Health beliefs play an important role in the implementation of cancer prevention and treatment, and health beliefs may differ between older patients and younger patients. In their report in this issue, Clipp et al. make the case for assessing health beliefs in future clinical trials and correlating these beliefs with treatment outcome.
See also pages 1085–94, this issue.
Cancer is more likely to present in individuals age ≥ 65 years than in younger individuals,1 and the percentage of older individuals with cancer is expected to increase with the expansion of the older population. Despite the size of the problem, the information related to effectiveness of cancer prevention and treatment in the elderly is limited due to the exclusion of older individuals from clinical trials,2, 3 despite it having been demonstratedn that they may benefit from preventative4 and therapeutic interventions to the same extent as younger individuals5 and also are as willing as younger patients to participate in clinical research.6, 7 It appears that one of the main obstacles to the enrollment of older individuals in intervention studies has been the diversity of the older population and the scarcity of evaluation instruments capable of accounting for this diversity. The article by Clipp et al.8 in this issue of Cancer explores a very important and virtually unknown aspect of geriatric oncology: the difference in health beliefs between older and younger patients at risk for colorectal carcinoma (CRC). The authors performed an age-weighted analysis of a data base of 1275 patients ages 40–75 years who were enrolled in a randomized clinical trial on the chemoprevention of CRC after polypectomy: Each patient was asked questions related to their perceived risk of CRC, their readiness to change their lifestyle, and their expectancies and self-efficacy. In the same study, issues of physical vulnerability (comorbidity and self-rated health) emotional vulnerability (depressive mood and quality of life), social vulnerability (presence of a cohabitant and number of confidants), and financial vulnerability (income) also were explored. Patients age ≥ 60 years were less likely than younger patients to have graduated from college and were more likely to be widowed and to have a yearly income < $45,000. Despite a similar risk of CRC, compared with younger patients, older patients felt less at risk for CRC, had a lower level of concern for the development of the disease, had less motivation to change their health habits, and had less confidence in the effectiveness of these changes for preventing CRC. It is interesting to note that older individuals were as confident as younger individuals that they would be able to change their health habits. The degrees of physical, social, and financial vulnerability were greater for older patients, but the prevalence of a depressive mood was lower and the rating of quality of life was similar for the younger and older groups. The concern for CRC and the motivation to change were lower for older patients with physical vulnerabilities, and the outcome expectancies and motivations to change were lower for those with poorer social support.
In two respects, the importance of this work cannot be over-emphasized. First, it highlights the importance of health beliefs in implementing cancer prevention. Health beliefs may be influenced by education and new information; therefore, beliefs may provide a target for promoting cancer prevention. This finding is new and deserves to be addressed in future studies of public health concern. Second, the authors document an age-related difference in health beliefs. The fact that, of necessity, this finding is painted in broad brush strokes does not make it less convincing. Affecting the health beliefs of the elderly is particularly important today because estimates of life expectancy4 and treatment complications9 are becoming more precise, allowing health care providers to tailor prevention and treatment programs to the characteristics of individual patients, whereas the risks of the most common complications of cancer chemotherapy have been ameliorated by effective antidotes.10 The unnecessary shortening of human life due to incorrect beliefs would be especially tragic in this information era.
Another important (albeit not new) aspect of the study by Clipp et al. is the methodology. After almost 40 years of experience in clinical trials of cancer treatment, the academic community is understandably skeptical of retrospective subset analysis. However, this type of analysis is essential to obtain hypothesis-generating information in older individuals. Without this type of study, our knowledge of geriatric oncology and our ability to conduct prospective trials in older cancer patients would be limited even more than it is currently. The limitations of the study are those to be expected in a retrospective analysis of a large population. First, significant portions of older individuals (age > 75 years) are not represented, and individuals age > 70 years most likely also are under-represented. Second, there is no documentation of a correspondence between the questions asked and the actual ability to institute behavioral modifications. One cannot assume that the two are equivalent. However, this flaw does not affect the difference in health beliefs the study purports to document: rather, it emphasizes the need to document the effectiveness on lifestyle changes of interventions aimed at modifying health beliefs. Third, the questions, of necessity, were generic. Fourth, important elements of geriatric assessment are missing, especially the evaluation of functional status, which is independent from comorbidity.11 These flaws should be imputed not to the authors but to the design of the original study and may be corrected in future trials.
In conclusion, Clipp et al, convincingly demonstrate that health beliefs play an important role in the implementation of prevention of cancer and, presumably, other disease and that health beliefs of older and younger patients may be different. Their study makes the case for assessing health beliefs in future clinical trials and correlating these beliefs with treatment outcome. This important finding needs to be complemented by more complete information in individuals age ≥ 70 years by correlating these beliefs with other geriatric parameters, including functional status, depression, and polypharmacy, and by interventional studies aimed at changing health beliefs.