A multilevel analysis of diet and socio-economic status in Scotland: investigating the 'Glasgow effect'

Public Health Nutr. 2009 Sep;12(9):1351-8. doi: 10.1017/S1368980008004047. Epub 2008 Nov 25.


Objective: To investigate differences between dietary habits in Glasgow and those in the rest of Scotland and the role that socio-economic factors have in explaining these.

Design: Data on age, sex, area deprivation, social class, educational qualifications, economic activity, health board region, postcode sector area and informants' usual intake of foods covering sugary foods, snacks, fibre, starch, meat, fish, spreading fats, dairy products, salt, dietary supplements, fruit and vegetables were available from the 1995, 1998 and 2003 Scottish Health Surveys. Multilevel logistic regression was used to model the relationship between diet and living in Greater Glasgow compared with elsewhere in Scotland, unadjusted and adjusted for age, survey year and socio-economic factors, accounting for the clustering within postcode sector area.

Setting: Scotland.

Subjects: Subjects comprised 11 075 male and 14 052 female respondents.

Results: Lower consumption of high-fibre bread and potatoes/pasta/rice (among men and women), of cakes (men) and of cereals, meat, skimmed/semi-skimmed milk and green vegetables (women) in Glasgow was explained by socio-economic factors, as was higher consumption of non-diet soft drinks among women; lower consumption of ice cream, bread, cereals, meat and green vegetables (men) and high butter and salt consumption (women) in Greater Glasgow were not.

Conclusion: Associations between unhealthy eating and deprivation accounted for much of the tendency of people in Glasgow to have poor diets. Policies are needed to encourage improvements in diet in Glasgow and more effort is required to reduce social inequalities in eating habits. Glasgow's poor diet will remain unless problems associated with poverty are tackled.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Age Distribution
  • Aged
  • Cluster Analysis
  • Diet / statistics & numerical data*
  • Feeding Behavior
  • Female
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Poverty*
  • Scotland
  • Sex Distribution
  • Social Class*
  • Socioeconomic Factors
  • Young Adult