Table 4.
Adjusted odds ratio* in a population of HIV+ French Guyanese patients
| Variable | HTLV1+ N (%) | HTLV1− N (%) | Adjusted odds ratio* (95% confidence interval) | P |
|---|---|---|---|---|
| Sex | ||||
| Male | 28 (36.84) | 638 (47.86) | 1 | |
| Female | 48 (63.16) | 695 (52.14) | 1.91 (1.13–3.24) | 0.02 |
| Age | ||||
| < 31 years | 6 (7.89) | 191 (14.33) | 1 | |
| 31–40 years | 9 (11.84) | 383 (28.73) | 0.96 (0.33–2.80) | 0.944 |
| 41–50 years | 23 (30.26) | 347 (26.03) | 3.64 (1.41–9.40) | 0.008 |
| 51–60 years | 16 (21.05) | 273 (20.48) | 4.88 (1.74–13.67) | 0.003 |
| > 60 years | 22 (28.95) | 139 (10.43) | 14.83 (5.30–41.5) | < 0.001 |
| Nationality | ||||
| French | 17 (22.37) | 346 (25.96) | 1 | |
| Brazilian | 2 (2.63) | 83 (6.23) | 0.47 (0.1–2.14) | 0.326 |
| Guyanian | 7 (9.21) | 123 (9.23) | 1.48 (0.58–3.82) | 0.415 |
| Haitian | 17 (22.37) | 397 (29.78) | 0.69 (0.33–1.42) | 0.311 |
| Caribbean but Haitian | 3 (3.95) | 23 (1.73) | 1.94 (0.47–7.99) | 0.356 |
| Others | 2 (2.63) | 23 (1.73) | 2.12 (0.42–10.62) | 0.360 |
| Surinamese | 19 (25) | 199 (14.93) | 2.65 (1.25–5.61) | 0.011 |
| Unknown | 9 (11.84) | 139 (10.43) | 1.21 (0.47–2.7) | 0.798 |
| Viral Load | ||||
| > 30,000 copies/mL | 12 (15.79) | 152 (11.4) | 2.56 (1.20–5.45) | 0.015 |
| 10,000–30000 | 8 (10.53) | 46 (3.45) | 5.68 (2.22–14.55) | < 0.001 |
| 1,000–9,999 | 11 (14.47) | 100 (7.5) | 3.27 (1.48–7.24) | 0.003 |
| 400–999 | 2 (2.63) | 53 (3.98) | 1.05 (0.23–4.74) | 0.946 |
| 50–399 | 8 (10.53) | 179 (13.43) | 1.17 (0.51–2.69) | 0.718 |
| < 50 | 28 (36.84) | 640 (48.01) | 1 | |
| Unknown | 7 (9.21) | 163 (12.23) | 1.62 (0.67–3.95) | 0.284 |
| CD4 count at the time of diagnosis | ||||
| < 200 CD4/mm3 | 21 (27.63) | 230 (17.25) | 1 | |
| 200–349 | 15 (19.74) | 196 (14.7) | 2.5 (0.98–5.18) | 0.057 |
| 350–499 | 7 (9.21) | 181 (13.58) | 0.92 (0.33–2.56) | 0.871 |
| > 500 | 11 (14.47) | 295 (22.13) | 2.34 (1.04–5.25) | 0.040 |
| Unknown | 22 (28.95) | 431 (32.33) | 1.58 (0.72–3.45) | 0.253 |
*
Adjusted using a logistic regression model. The saturated model included sex, age, nationality, viral load, CD4 count at the time of diagnosis, treatment, cardiopathies, diabetes, nadir CD4, and aspartate amino transferase (ASAT) as explanatory variables, the most parsimonious model was obtained using the Likelihood ratio test.