Table 4

Predicted adjusted population attributable risk fractions for key predictors across Kenya

 Population attributable risk fractions by province, % (95% BCI)
CentralCoastEasternNyanzaRift ValleyWestern
14-year-old girls
Observed prevalence of anaemia3.337.334.730.819.425.0
PAF for
 Med/high-intensity hookwormn/a13.0 (3.7, 23.1)5.9 (0, 12.5)8.2 (0, 30.0)5.0 (0, 12.5)29.0 (1.4, 55.1)
 Malarian/a5.8 (2.0, 10.4)n/a27.6 (0, 49.1)n/a23.6 (0, 40.9)
 Schistosomiasisn/a3.1 (0, 0.08)n/an/an/an/a
 Acute malnutrition*n/a18.5 (0, 37.9)18.4 (0, 73.0)n/a16.0 (0, 52.5)n/a
7-year-old boys
Observed prevalence of anaemia5.043.323.243.222.419.2
PAF for
 Med/high-intensity hookwormn/a5.6 (1.0, 10.3)n/a11.0 (0, 21.6)3.4 (0, 6.6)13.5 (0, 27.5)
 Malarian/a1.5 (0.3, 6.1)n/a14.3 (0, 32.0)n/a15.8 (0, 28.1)
 Schistosomiasisn/a0 (0, 5.5)n/an/an/an/a
 Acute malnutrition*n/a16.4 (0, 33.5)16.4 (0, 67.1)n/a14.1 (0, 45.9)n/a
  • Estimates are based on the predicted numbers anaemic in each risk group given by the mean and individual-level SD at each realisation of the Bayesian multivariate model for Hb, and are shown for two demographic groups by Province.

  • N/a, no children are exposed (ie, prevalence of predictor is zero in this demographic group).

  • *Exposure to ‘acute malnutrition’, prevalence of wasting in under 5s exceeds 5%.