Individual- and City-Level Socioeconomic Factors and Tooth Loss among Elderly People: A Cross-Level Multilevel Analysis

Int J Environ Res Public Health. 2020 Mar 30;17(7):2345. doi: 10.3390/ijerph17072345.

Abstract

This study aimed to test the association of contextual and individual socioeconomic status with tooth loss among Brazilian elderly people aged 65-74 years. Data from 5435 elderly participants from the Brazilian National Oral Health Survey (2010) were linked to city-level data for 27 state capitals and the Federal District. Tooth loss was clinically assessed according to the number of missing natural teeth. Contextual social variables included Human Development Index income (HDI-income) and HDI-education. Individual socioeconomic measures were monthly family income and years of schooling. Covariates included sex, skin colour, number of residents per room and number of goods. Multilevel Negative Binomial regression models were used to estimate rate ratios (RR) and 95% confidence intervals between contextual and individual variables and tooth loss. Contextual and individual income and education measures were consistently associated with tooth loss. Elderly people living in cities with low HDI-income and low HDI-education were respectively 21% and 33% more likely to present tooth loss. Cross-level interaction suggested that the relationship of lower income and lower schooling with tooth loss is different across levels of city-level income and city-level education inequality, respectively. Public policies aiming to reduce the income and education gaps and preventive dental interventions are imperative to tackle tooth loss among elderly people.

Keywords: edentulous; elderly people; multilevel analysis; social inequity; tooth loss.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Brazil / epidemiology
  • Cities
  • Cross-Sectional Studies
  • Female
  • Humans
  • Income
  • Male
  • Multilevel Analysis
  • Oral Health
  • Socioeconomic Factors*
  • Tooth Loss / epidemiology*