| Total | Recurrent, n = 24 (%) | Non-recurrent, n = 72 (%) | p Value* | |
| Gender | ||||
| Male | 63 | 18 (75) | 45 (62.5) | NS |
| Female | 33 | 6 (25) | 27 (37.5) | |
| Ethnicity | ||||
| Black African | 45 | 8 (33.3) | 37 (51.4) | NS |
| Caucasian | 40 | 12 (50) | 28 (38.9) | |
| Black Caribbean | 8 | 4 (16.6) | 4 (5.6) | |
| Other | 3 | 0 | 3 (4.2) | |
| CD4 count | ||||
| <200 cells/ml | 5 | 2 (8.3) | 3 (4.2) | NS |
| 200–350 cells/ml | 19 | 4 (16.7) | 15 (20.8) | |
| >350 cells/ml | 72 | 18 (25) | 54 (75) | |
| Anti-retrovirals | ||||
| Off treatment | 23 | 9 (37.5) | 14 (19.4) | p = 0.073 |
| On treatment | 73 | 15 (62.5) | 58 (80.6) | |
| Syphilis serology | ||||
| New infection** | 9 | 1 (11.1) | 8 (88.9) | NS |
| Negative/previously treated*** | 87 | 232 (6.4) | 64 (73.6) | |
| APOE-ϵ4 allele | ||||
| APOE-ϵ4 present | 48 | 19 (79.2) | 29 (40.3) | p = 0.001 |
| APOE-ϵ4 not present | 48 | 5 (20.8) | 43 (59.7) |
*χ2 or Fisher’s exact test as appropriate; **new infection or >fourfold rise in Venereal Disease Research Laboratory (VDRL) test. Of these nine, one patient had a syphilitic chancre (not classified as an ulcerative episode for study purposes), two had rash only and the rest were asymptomatic; ***73 patients had negative syphilis serology during the 12 months preceding recruitment. 14 patients had positive serology due to previously treated, old infections with no VDRL rise in the period studied.