Methods: Health Information National Trends Survey 5 (HINTS5) Cycle 1, 2 and 3 data were merged and analyzed in this cross-sectional study (N=12,227). The dependent variable, self-rated health (SRH), was measured on a 5-point Likert scale. The independent variable, PA adherence, was measured categorically: no adherence, partial adherence and full adherence. Cancer history (0=no, 1=yes) was tested as a moderator. Age was measured categorically: young (18-44), middle-age (45-64), and older (65<). Demographic characteristics, socioeconomic status, cancer beliefs, smoking status, BMI and chronic diseases were included as covariates. Multilinear regressions were performed without and with the interactions to test effects of PA adherence on SRH and the moderation effect of having cancer history on this association. Then, multilinear regressions were conducted to investigate how this moderation effects differed across age groups.
Results: People previously diagnosed with cancer were more likely to be white, older and experience chronic disease while less likely to be educated or adherent to PA guidelines. Multiple linear regression among total samples showed that, compared with no PA adherence, partial adherence yielded a .22-unit SRH increase (p<.001), whereas full adherence yielded a .41-unit SRH increase (p<.001). After including interactions, this significant effect still held, while interaction1 (cancer history × partial adherence) was significant (ß=.14, p<.05) and interaction2 (cancer history × full adherence) was marginally significant (ß=.11, p=.051). Multilinear regressions across all age groups indicated consistent main effects of partial and full adherence on SRH. Yet, interactions between cancer history and PA guidelines adherence revealed interesting distinctions: Among young adults, interaction2 (cancer history × full adherence) was significant (ß=.90, p<.01); among middle-aged adults, no significant interactions were found; among older adults, interaction1 (cancer history × partial adherence) was significant (ß=.17, p<.05).
Conclusion: These findings suggest people with a cancer history would benefit more from adhering to PA guidelines than their general-population peers at the same adherence level. Additionally, the health benefit patterns of PA adherence are different across age groups: young cancer survivors would benefit more from full adherence; older cancer survivors would benefit more from partial adherence; no significant differences are found among middle-aged group. Therefore, future PA interventions for cancer survivors should consider these distinct patterns across age groups and generate age-tailored interventions to realize greater benefits.