Neighborhood clustering of non-communicable diseases: results from a community-based study in Northern Tanzania

BACKGROUND In order to begin to address the burden of non-communicable diseases (NCDs) in sub-Saharan Africa, high quality community-based epidemiological studies from the region are urgently needed. Cluster-designed sampling methods may be most efficient, but designing such studies requires assumptions about the clustering of the outcomes of interest. Currently, few studies from Sub-Saharan Africa have been published that describe the clustering of NCDs. Therefore, we report the neighborhood clustering of several NCDs from a community-based study in Northern Tanzania. METHODS We conducted a cluster-designed cross-sectional household survey between January and June 2014. We used a three-stage cluster probability sampling method to select thirty-seven sampling areas from twenty-nine neighborhood clusters, stratified by urban and rural. Households were then randomly selected from each of the sampling areas, and eligible participants were tested for chronic kidney disease (CKD), glucose impairment including diabetes, hypertension, and obesity as part of the CKD-AFRiKA study. We used linear mixed models to explore clustering across each of the samplings units, and we estimated absolute-agreement intra-cluster correlation (ICC) coefficients (ρ) for the neighborhood clusters. RESULTS We enrolled 481 participants from 346 urban and rural households. Neighborhood cluster sizes ranged from 6 to 49 participants (median: 13.0; 25th-75th percentiles: 9-21). Clustering varied across neighborhoods and differed by urban or rural setting. Among NCDs, hypertension (ρ = 0.075) exhibited the strongest clustering within neighborhoods followed by CKD (ρ = 0.440), obesity (ρ = 0.040), and glucose impairment (ρ = 0.039). CONCLUSION The neighborhood clustering was substantial enough to contribute to a design effect for NCD outcomes including hypertension, CKD, obesity, and glucose impairment, and it may also highlight NCD risk factors that vary by setting. These results may help inform the design of future community-based studies or randomized controlled trials examining NCDs in the region particularly those that use cluster-sampling methods.