Table 3.
Multivariate linear and logistic regression models predicting reduced severity and reduced susceptibility optimism among young adults in Balaka, Malawi (N=1745)
| Severity scale | Susceptibility | |||||||
|---|---|---|---|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | |||||
| β | β | OR | OR | |||||
| Gender | ||||||||
| Male (ref) | 0 | 0 | 1 | 1 | ||||
| Female | 0.637 | *** | 0.655 | *** | 1.949 | *** | 1.908 | *** |
| Age | 0.072 | ** | 0.066 | * | 1.057 | * | 1.051 | * |
| Years of education | −0.288 | *** | −0.281 | *** | 0.898 | *** | 0.898 | *** |
| Household goods | −0.060 | −0.051 | 0.984 | 0.988 | ||||
| Income quartiles | ||||||||
| 1 lowest (ref) | 0 | 0 | 1 | 1 | ||||
| 2 | −0.060 | −0.013 | 0.934 | 0.944 | ||||
| 3 | −0.150 | −0.07 | 1.244 | 1.275 | ||||
| 4 highest | −0.435 | * | −0.408 | * | 0.901 | 0.920 | ||
| Marital status | ||||||||
| Married (ref) | 0 | 0 | 1 | 1 | ||||
| Never married | −0.466 | * | −0.468 | * | 0.934 | 0.924 | ||
| Formerly married | 0.156 | 0.061 | 0.977 | 0.910 | ||||
| HIV status and ART use | ||||||||
| HIV negative or unknown status (ref) | 0 | 1 | ||||||
| HIV positive and not on ART | 0.226 | 1.449 | ||||||
| HIV positive and on ART | 1.909 | ** | 2.844 | * | ||||
| Exposure to people on ART | ||||||||
| Do not know anyone on ART (ref) | 0 | 1 | ||||||
| Know someone on ART but do not discuss | −0.539 | *** | 0.925 | |||||
| Know someone on ART and regularly discuss | −0.040 | 0.962 | ||||||
| Constant | 7.436 | *** | 7.684 | *** | 0.138 | *** | 0.162 | ** |
Models are weighted to account for the gender stratified sampling design
*
p<0.05,
**
p<0.01,
***
p<0.001