| Popis: |
Introduction: Chronic kidney disease of uncertain etiology (CKDu) has been found at high frequency in several lowland agricultural areas including Meso-America and Sri Lanka. Whether CKDu also occurs in other countries with large agricultural populations remains uncertain, primarily due to lack of systematic data on kidney function. Hemoglobin (Hgb) levels could be an ancillary marker for presence of kidney dysfunction. Our goal is to estimate the causal effect of agricultural work on Hgb level in men. A causal effect may indicate the presence of CKDu. Methods: We use Demographic and Health Surveys (DHS) data from seven African and Asian countries (Ethiopia, Lesotho, Namibia, Senegal, Uganda, Zimbabwe, and India). We use optimal full matching and permutation inference to estimate the causal effect of agricultural work on altitude-adjusted Hgb levels after adjusting for six known confounders. To assess potential bias due to unmeasured socioeconomic differences, we use multiple control groups that differ in non-agricultural occupation. We then conduct sensitivity analyses to assess the robustness of our causal conclusions to unmeasured confounding. Results: Data were available for 41,180 agricultural workers and 55,705 non-agricultural workers in seven countries. On average, Hgb levels were 0.09 g/dL (95% CI [-0.12, -0.07]) lower among agricultural workers compared to matched controls. Significant effects were observed in Ethiopia, India, Lesotho, Senegal, and Uganda, with effects from 0.10 to 0.32 g/dL lower hemoglobin among agricultural workers. The effect was not attenuated in sensitivity analyses involving both better-off and worse-off occupational controls. Discussion: We find evidence that men engaged in agricultural work in five of the seven countries studied have modestly lower Hgb levels compared with other men living in rural areas. Since underlying kidney disease could be a potential explanation for this finding, our data support consideration to integrating kidney function assessments within DHS surveys and other population-based surveys. |
