Table 2.
Analysis of potential risk factors for ocular surface squamous neoplasia adjusted for age group, sex and study centre
| Exposure | Cases (N = 131) | Controls (N = 131) | OR (95% CI) | P value |
|---|---|---|---|---|
| Marital status, No. (%) | 0.37 | |||
| Married | 83 (63.4) | 89 (67.9) | 1 [Reference] | |
| Single | 22 (16.8) | 26 (19.9) | 0.91 (0.48–1.73) | |
| Widowed | 18 (13.7) | 10 (7.6) | 1.93 (0.84–4.45) | |
| Divorced or separated | 8 (6.1) | 6 (4.6) | 1.43 (0.47–4.31) | |
| Highest education level, no. (%) | 0.04 | |||
| Completed secondary or more | 46 (35.1) | 68 (51.9) | 1 [Reference] | |
| Completed primary or some secondary | 53 (40.5) | 39 (29.8) | 1.82 (1.03–3.21) | |
| None or some primary | 32 (24.4) | 24 (18.3) | 1.78 (0.90–3.52) | |
| Residential area (province) | 0.21 | |||
| Nairobi | 28 (21.4) | 42 (32.1) | 1 [Reference] | |
| Rift Valley | 36 (27.5) | 28 (21.4) | 1.93 (0.96–3.88) | |
| Central | 27 (20.6) | 34 (26.0) | 1.19 (0.59–2.40) | |
| Eastern | 13 (9.9) | 9 (6.9) | 2.17 (0.81–5.85) | |
| Nyanza | 14 (10.7) | 7 (5.3) | 3.00 (1.04–8.65) | |
| Western | 10 (7.6) | 10 (7.6) | 1.50 (0.55–4.11) | |
| Coast | 1 (0.8) | 1 (0.8) | 1.50 (0.09–25.52) | |
| North‐eastern | 2 (1.5) | 0 | – | |
| Duration of residence in home province, median (IQR), y | 24 (10–39) | 28 (14–40) | – | 0.10a |
| History of allergic conjunctivitisb | 28 (21.5) | 1 (0.9) | 30.78 (4.05–233.75) | <0.001 |
| Location of current occupation, no. (%)b | 0.02 | |||
| Indoor | 45 (34.9) | 62 (48.1) | 1 [Reference] | |
| Outdoor | 84 (65.1) | 67 (51.9) | 1.73 (1.03–2.91) | |
| Duration in current occupation, median (IQR), y | 10 (5–20) | 15 (7–23) | – | 0.04a |
| Sun exposure per day in current occupation, mean (SD), h | 6.9 (3.8) | 4.6 (3.2) | 1.22 (1.12–1.32) | <0.001c |
| Hat or cap worn outdoors, no. (%)b | 19 (15.0) | 32 (24.8) | 0.60 (0.30–1.20) | 0.07 |
| Sunglasses worn outdoors, no. (%)b | 10 (7.9) | 16 (12.4) | 0.66 (0.28–1.55) | 0.11 |
| Cigarette smoking (past), no. (%)b | 0.10 | |||
| No | 93 (71.5) | 90 (70.3) | 1 [Reference] | |
| Passive (spouse/partner smokes) | 17 (13.1) | 21 (16.4) | 1.09 (0.47–2.53) | |
| Yes | 20 (15.4) | 17 (13.3) | 1.11 (0.54–2.31) | |
| Cigarette smoking (current), no. (%)b | 0.03 | |||
| No | 115 (89.2) | 111 (88.1) | 1 [Reference] | |
| Passive (spouse/partner smokes) | 4 (3.1) | 9 (7.1) | 0.54 (0.15–1.94) | |
| Yes | 10 (7.8) | 6 (4.8) | 1.62 (0.56–4.75) | |
| Duration of cigarette smoking, mean (SD), y | 13.5 (8.6) | 12.7 (10.0) | – | 0.79c |
| No. of cigarettes smoked daily, median (IQR), y | 9 (5–20) | 7 (5–10) | – | 0.60a |
| HIV infectiond, no. (%) | 62 (74.7) | 14 (13.6) | 18.96 (8.87–40.51) | <0.001 |
| HIV infection and ART usee, no. (%) | <0.001 | |||
| HIV− | 21 (16.0) | 89 (67.9) | 1 [Reference] | |
| HIV+/ART− | 17 (13.0) | 2 (1.5) | 38.54 (8.19–181.39) | |
| HIV+/ART+ | 40 (30.5) | 12 (9.2) | 14.30 (6.35–32.20) | |
| No HIV or ART data | 53 (40.5) | 28 (21.4) | 7.18 (3.67–14.05) | |
| CD4 countf, median (IQR), cells/mm3 | 251 (118–670) | 796 (619–1063) | – | <0.001a |
| CD4 WHO category, no.(%), cells/mm3 | <0.001 | |||
| ≥500 | 20 (15.3) | 64 (48.9) | 1 [Reference] | |
| 350–499 | 6 (4.6) | 9 (6.9) | 2.39 (0.74–7.68) | |
| 200–349 | 11 (8.4) | 1 (0.8) | 39.24 (4.68–328.33) | |
| <200 | 22 (16.8) | 2 (1.5) | 37.33 (7.98–174.50) | |
| No CD4 data | 72 (55.0) | 55 (42.0) | 3.84 (2.06–7.17) | |
| Vitamin A deficiency (<30 μg/dl)g, no. (%) | 5 (8.8) | 5 (5.8) | 1.30 (0.32–5.32) | 0.50 |
| Serum retinol, median (IQR), μg/dl | 44.9 (37.6–56.2) | 51.0 (39.7–64.7) | – | 0.03a |
SD, standard deviation; IQR, interquartile range.
a
Wilcoxon–Mann–Whitney U‐test.
b
There were missing data on allergic conjunctivitis (24 participants), occupational location 4, hat or sunglasses use 6 and cigarette smoking 7.
c
t‐test.
d
The denominator here is 83 cases and 103 controls who were tested for HIV.
e
We used the combined HIV/ART variable because the effect of HIV would be related to ART use.
f
There were 59 cases and 76 controls who had a CD4 test.
g
The denominator here is 57 cases and 86 controls who had a vitamin A test.