Respondents demographic characteristics
The cooperation rate was 30.4% (1216 responses from subjects in 4000 sampled households) in this study. Of those who cooperated, 98.2% provided an answer, including I dont know, to all questions, while the rest did not answer the question about the degree to which they believed cancer could be prevented by improving lifestyle Reasons for lack of responses were refusal to participate (n=1122), absence from home in the survey period (n=1071), change of address after sampling (n=143), lack of knowledge about the address (n=9), and other undetermined reasons (n=439). However, response rates could not be calculated according to the formula determined by the American Association for Public Opinion Research due to lack detailed data on reasons for non-responses in this study16. The response rate did not differ by geographical area or city-scale of the study areas (Supplementary Table S1). Mean age of the respondents was 54.8years, and 46.3% (n=563) of respondents were men (Table 1). The mean age was statistically significantly lower and the educational status was statistically significantly higher among men than women.
Although women generally indicated higher attributable fractions of cancer risk than men, the order of magnitude was similar in both sexes (Table 2). Tobacco smoking (55.7%, mean attributable fraction of cancer risk overall) and cancer-causing viral and bacterial infection (52.0%) were regarded highly as causes of cancer. In contrast, participants regarded the attributable fraction of cancer risk of other lifestyle factors to be much lower (obesity [36.6%], physical inactivity [31.9%], unbalanced diet [30.9%], and alcohol consumption [26.2%]) than that of other environmental factors (endocrine-disrupting chemicals [42.7%], air pollution [40.0%], occupational exposure [38.1%], and food additives and pesticides [33.9%]). While respondents thought the attributable fraction of cancer risk of genetic factors was high (51.7%), they thought a small fraction of cancers were preventable by improving lifestyle (34.6%). Missing data were found in calculation of the attributable fraction due to exclusion of responses of "I don't know" (4.4% to 13.6%) and no answer for the degree prevented by improving lifestyle (1.8%) (Table 2).
A large proportion of respondents indicated they were interested in cancer prevention (n=980, 80.8%), with the rate being significantly higher in women (n=562, 86.3%) than in men (n=418, 74.5%) (Supplementary Table S2). Respondents who indicated they were interested in cancer prevention tended to be older and marginally highly educated. The presence of interest in cancer prevention did not differ by the city-scale of study areas.
The demographic characteristics (i.e. sex, age, educational status, and study area) of the respondents who indicated they engaged in any health behavior for cancer prevention were similar to those who indicated they were interested in cancer prevention (Supplementary Table S2). Among the individual health behaviors (Table 3), abstinence from smoking (38.4%) accounted for the highest proportion of health behaviors for cancer prevention in men, followed by improving diet (30.5%), whereas improving diet (44.6%) in women, followed by cancer screening/health check-up (40.0%). While the proportion who engaged in health behaviors tended to increase with age, the proportion who indicated they abstained from smoking and drinking tended to be high among both younger and older generations (Supplementary Table S3).
The majority of respondents indicated they obtained information on cancer prevention from any source (n=1158, 95.2%) (Table 4). The most common source was television (n=986, 81.2%), followed by print media (n=754, 62.3%; including newspapers, books, magazines, brochures provided by pharmacies/hospitals, and advertisements), interpersonal sources including health professionals (n=337, 27.7%; such as instructions from professionals and health classes) and family/friends (n=333, 27.3%), the internet (n=280, 23.1%; including websites of public institutions and other organizations, and social media), and radio (n=111, 9.2%). Among the types of print media, newspapers were used by 42.6% (n=515) of respondents, while books were used by 9.0% (n=109).
Older respondents were more likely to use radio (age [continuous], odds ratio [OR]=1.03, 99.9231% CI by Bonferroni correction: 1.011.05; multivariate-adjusted model), newspapers (OR=1.04, 99.9231% CI 1.021.05), while younger respondents were more likely to use social media (OR=0.96, 99.9231% CI 0.930.98) (Supplementary Tables S3, S4). Further, women were more likely to use interpersonal sources compared with men: health classes (women vs. men, OR=2.44, 99.9231% CI 1.195.00) and family/friends (OR=1.76, 99.9231% CI 1.142.70). Moreover, respondents with higher levels of education tended to be more likely to use sources of print media, except advertisements, and internet sources, except social media, but not statistically significant. Preference for information sources did not differ by city-scale of the study areas.
We investigated the association between information sources and health behaviors after adjusting for age, sex, educational status, city-scale of study area, and interest in cancer prevention (Table 5). Among print media, books were associated with improving diet (OR=2.52, 99.9231% CI 1.185.39) and exercise (OR=2.33, 99.9231% CI 1.124.85), and newspapers and magazines were associated with a broad range of health behaviors. Brochures provided by pharmacies/hospitals were associated with cancer screening/health check-up (OR=2.31, 99.9231% CI 1.293.38), and advertisements were associated with abstinence from smoking (OR=2.14, 99.9231% CI 1.193.83) and drinking (OR=2.28, 99.9231% CI 1.303.98). Among the online sources, websites of public institutions were associated with a broad range of health behaviors including improving diet (OR=2.11, 99.9231% CI 1.074.17), cancer screening/health check-up (OR=2.12, 99.9231% CI 1.114.04), and abstinence from smoking (OR=2.56, 99.9231% CI 1.225.37). Information obtained from health professionals including instructions and health classes was associated with a broad range of health behaviors including improving diet, exercise, cancer screening/health check-up, and abstinence from smoking/drinking (OR=1.92 to 3.11, P-value=0.0039 to<0.0001). Information obtainment from family/friends was associated with abstinence from smoking (OR=1.68, 99.9231% CI 1.032.73). In contrast, television and social media were not associated with an increase in any type of health behavior.
Read more from the original source:
A cross-sectional survey on awareness of cancer risk factors, information sources and health behaviors for cancer prevention in Japan | Scientific...