Table 3.
Country-level linear regression analyses of the predictors of the average age of PWID and average age of onset of injecting
|
Average age of PWID |
Average age of onset of injecting |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Univariable analysis (n=94) |
Multivariable analysis (n=79) |
Univariable analysis (n=66) |
Multivariable analysis (n=61) |
|||||||
| n | Coefficient (95% CI) | p value | Coefficient (95% CI) | p value | n | Coefficient (95% CI) | p value | Coefficient (95% CI) | p value | |
| Urban population growth | 92 | −0·52 (−1·18 to 0·15) | 0·129 | .. | .. | 64 | 1·20 (0·62 to 1·78) | 0·0001 | 1·40 (0·41 to 2·40) | 0·007 |
| Youth unemployment | 91 | −0·01 (−0·09 to 0·08) | 0·887 | .. | .. | 64 | −0·09 (−0·16 to −0·01) | 0·032 | −0·05 (−0·14 to 0·04) | 0·257 |
| GDP (per US$1000 increase) | 89 | 0·12 (0·07 to 0·17) | <0·0001 | 0·13 (0·06 to 0·20) | 0·0004 | 63 | −0·01 (−0·07 to 0·04) | 0·631 | .. | .. |
| Gini coefficient (per score increase) | 87 | −0·06 (−0·23 to 0·11) | 0·496 | .. | .. | 63 | 0·05 (−0·12 to 0·22) | 0·539 | .. | .. |
| Opioid substitution therapy coverage (per % increase in coverage) | 76 | 0·03 (−0·00 to 0·06) | 0·078 | .. | .. | 54 | −0·02 (−0·06 to 0·01) | 0·154 | .. | .. |
| Percentage of PWID who are women | 83 | 0·06 (−0·05 to 0·17) | 0·276 | −0·07 (−0·22 to 0·07) | 0·293 | 63 | −0·13 (−0·22 to −0·04) | 0·005 | −0·14 (−0·26 to −0·02) | 0·019 |
| Proportion of the general population aged 15–24 years | 90 | −0·32 (−0·63 to 0·02) | 0·038 | −0·06 (−0·44 to 0·31) | 0·731 | 64 | 0·36 (0·09 to 0·64) | 0·011 | −0·45 (−0·96 to 0·06) | 0·082 |
n refers to number of countries. For age of PWID and age of onset of IDU, we used a median (range) where available and a mean (range) where a median was not available but a mean was; hence, we refer to these as averages throughout. GDP=gross domestic product. PWID=people who inject drugs.