Scottish Government

CHAPTER SEVEN HEALTH BEHAVIOURS

7.1 In this chapter we look at 4 health related behaviours: smoking; drinking alcohol; eating 5 or more portions of fruit and vegetables per day; and taking exercise by walking around the neighbourhood.

Change over time in smoking

7.2 Participants were asked about their smoking habits; all respondents were asked at both waves if they were currently smoking and current smokers in the Intervention Group were also asked at both waves if they were thinking of giving up ⁷.

7.3 In Wave 1, the Intervention Group contained slightly more current smokers (58.7% vs. 54.5% in the Control Group), though the difference was not significant (p=0.23, Table 7.1). Respondents in older households, longer-term residents, and those who were working were less likely to be current smokers.

7.4 Comparisons of intention to give up smoking are restricted to current smokers in the Intervention Group only. In Wave 1, the healthier behaviour of working respondents in terms of current smoking was echoed in a greater proportion of smokers who were thinking of quitting (72.3% of working smokers vs. 46.1% of non-working smokers, p=0.001). Older households and longer-term residents were less likely to be thinking of quitting, which seems to be in contrast with the healthier behaviour suggested by fewer current smokers in these groups. However it is possible, given their greater age, that many of the less committed smokers in these groups had already given up, so that only the more committed smokers remain as 'current smokers' and thus are less likely to be 'thinking of quitting'. As can be seen from Table 7.2, the number of current smokers in the Intervention Group thinking of quitting fell over time from 55% to 47%.

7.5 By Wave 3, the difference in current smoking habits between the intervention and Control Groups had widened. There were fewer current smokers in both groups but the reduction was much more marked in the Control Group and, after adjustment for Wave 1 smoking status, respondents in the Intervention Group were almost 50% more likely to be current smokers than those in the Control Group, though the difference is not significant (Table 7.1).

Table 7.1 Current smoking habits

Intention to quit smoking by location

7.6 Comparisons of intention to give up smoking by relocation status among current smokers in the Intervention Group are shown in Table 7.2. As already noted, intention to quit smoking fell over time among the Intervention Group, so that rehousing overall did not prompt many people to consider a fresh start - recall that smoking only fell by 3% among the Intervention Group whilst intention to quit also fell by more than this, namely 8%. However, those who moved neighbourhood as well as moving home (relocators) were around 50% more likely to be thinking of quitting than those who moved home within the same neighbourhood (non-relocators), indicating that if there is any effect of attempting a 'fresh start', it is more likely to operate where people move location as well as home. Again however this finding is tentative as the effect is not statistically significant.

Table 7.2 Thinking of quitting smoking at Wave 1 and Wave 3, and by relocation status (Intervention Group, current smokers only)

Smoking and dwelling changes

7.7 Differences among the Intervention Group in current smoking and in thinking of quitting smoking at Wave 3, by dwelling transition, were not statistically significant. There was however an indication that people who moved from a house to a house were less likely to be current smokers than other movers: 60% of those who moved from a flat to a house were current smokers, compared with 40% of those who moved from a house to a house (adjusted odds ratio 0.50, p=0.08). This may reflect the fact that at Wave 1, flat dwellers are far more likely to be smokers than occupants of houses (62% vs 43%).

7.8 There were several instances where reported improvements in dwelling conditions were associated with higher rates of current smoking at Wave 3. Thus, those people in the Intervention Group who reported the most gains in dwelling fabric conditions (e.g. damp, state of repair), dwelling amenities (e.g. heating, double glazing) and privacy (including noise), were also more likely to be current smokers. These associations are hard to explain, and may have no causal relationship to one another. They do not, however, lend any support to the hypothesis that smoking may in part be a stress-response to poor housing conditions; at least, that has not been found here.

7.9 It was also the case that those people who reported a worsening in outlook from their home, were both more likely to be current smokers at Wave 3, but also more likely to be thinking of quitting at Wave 3 - see Table 7.3. Further, a better outlook through rehousing did not result in more people thinking of quitting smoking.

Table 7.3 Current smoking and thinking of quitting, by change in outlook from the home, at Wave 3 (Intervention Group)

Smoking and neighbourhood changes

7.10 We found no association between reported changes in neighbourhood conditions (either neighbourhood environment and infrastructure nor neighbourhood crime and anti-social behaviour) and smoking habits. Neighbourhood changes were not related to changes in current smoking nor intention to quit smoking.

