Links across disabilities: unveiling associations between functional domains

Background Persons with disabilities experience higher risks of mortality as well as poorer health as compared to the general population. The aim of this study is to estimate the correlations between functional difficulties across several domains in six countries. Methods National census data with questions on disability from six countries (Mauritius, Morocco, Senegal, Myanmar, Vietnam, and Uruguay) was used in this study. We performed logistic regressions to assess the extent to which having a functional difficulty in one domain is correlated with having a functional difficulty in each of the other domains and report weighted odds ratios (ORs) overall and within age-groups (‘18–44’ years and ‘45+’ years). Models adjust for age, sex, and location (rural or urban). Sensitivity analyses around different choices of predictors and response variables were conducted. Findings For all countries, reporting a functional difficulty in one domain was consistently and significantly positively correlated with reporting a functional difficulty in other domains (overall) and for each of the two age-groups considered - ‘18–44’ years and ‘45+’ years. All ORs were greater than one. Cognition, mobility, and hearing were the domains that were the most correlated ones with other domains. The highest pairwise correlations were for i/ hearing and cognition; ii/ mobility and cognition. Results were robust to changing the severity thresholds for functional difficulties. Across countries, Uruguay, the only high-income country among the six countries under study, had the lowest correlations between functional domains. Conclusions There are consistent positive associations in the experience of functional difficulties in various domains in the six countries under study. Such correlations may reflect barriers to social services including healthcare services and resources (e.g. assistive devices) that may lead to an avoidable deterioration of functioning across domains. Further research is needed on the trajectories of functional difficulties and on structural barriers that people with functional difficulties may experience in their communities and in healthcare settings in particular. This is important as some functional difficulties may be preventable. Supplementary Information The online version contains supplementary material available at 10.1186/s12889-023-17523-5.


Table of Contents
Table S1: An overview of the datasets used in the analysis.Table S2: Main analysis (adjusted estimates): Age-specific odds ratios in the presence of all control variables Table S3: Sensitivity analysis (adjusted estimates) -first threshold: Overall odds ratios in the presence of all control variables Table S4: Sensitivity analysis (adjusted estimates) -second threshold: Overall odds ratios in the presence of all control variables Table S5: Main analysis (crude estimates): Odds ratios of having a difficulty in a given domain (e.g. hearing) given a difficulty in another domain (e.g.seeing) Table S6: Main analysis (crude estimates): Odds ratios of having a difficulty in a given domain (e.g. hearing) given a difficulty in another domain (communication or self-care) Table S7: Main analysis (crude estimates): Age-specific odds ratios in the presence of all control variables Table S8: Sensitivity analysis (crude estimates) -first threshold: Overall odds ratios in the presence of all control variables Table S9: Sensitivity analysis (crude estimates) -second threshold: Overall odds ratios in the presence of all control variables ,    =    +  +  +  where = one among the hearing, seeing, mobility, cognition, communication, and self-care  domains.
For instance, if the aim is to understand how having functional difficulties ('difficulty') in the hearing-domain could be correlated with having functional difficulties in the seeing-and mobility-domains, then the resulting regression equations look like: , _ = _ℎ +  +  +  _ = _ℎ +  +  + , .

Sensitivity analyses:
Country-specific estimates: -First threshold: A functional difficulty in a given domain was assigned 0 if people responded with either "no difficulty" or "some difficulty" and 1 if people responded with either "a lot of difficulty" or "unable to do" to the disability questions asked.
Table S3: Sensitivity analysis (adjusted estimates): Overall odds ratios in the presence of all control variables, along with their 95% CIs.-Second threshold: A functional difficulty in a given domain was assigned 0 if people responded with "no difficulty" or "some difficulty" or "a lot of difficulty" and 1 if people responded with "unable to do" to the disability questions asked.
Table S4: Sensitivity analysis (adjusted estimates): Overall odds ratios in the presence of all control variables, along with their 95% CIs.All the p-values were <0.001 corresponding to each odd ratio implying statistically significant results at the 5% threshold.988060, 23983300, 7288742, 35894738, 64267057, 2478630, were

Age-specific ORs estimates (crude):
Table S7: Main analysis (crude estimates): Age-specific odds ratios in the presence of all control variables, along with their 95% CIs.All the p-values were <0.001 corresponding to each odd ratio implying statistically significant results at the 5% threshold.388.9 (349.6, 432.6) 191.3 (176.4, 207.3)Self-care 168.9 (151.8,188) - -168.9 (151.8, 188) -First threshold: A functional difficulty in a given domain was assigned 0 if people responded with either "no difficulty" or "some difficulty" and 1 if people responded with either "a lot of difficulty" or "unable to do" to the disability questions asked.

Table S1 :
An overview of the datasets used in the analysis.

Table S2 :
Main analysis (adjusted estimates): Age-specific odds ratios in the presence of all control variables, along with their 95% CIs.All the p-values were <0.001 corresponding to each odd ratio implying statistically significant results at the 5% threshold.

1. 45+ age-group:
2. '18-44' age-groupORs corresponding to the four core functional domains among all countries included (n = 6):ORs corresponding to the two functional domains (communication and self-care) inMauritius,  Morocco, and Senegal: All the p-values were <0.001 corresponding to each odd ratio implying statistically significant results at the 5% threshold.988060, 23983300, 7288742, 35894738, 64267057, 2478630, were the cell counts used in the computation of ORs for Mauritius, Morocco, Senegal, Myanmar, Vietnam, and Uruguay respectively.

Table S5 :
the cell counts used in the computation of ORs for Mauritius, Morocco, Senegal, Myanmar, Vietnam, and Uruguay respectively.Main analysis (crude estimates): Odds ratios of having a difficulty in a given domain (e.g.hearing) given a difficulty in another domain (e.g.seeing), along with their 95% CIs.All the p-values were <0.001 corresponding to each odd ratio implying statistically significant results at the 5% threshold.988060,23983300,7288742, 35894738, 64267057, 2478630, were the cell counts used in the computation of ORs for Mauritius, Morocco, Senegal, Myanmar, Vietnam, and Uruguay respectively.
ORs corresponding to the four core functional domains among all countries included (n = 6):

Table S6 :
Main analysis (crude estimates): Odds ratios of having a difficulty in a given domain (e.g.hearing) given a difficulty in another domain (communication or self-care), along with their 95% CIs.Only Mauritius, Morocco, and Senegal, have data available for communication and self-care domains.All the p-values were <0.001 corresponding to each odd ratio implying statistically significant results at the 5% threshold.988060,23983300,7288742, 35894738,  64267057, 2478630, were the cell counts used in the computation of ORs for Mauritius, Morocco, Senegal, Myanmar, Vietnam, and Uruguay respectively.
ORs corresponding to the two functional domains (communication and self-care) inMauritius,  Morocco, and Senegal: the cell counts used in the computation of ORs for Mauritius, Morocco, Senegal, Myanmar, Vietnam, and Uruguay respectively.ORs corresponding to the four core functional domains among all countries included (n = 6):