Is the association between social support and oral health modified by household income? Findings from a national study of adults in Australia.

Abstract

To investigate whether the association between social support and oral health outcomes is modified by levels of household income. Data were from the National Study of Adults Oral Health (NSAOH 2004-06), a nationally representative study comprising n=3619 adults in Australia. Effect measure modification (EMM) analysis was adopted to verify whether the association between social support and poor/fair self-rated oral health, lack of a functional dentition (<21 teeth) and low Oral Health Related Quality of Life (OHRQoL; measured using OHIP-14) varies according to levels of income. Poisson regressions adjusted for age, sex, education, country of birth, main language spoken at home and remoteness were used to estimate prevalence ratios (PR) for oral health outcomes for each stratum of social support (overall, family, friends and significant other) and income (effect modifier). We then computed the Relative Excess Risk due to Interaction (RERI), which represents the risk that is over what would be expected if the combination of low social support and low income was entirely additive. Sensitivity analyses for different cut-offs of household income were performed. Adults with lower levels of social support had a 2.1, 1.2 and 1.9 times higher prevalence of fair/poor self-rated oral health, <21 teeth and poor OHRQoL respectively. The RERIs observed were 0.98 (95% CI -0.01; 1.96) for poor/fair self-rated oral health; 0.52 (95% CI -0.06; 1.10) for lack of a functional dentition and 0.50 (95% CI -0.16; 1.15) for poor OHRQoL. For all outcomes and all individual domains of social support, the positive RERIs indicated that the joint association of low social support and low household income surpassed the sum of their separate associations with objective and subjective oral health indicators. Individuals with lower levels of social support had poorer oral health than those with high levels of social support, although this association was small for the outcome <21 teeth. The association between social support and poor oral health indicators is modified by levels of household income. Hence, the provision of social support had a stronger association with the oral health of low-income participants, suggesting that socioeconomically disadvantaged individuals would mostly benefit from a social support intervention.