233 (5), pp. 64-78.

Kasl, S. V., and Cobb, S. "Health Behavior, Illness Behavior, and Sick Role Behavior." Archives of Environmental Health, 1966, 12, 246-266.

Cooper, P. D., Kehoe, W. J. and Murphy, P. E.(eds.) Preventative Health Care and Marketing: Interdisciplinary and Interorganizational Perspectives. Chicago: American Marketing Association, 1978.

Keyfitz, Nathan. "Improving Life Expectancy: An Uphill Road Ahead." American Journal of Public Health, October, 1978, Vol. 68, No. 10, 954-956.

Mehrota, S. and Wells, W. D. "Psychographics and Buyer Behavior: Theory and Recent Empirical Findings" In A. G. Woodside, J. N. Sheth and P. D. Bennett(eds.) Consumer and Industrial Buying Behavior, New York, North-Holland, 1977, 49-66.

Ng, L. K., Davis, D. L. and Mendorsheid, R. W., "The Health Promotion Organization: A Practical Intervention Designed to Promote Healthy Living." Public Health Reports, 93(5) September-October 1978, 446-455. Plummet, 3.T., "Applications of Life Style Research to the Creation of Advertising Campaigns" in W. D. Wells, Life Style and Psychographics, Chicago: American Marketing Association, 1974, 167-180.

Preston, Samuel H. "Mortality Trends." Annual Review of Sociology, 1977, Vol., 3, 163-178.

Rosenstock, I. M. "The Health Belief Model and Preventative Health Behavior" in M. H. Backer(ed.) The Health Belief Model and Personal Health Behavior, New Jersey, Charles B. Slack, Inc., 1974, 27-59.

Stone, G. C., Cohen, F. and Adler, N.E.(ed.) Health Psychology, San Francisco: Jossey-Bass Publishers, 1979.

Wells, W. D. and Tigert, D. J. "Activities, Interests, and Opinions," Journal of Advertising Research, 11, August 1971, 27-35.

Wildavsky, A., "Doing Better and Feeling Worse: The Political Pathology of Health Policy." Daedalus, 106, Winter 1977, 105-123.

Wind, Y., and Green, P. E., "Some Conceptual, Measurement, and Analytical Problems in Life Style Research," in W. D. Wells, Life Style and Psychographics, Chicago; American Marketing Association, 1974, 97-126.

Wortzel, L. H. "The Behavior of the Health-Care Consumer: A Selective Review," in B. B. Anderson, Advances in Consumer Research. Vol. 3, 1976, 296-301.