Baseline characteristics among the 215 case–control pairs (matched on gender and village and adjusted for age)
| Exposure | Cases (215) | Controls (215) | P value |
| n (%) | n (%) | ||
| Married (yes) * | 154 (72) | 143 (67) | 0.215 |
| Head of household (yes) † | 146 (68) | 140 (65) | 0.441 |
| Education status ‡ | |||
| None | 60 (28) | 48 (22) | 0.148 |
| Primary | 110 (51) | 114 (53) | |
| Secondary | 31 (14) | 32 (15) | |
| Tertiary | 14 (7) | 21 (10) | |
| Farmer (yes) ‡ | 157 (73) | 168 (78) | 0.144 |
| Size of the household § | |||
| Small (1–4 people) | 50 (23) | 109 (51) | 0.05 |
| Medium (5–10 people) | 115 (54) | 94 (44) | |
| Large (>11 people) | 50 (23) | 12 (5) | |
| Self-reported wealth status ¶ | |||
| Poor | 36 (17) | 20 (9) | 0.003 |
| Middle | 158 (73) | 188 (88) | |
| Upper | 21 (10) | 6 (3) |
*People who were married, or cohabiting were considered as married while those who were divorced, single or widowed were considered as not married.
†Being head of the household meant people who were responsible for the overall care of the family, this was regardless of gender: among the cases and controls, 31% and 23% were female heads of households, respectively.
‡Majority of the participants had no or minimal education (primary level) which is not uncommon for a predominantly rural population in Uganda. Subsistence farming is the main occupation for this population.
§Majority of the household sizes were medium to large (five people or more). This is not uncommon since most of the living in rural Uganda is largely in an extended family setting.
¶Self-reported wealth status was classified as poor (1, ‘very poor’ and 2 ‘poor’), middle (3, ‘neither poor nor rich’) upper (4, ’rich’ and 5, ‘very rich’). There was one missing value among the control group. Participants were asked to compare themselves to their neighbours and give a score of their economic status.