Table 4. Multivariable association of socio-demographic and environmental factors with S. haematobium infection.
| All study sites (N = 17280) | Multivariable non-spatial* | Multivariable spatial§ | ||||||
|---|---|---|---|---|---|---|---|---|
| Covariate | N | % pos. | OR | 95% CI | p value | OR | 95% CI | p value |
| Sex | ||||||||
| Female* | 9194 | 4.95 | 1.00 | - | - | 1.00 | ||
| Male | 8086 | 5.68 | 1.03 | 0.87–1.21 | 0.7588 | 1.02 | 0.88 – 1.20 | 0.7653 |
| Age (years) | ||||||||
| below 5 | 2233 | 2.15 | 1.15 | 0.77–1.74 | 0.4935 | 1.12 | 0.74 – 1.68 | 0.5932 |
| 5–15 | 5534 | 8.98 | 7.82 | 5.85–10.43 | <0.0001 | 7.26 | 5.50 – 9.60 | <0.0001 |
| 15–25 | 3063 | 7.61 | 5.84 | 4.30–7.95 | <0.0001 | 5.51 | 4.09 – 7.42 | <0.0001 |
| 25–35 | 2336 | 2.65 | 1.64 | 1.12–2.40 | 0.0119 | 1.57 | 1.07 – 2.29 | 0.0213 |
| 35 and above* | 4114 | 1.80 | 1.00 | - | - | 1.00 | - | - |
| HIV infection | ||||||||
| No* | 15700 | 5.57 | 1.00 | - | - | 1.00 | - | - |
| Yes | 1209 | 2.07 | 0.64 | 0.40–1.03 | 0.0637 | 0.63 | 0.39 – 1.00 | 0.0524 |
| No information | 371 | 4.04 | 0.70 | 0.36–1.33 | 0.2752 | 0.72 | 0.39 – 1.34 | 0.3021 |
| SES score (per 1 unit) | 0.90 | 0.79–1.02 | 0.1004 | 0.89 | 0.79 – 1.01 | 0.0631 | ||
| Population density (per 1000 persons/km2) | 1.08 | 0.97–1.20 | 0.1559 | 1.27 | 1.10 – 1.46 | 0.0008 | ||
| Elevation (per 100 m) | 0.74 | 0.64–0.84 | <0.0001 | 0.69 | 0.46 – 1.04 | 0.0750 | ||
| EVI minimum (per 0.1 units) | 0.21 | 0.11–0.39 | <0.0001 | 0.15 | 0.07 – 0.33 | <0.0001 | ||
| Distance to water course (in km) | 1.37 | 1.20–1.57 | <0.0001 | 1.17 | 1.03 – 1.34 | 0.0186 | ||
| Distance to Lake Nyasa (in km) | ||||||||
| below 1 | 195 | 9.23 | 4.53 | 1.80–11.38 | 0.0013 | 4.96 | 0.91 – 27.08 | 0.0642 |
| 1–2 | 372 | 9.14 | 3.51 | 1.65–7.47 | 0.0011 | 4.11 | 1.06 – 15.94 | 0.0413 |
| 2–4 | 428 | 11.92 | 3.33 | 1.68–6.59 | 0.0006 | 4.31 | 1.56 – 11.93 | 0.0048 |
| 4 and above* | 16285 | 4.98 | 1.00 | - | - | 1.00 | - | - |
*Multivariable mixed effects logistic model with random effects for households and study sites (N = 17,280).
§Multivariable generalized additive mixed model of binomial family with spatially correlated effects base on the location of the observation, with additional random effects for households and study sites (N = 17,280). The outcome for both modelling approaches was the binary S. haematobium infection status. N = number of participants, % pos. = prevalence of S. haematobium infection for categorical variables, OR = prevalence ratio, CI = confidence interval, SES = socio-economic status, EVI = enhanced vegetation index. The reference category for categorical variables is denoted by the asterisk *.