Change in drinking habits

7.11 All participants were asked whether they ever drank alcohol and, if so, how much on average they drank per week. Current drinkers were defined as those answering "yes" to drinking alcohol, even if it was only occasionally (14% of current drinkers reported drinking less than 1 unit per week); Heavy drinkers were defined as those exceeding government "recommended sensible limits" of 21 units per week for men and 14 units per week for women, and were compared with drinkers below these levels and non-drinkers combined (i.e. light drinkers, non-drinkers and ex-drinkers combined).

7.12 Drinking was more common among the Intervention Group in Wave 1 (77.0% were current drinkers vs. 66.4% of the Control Group, p=0.002 - see Table 7.4). Drinking was more common in men (78.9% vs. 68.8% in women, p=0.01), in families with children and shorter-term residents (both of which are consistent with patterns of more drinking in younger respondents) and in those who were working (85.7% vs. 64.3% in non-working, p<0.001). This final observation is in contrast to other factors where working respondents display "better" health and are less likely to report unhealthy behaviours, but may reflect financial advantage and therefore better access to alcohol.

7.13 The rates of heavy drinking in Wave 1 were similar in the Intervention and Control Groups. However, it was markedly more common in men in spite of the use of sex-specific definitions (13.6% of men vs. 4.7% of women, p<0.001).

Table 7.4 Drinking habits in Intervention vs. Control Groups

7.14 The proportions of current drinkers in the Intervention and Control Groups in Wave 3 were similar to those in Wave 1, with again a higher rate of current drinking in the Intervention Group. However, adjusted analyses indicate that there was no differential change in drinking habits between the 2 groups (i.e. the differences in Wave 1 remained in Wave 3, suggesting that the intervention had no impact on current drinking status). In contrast there was a modest but non-significant increase in heavy drinking between Waves in the Control Group compared with a small decline in the Intervention Group (see final column of Table 7.4 for Wave 3 odds ratios adjusted for Wave 1 drinking habits).

Drinking habits and dwelling changes

7.15 The type of dwelling transition experienced by Intervention Group households had no association with rates of current drinking or heavy drinking at Wave 3. Many aspects of dwelling change also bore no relation to changes in drinking habits: this was true for changes in floor level, having a garden, dwelling amenities, space, and dwelling safety and security.

7.16 Where people reported major improvements in the quality of dwelling fabric, they were also more likely to have moved from being a heavy to a lighter drinker over time, and less likely to be a heavy drinker at Wave 3 - see Table 7.5.

Table 7.5 Changes in drinking habits by changes in dwelling fabric (Intervention Group)

Drinking habits and neighbourhood change

7.17 There was no significant difference in rates of current drinking among Intervention Group members at Wave 3 according to whether or not they had moved neighbourhood (73.6% among non-relocators vs. 78.8% among relocators, p=0.60, adjusted for Wave 1 drinking). Heavy drinkers were more common among non-relocators, but the difference again was not significant (9.7% vs 6.1% respectively, p=0.49, adjusted for Wave 1 drinking).

7.18 Where people reported improvements in neighbourhood crime and anti-social behaviour, they were also more likely to be heavy drinkers at Wave 3 - see Table 7.6. It may be that those people who have been longest exposed to neighbourhood social problems are both more likely to report improvements in the situation over time and more likely to drink heavily, partly as a result of long term exposure to such environments.

Table 7.6 Changes in drinking habits by changes in neighbourhood crime and antisocial behaviour (Intervention Group)

Change in diet: fruit and vegetable consumption

7.19 Respondents were asked, on average, how many portions of "fresh fruit or fruit juice" and how many portions of "fresh or frozen vegetables" they had per day. The number of portions of fruit and vegetables were added and respondents sub-divided into those who did or did not have 5 or more per day.

7.20 There was little difference between Intervention and Control Groups in Wave 1 with around a quarter in each reporting having 5 or more portions of fruit or vegetables per day (Table 7.7). Wave 1 fruit and vegetable consumption was greater among respondents who were working (33.0% vs. 23.9% in those not working, p=0.01).

Table 7.7 Five or more portions of fruit or vegetables per day

7.21 In Wave 3, the Intervention and Control Groups were even more similar with 27.1% of Control and 28.2% of Intervention respondents reporting 5 or more per day. The difference between the groups was not significant, so we must conclude that rehousing itself had little or no impact upon fruit and vegetable consumption. Furthermore, analysis showed that particular aspects of dwelling and neighbourhood change were not related to changes in eating habits.

