We also found that particular socioeconomic

We also found that particular socioeconomic and demographic subgroups in Ontario may have a heightened potential to discriminate against the poor because of bad teeth, including men, older people, those with higher levels of education, and those who identify with conservative politics. Further, an income gradient in this neuropeptides outcome appears to suggest that respondents living in higher income households may be more likely to display discriminatory attitudes towards the poor. These findings may be based upon a lack of lived experience by some respondents, or may be influenced by social norms (Lofters, Slater, Kirst, Shankardass & Quiñonez, 2014). Our study supports several others in suggesting that the poor face significant social barriers because of bad teeth (Bedos et al., 2009; Hyde, Satariano & Weintraub, 2006; Gift, Reisine & Larach, 1992; Hollister & Weintraub, 1993). Therefore, providing dental care to socially marginalized groups can create a context that mitigates things like discrimination. Given this, an expansion of those dental care programs that are targeted at the poor may be warranted – either by extending these programs to more of society’s poorest members (such as the working poor or seniors on fixed incomes), or by increasing the range of benefits that are offered to existing recipients. More importantly, considering that oral health is a product of a wide range of social and economic factors suggests a broader policy approach that focuses on providing solutions that address the social determinants of health. The study has several limitations, which have been discussed previously (Shankardass et al., 2012), and include: the potential under-sampling of those from lower socioeconomic strata as a result of preferences for cellular telephones over conventional landlines; the absence of respondents who were able to only speak a language other than English; and a low response rate. With that said, given our quota sampling, sample weighting, and annual household incomes and education levels that closely paralleled those observed in the general population, cell-mediated immunity is arguable that are sample is representative of the Ontario population. Respondents may have also displayed social desirability biases, in which their responses did not reflect their true opinions because they believed that others would unfavourably interpret those responses. Nevertheless, this just means that it is possible that our results underestimate the true degree to which respondents may demonstrate a potential to discriminate against the poor on the basis of dental appearance (Hyde et al., 2006). Additional limitations exist with respect to particular information that was not collected through the survey. Of note, the survey did not gather data related to participants’ clinical health (dental or systemic), which could shape participants’ responses based on their personal experiences. On the other hand, we did control for the age of respondents in our binary logistic regression model, which may, at least in part, account for respondents’ dental experience. We also explored household income as a variable in the binary logistic regression model, which is the strongest predictor of access to and utilization of dental care services among Canadians (Locker, Maggirias & Quinõnez, 2011). In addition, we had limited information related to the type of dental insurance that individuals had for the purpose of this analysis. Again, however, we may consider household income to serve as an effective proxy for dental insurance, given that there is a strong correlation between a person’s income and their insurance coverage in Ontario (Leake, 2006). Finally, the statements presented to participants framed around blaming the poor for differences in health were not tested for their validity. However, their use has been supported by health inequities literature (Niederdeppe, Bu, Borah, Kindig & Robert, 2008), and has been discussed in previous publications (Shankardass et al., 2012).