Sanitation and Sociocultural Correlates of Selected Dietary Behaviors in Rumah Lanting: Cross-Sectional Evidence

Abstract

Stunting reflects cumulative early-life exposures, whereas current dietary behavior is proximal and time-specific. Sanitation and sociocultural exposures may coexist within river-dependent rumah lanting communities. This study examined associations of sociocultural conditions and environmental sanitation with selected dietary behaviors and, secondarily, with quality-controlled height-for-age z-scores (HAZ). The cross-sectional analysis included 193 children aged 3–59 months from Pahandut and Pahandut Seberang, Palangka Raya. The selected dietary-behavior construct comprised AL2, AMS1, and AMS2. PLS-SEM evaluated sociocultural and sanitation associations with this construct using 5,000 bootstrap resamples. HAZ was independently recalculated using sex, completed age in months, and recorded length/height according to the 2006 WHO Child Growth Standards. Implausible HAZ values flagged by WHO criteria (<−6 or >6) were excluded from HAZ analyses, leaving 188 children. Sociocultural factors (β=−0.210, 95% CI −0.375 to −0.064; p=0.008) and environmental sanitation (β=−0.171, 95% CI −0.331 to −0.022; p=0.032) were associated with selected dietary behaviors, jointly explaining 9.1% of the variance (Q²=0.019). In an age-adjusted composite-score sensitivity analysis, the sociocultural association remained statistically significant (β=−0.200, p=0.027), whereas the sanitation estimate was attenuated (β=−0.124, p=0.095). In the quality-controlled HAZ analysis, none of the three constructs was significantly associated with HAZ (R²=0.043); associations remained non-significant after age adjustment (R²=0.048). Sociocultural and sanitation conditions were associated with a narrow set of current dietary behaviors in the primary PLS model. However, the sanitation association was less robust after age adjustment, and the measured constructs explained little variation in concurrent HAZ. These cross-sectional findings should not be interpreted causally.