Effect of neighborhood social participation on individual use of hormone replacement therapy and antihypertensive medication: a multilevel analysis.

Research output: Contribution to journalArticle


The authors investigated a possible contextual effect of neighborhood on individual use of hormone replacement therapy (HRT) and antihypertensive medication (AHM) and the impact of neighborhood social participation on individual use of these medications. They attempted to disentangle contextual from individual influences. Multilevel logistic regression modeling was used to analyze data on 15,456 women aged 45–73 years (first level) residing in 95 neighborhoods (second level) of the city of Malmö, Sweden (250,000 inhabitants) who participated in the Malmö Diet and Cancer Study (1991–1996). AHM use was studied among 7,558 participants with defined hypertension. Of the total variability in medication use in this population, only 1.7% (HRT) and 0.5% (AHM) was between neighborhoods. After adjustment for age, individual socioeconomic factors, individual low levels of social participation, and health and behavioral variables, no neighborhood effect on AHM use was found. However, women living in neighborhoods with low social participation were much less likely to use HRT (odds ratio = 0.36, 95% confidence interval: 0.21, 0.63), especially if they themselves experienced low social participation (synergy index, 1.53) or were immigrants (synergy index, 1.68). The Malmö neighborhoods were homogeneous with regard to HRT and especially AHM use. However, differences in neighborhood social participation affected HRT use independently of individual characteristics.


Research areas and keywords

Subject classification (UKÄ) – MANDATORY

  • Public Health, Global Health, Social Medicine and Epidemiology


  • Prospective Studies, Pharmacoepidemiology, Middle Aged, Human, Female, Estrogen Replacement Therapy: utilization, Emigration and Immigration, Cohort Studies, Aged, Antihypertensive Agents: supply & distribution, Patient Acceptance of Health Care: statistics & numerical data, Age Distribution, Regression Analysis, Residence Characteristics: statistics & numerical data, Social Environment, Support, Non-U.S. Gov't, Sweden: epidemiology, Women's Health
Original languageEnglish
Pages (from-to)774-783
JournalAmerican Journal of Epidemiology
Issue number9
Publication statusPublished - 2003
Publication categoryResearch