Table 1.
Participant characteristics
| Serodiscordant couples cohorts | FSW cohort | |||
|---|---|---|---|---|
| HIV seroconverted (N = 245) | Controls (N = 713) | HIV seroconverted (N = 330) | Controls (N = 962) | |
| Age a | ||||
| 16 to 24 | 52 (21%) | 91 (13%) | 72 (22%) | 205 (21%) |
| 25 to 34 | 120 (50%) | 327 (46%) | 176 (53%) | 453 (47%) |
| ≥35 | 73 (30%) | 295 (41%) | 82 (25%) | 304 (32%) |
| Sex | ||||
| Female | 128 (52%) | 266 (37%) | 330 (100%) | 962 (100%) |
| Male | 117 (48%) | 447 (63%) | – | – |
| Educationb | ||||
| <9 years | 167 (68%) | 437 (61%) | 212 (64%) | 607 (63%) |
| ≥9 years | 78 (32%) | 276 (39%) | 118 (36%) | 355 (37%) |
| Married c | ||||
| Yes | 238 (97%) | 695 (97%) | 176 (53%) | 512 (53%) |
| No | 7 (3%) | 18 (3%) | 154 (47%) | 450 (47%) |
| Enrolment locationd | ||||
| Kenya | 122 (50%) | 381 (53%) | 330 (100%) | 962 (100%) |
| Uganda | 123 (50%) | 332 (47%) | – | – |
| Any unprotected sex e | ||||
| Yes | 72 (30%) | 139 (20%) | 162 (49%) | 447 (46%) |
| No | 172 (70%) | 572 (80%) | 168 (51%) | 515 (54%) |
| Number of sex partners e | ||||
| ≤1 | 219 (91%) | 626 (90%) | 259 (78%) | 729 (76%) |
| >1 | 22 (9%) | 72 (10%) | 71 (22%) | 233 (24%) |
| Sexually transmitted infections f | ||||
| Yes | 27 (11%) | 57 (8%) | 56 (17%) | 77 (9%) |
| No | 218 (89%) | 656 (92%) | 270 (83%) | 769 (91%) |
| Contraceptive use, females only | ||||
| None | 81 (63%) | 153 (58%) | 176 (53%) | 629 (65%) |
| IUD/surgical | 4 (3%) | 17 (6%) | 9 (3%) | 49 (5%) |
| Implant/injectable | 34 (27%) | 75 (28%) | 102 (31%) | 192 (20%) |
| Oral contraceptive | 9 (7%) | 21 (8%) | 43 (13%) | 92 (10%) |
| Serodiscordant couples cohort | ||||
| Partners HSV/HIV transmission study | 94 (38%) | 262 (37%) | – | – |
| Couples observational study | 13 (5%) | 39 (5%) | – | – |
| Partners PrEP study | 138 (56%) | 412 (58%) | – | – |
| Workplace | ||||
| Nightclub | – | – | 41 (12%) | 249 (26%) |
| Bar/other | – | – | 289 (88%) | 713 (74%) |
FSW, female sex worker; IUD, intrauterine device; PrEP, pre‐exposure prophylaxis.
For the serodiscordant couples cohorts, age at enrolment was assessed since the longest study duration was three years. For the FSW cohort, age was time‐varying
years of education at time of cohort enrolment
for the serodiscordant couples cohorts, marital status at the time of study enrolment was assessed. For the FSW cohort, marital status at enrolment was categorized as ever married vs. never married because few participants were married (18/1292)
for the serodiscordant couples cohorts, cases and controls were enrolled at four sites in Kenya [Kisumu (n = 192), Nairobi (n = 119), Eldoret (n = 100), and Thika (n = 92)) and five sites in Uganda (Kampala (n = 210), Tororo (n = 79), Mbale (n = 71), Kabwohe (n = 59) and Jinja (n = 36)]. Enrolment for the FSW cohort was done in Mombasa, Kenya
for the serodiscordant couples cohorts, sexual behaviours were assessed over the prior month. Some individuals had missing values for unprotected sex (n = 3) and number of sexual partners (n = 19). For the FSW cohort, average sexual behaviours were calculated for each year of cohort follow‐up. For both cohorts, sexual behaviours were assessed at all study visits and was time‐varying
for the serodiscordant couples cohorts, testing for sexually transmitted infections (trichomoniasis, gonorrhoea and chlamydia) was done at enrolment. For the FSW cohort, sexually transmitted infection testing (trichomoniasis and gonorrhoea) occurred at each study visit and was time‐varying, and some individuals lacked test results (n = 116).