Diet and location

7.22 The effects of relocation status in the Intervention Group are shown in Table 7.8. Once again there was evidence to suggest that the new start offered by moving location may be a spur to improvement in health behaviour with relocators more than twice as likely to report eating 5 or more portions per day than non-relocators.

Table 7.8 Self-reported consumption of five or more portions of fruit or vegetables per day by relocation status (Intervention Group only)

Change in exercise: walking around the neighbourhood

7.23 Respondents in the Intervention Group were asked at Waves 1 and 3 when they "last went for a walk around (their) neighbourhood" ⁸. In Wave 1, 53.8% of the Intervention Group said they had last walked round their neighbourhood a day ago (vs. a week or more). There were no marked differences in Wave 1 by sex, SIP status, household type, length of residence or working status.

7.24 By Wave 3, the number of Intervention Group respondents who reported walking around their neighbourhood the previous day had fallen to 41.2%; this was however still higher than the rate found in the Control Group, of 29.9% ⁹. Thus, rehousing itself did not increase the taking of local exercise in the form of walking, but it may have prevented it reducing even further than it might otherwise have done, although we cannot be certain about this.

7.25 In Wave 3, respondents in families with non-dependent children and older households were less likely to have walked around their neighbourhood the previous day (see Table 7.9); this was independent of walking habits in Wave 1.

Table 7.9 Last walked round neighbourhood a day ago (vs. a week or more) in Wave 3 by household type (Intervention Group only)

Dwelling type and walking around the neighbourhood

7.26 Respondents living in flats and houses in Wave 1 were very similar in their responses, with 59.7% of house occupants and 55.4% of flat occupants reporting having walked around the neighbourhood in the previous day. However, respondents who moved into a flat in Wave 3, were only half as likely to have walked round their neighbourhood in the previous day, regardless of whether they had previously lived in a house or a flat (see Table 7.10). Almost half (47.6%) of those people in the Intervention Group living above the ground floor in Wave 3 reported a worsening of their walking habits over time, nearly twice as many as among those living on the ground floor ( p=0.02).

Table 7.10 Last walked round neighbourhood a day ago (vs. a week or more) in Wave 3 by dwelling type in Wave 3 (Intervention Group only)

7.27 Where people reported that they had gained in terms of internal dwelling space through rehousing, they were nearly 3 times as likely to report an improvement in their walking habits as those who reported either no change or a worsening of dwelling space (24.4% vs 8.9%, p=0.004).

Summary

In this chapter we looked at smoking, drinking, diet and exercise. Our key findings were:

• Levels of current smoking were higher in the Intervention Group at Wave 1 and this difference widened over time. Those people rehoused were nearly 50% more likely to be smokers at Wave 3, and thus rehousing itself did not help reduce smoking rates.
• However, where people relocated to a new neighbourhood through rehousing, they were 50% more likely to be thinking of quitting smoking than those rehoused in the same neighbourhood. It would appear that where residential change is more comprehensive (involving area as well as dwelling) people may be more prepared to make a 'fresh start'. But, contrary to expectations, having a better outlook from the home after rehousing was not associated with better smoking outcomes.
• People who reported a worsening of outlook or view were more likely than others to be smokers at Wave 3, but also more likely to be thinking of quitting.
• Levels of current drinking were also higher in the Intervention Group at Wave 1 and at Wave 3, with no differential change over time. Again, rehousing did not help attenuate drinking habits.
• Reported improvements in the dwelling fabric were associated with better drinking outcomes at Wave 3, among the Intervention Group.
• Relocation to a new neighbourhood was not associated with improved drinking outcomes, unlike in the case of smoking. What is more, drinking habits were worse at Wave 3 among those people who reported improvements in neighbourhood crime and anti-social behaviour. It may be that any negative effect of being concerned about local crime persists, despite change in conditions over time.
• Eating habits were no different between the Intervention and Control Groups at either Wave 1 or Wave 3. Rehousing itself had no impact on diet.
• However, those people who were relocated to a new neighbourhood as a result of rehousing were twice as likely to eat 5 portions of fruit and vegetables a day as others (after adjustment for diet at Wave 1), suggesting that relocation may prompt the maintenance of a healthy diet.
• Walking around the neighbourhood fell over time among the Intervention Group, but was nonetheless higher at Wave 3 than in the Control Group. Rehousing did not improve walking rates, but may have prevented their worsening.
• At Wave 3, those people who were rehoused into a house rather than a flat were far more likely to have walked around their neighbourhood the previous day. Thus, rehousing to a house was associated with better walking habits, as was rehousing that provided more suitable internal dwelling space.