Sociodemographic characteristics of the clusters To facilitate specific target group prevention measures, sociodemographic indicators of the clusters were investigated and set in relation to Cluster 1 (‘No Risk Behaviours’). The odds ratios for cluster membership according to different social factors are shown in table 2. The composition of the ‘healthaware’ Cluster 1 was as follows: 60.4% female, 46.3% aged 50– 59 and 71% in the middle or upper SES categories. In contrast with the reference group, the ‘Physically Inactives’ (Cluster 2) were significantly more likely to live alone (OR=1.60). The ‘Fruit and Vegetable Avoiders’ (Cluster 3) were also significantly more likely to be without a partner (OR=1.98). In particular, this group was comprised of divorced or widowed men. The ‘Smokers with Risk Behaviours’ (Cluster 4) were less likely to be female (OR=0.37), to be within the oldest age group (OR=0.37) and to have middle or high SES (OR=0.35). This cluster largely included younger men with low SES, indicating an interaction effect between gender and SES (OR=3.41; 95% CI: 1.13–10.28). This group also was more likely to live alone (OR=2.62). As with the previous two clusters (Clusters 3 and 4), the ‘Drinkers with Risk Behaviours’ were significantly less likely to be female (OR=0.60) and also showed an interaction effect between gender and SES (OR=2.95; 95% CI: 1.36–6.39). The health behaviour clusters identified were plotted on a two-dimensional matrix by gender and SES. Figure 2 shows
the size and position of the clusters, which was based on the proportion of men and individuals in the upper social class. Generally, the higher the percentage of men in a cluster, the more unfavourable the health behaviour. The interaction effect between gender and SES is clearly demonstrated for Cluster 4 (‘Smokers with Risk Behaviours’).
Sociodemographic characteristics of the clusters To facilitate specific target group prevention measures, sociodemographic indicators of the clusters were investigated and set in relation to Cluster 1 (‘No Risk Behaviours’). The odds ratios for cluster membership according to different social factors are shown in table 2. The composition of the ‘healthaware’ Cluster 1 was as follows: 60.4% female, 46.3% aged 50– 59 and 71% in the middle or upper SES categories. In contrast with the reference group, the ‘Physically Inactives’ (Cluster 2) were significantly more likely to live alone (OR=1.60). The ‘Fruit and Vegetable Avoiders’ (Cluster 3) were also significantly more likely to be without a partner (OR=1.98). In particular, this group was comprised of divorced or widowed men. The ‘Smokers with Risk Behaviours’ (Cluster 4) were less likely to be female (OR=0.37), to be within the oldest age group (OR=0.37) and to have middle or high SES (OR=0.35). This cluster largely included younger men with low SES, indicating an interaction effect between gender and SES (OR=3.41; 95% CI: 1.13–10.28). This group also was more likely to live alone (OR=2.62). As with the previous two clusters (Clusters 3 and 4), the ‘Drinkers with Risk Behaviours’ were significantly less likely to be female (OR=0.60) and also showed an interaction effect between gender and SES (OR=2.95; 95% CI: 1.36–6.39). The health behaviour clusters identified were plotted on a two-dimensional matrix by gender and SES. Figure 2 showsthe size and position of the clusters, which was based on the proportion of men and individuals in the upper social class. Generally, the higher the percentage of men in a cluster, the more unfavourable the health behaviour. The interaction effect between gender and SES is clearly demonstrated for Cluster 4 (‘Smokers with Risk Behaviours’).
การแปล กรุณารอสักครู่..
