Background
Undernutrition among preschool children in Northern region is the highest in Ghana. However, there is scarcity of data on the factors that determine undernutrition in these children. This study investigated the effect of maternal and child factors on undernutrition among preschool children in Northern Ghana.

Methods
This study was a community based analytical cross-sectional survey on a sample of 425 mother- child pairs drawn from 25 clusters. A semi- structured questionnaire was used to collect data on maternal and child socio-demographic characteristics, feeding practices and anthropometry. Anthropometric indices of Height-for-age Z-scores (HAZ), Weight-for-height Z-scores (WHZ) and Weight-for – age Z-scores (WAZ) were used to classify child stunting, wasting and underweight respectively. Bivariate and multivariate analyses were performed to determine associations between explanatory variables and undernutrition.

Results
The prevalence of stunting, wasting and underweight were 28.2, 9.9 and 19.3% respectively. Multiple logistic regression analysis showed that, the odds of stunting was higher among male children [AOR = 1.99; 95% CI (1.26–3.13); p = 0.003], children of mothers less than 150 cm in height [AOR = 3.87; 95% CI (1.34–11.20); p = 0.01], mothers 155–159 cm tall [AOR = 2.21; 95% CI (1.34–3.66); p = 0.002], and older children aged 12–23 months [AOR 9.81; 95% CI (2.85–33.76); p < 0.001]. Wasting was significantly higher among male children [AOR = 2.40; 95% CI (1.189–4.844); p = 0.015], consumption of less than four food groups [AOR = 3.733; 95% CI (1.889–7.376); p < 0.001] and among children of underweight mothers [AOR = 3.897; 95% CI (1.404–10.820); p = 0.009]. Male children [AOR = 2.685; 95% CI (1.205–5.98); p = 0.016] and having low birth weight [AOR = 3.778; 95% CI (1.440–9.911); p < 0.001] were associated with higher odds of underweight in children.

Conclusion
Maternal height associated negatively with stunting but not wasting. Factors that affect low height –for-age z-score (HAZ) may not necessarily be the same as stunting. Infant and child feeding practices as measured by dietary diversity score associated positively with weight-for-height Z-scores than length-for-age Z-scores of young children. Surprisingly, consumption of some specific food groups including, animal source foods, legumes, staples and eggs were associated with lower HAZ but with increased likelihood of higher WHZ among children 6–59 months.