Table 3.
Best predictors of low functionality level as measured by machine learning methods listed in order (e.g., from the strongest to the weakest)
| Predictors of low functionality level | Type of subfactor and category of predictors of low functionality level |
|---|---|
| Speed of waking | Physical subfactors-PMHC |
| Body mass index | CMF |
| Age | Demographic factors |
| Employment | Occupation subfactor-CSCF |
| Grip force | Physical subfactor-PMHC |
| Poor institutional help | Socio-economic resources-SDH |
| Mental disease antecedents | Mental health subfactor-PMHC |
| Falls | Mental health subfactor-PMHC |
| Smoking antecedents | Lifestyle subfactors-PMHC |
| Moderate institutional help | Socio-economic resources-SDH |
| Widow | Marital status-CSCF |
| Hypertension | CMF |
| Housing in urban area | Socio-economic resources-SDH |
| Fear of falling | Mental health subfactor-PMHC |
| Diabetes | CMF |
| Housing | Socio-economic resources-SDH |
| Stroke | CMF |
| Health problems (las 30 days) | Medical conditions subfactor-PMHC |
| Mild fear of falling | Medical conditions subfactor-PMHC |
| Health problems (last 15 days) | Medical conditions subfactor-PMHC |
| Absence of institutional help | Socio-economic resources-SDH |
| Poor economic resources | Socio-economic resources-SDH |
| High medication’s consumption | Medical conditions-PMHC |
| Auditory functioning | Lifestyle subfactor-PMHC |
| Religious participation | Social participation-SDH |
CMF, cardiometabolic factors; SDH, social determinants of health; PMHC, physical and mental health conditions; CSCF, complementary social context factors