A Cross-sectional sample of households that were involved in farming of orange-fleshed sweet potato (OFSP) and
had 2–6 year old children were purposively selected from four divisions of Kampala Capital City Authority (Kawempe,
Rubaga, Makindye and Nakawa) to participate in a controlled, cohort intervention. Respondents in Kawempe division had
received training in production of OFSP and nutrition education; in Rubaga division respondents had only training in
production of OFSP; in Nakawa division they only had nutrition education while in Makindye division the respondents did not
training in nutrition education and production of OFSP and thus served as the control. A total of 457 households were involved
in the study and a coded questionnaire was used to collect data on caregiver’s knowledge, attitudes and practices of child
nutrition after the 12 month intervention. Chi-square tests were used to test for significant relationships (p=0.05) amongst
variables of interest. Results of the study show that majority of the children (2-6 years) take their meals from that of the family.
At least 20% of the respondents serve lunch and dinner separately for the child with no significant differences among the
divisions. Thus any major nutritional interventions should target the caregivers’ knowledge, attitudes and practices of child
feeding and the first meal the child takes. On preparation of meat, 71% of the respondents who received nutrition education
gave correct advice compared to 28% who did not receive the intervention. With regard to Maize porridge and beans
preparation, significant differences could not be attributed to the intervention. However, 40% of the respondents in Nakawa
and Makindye gave advice that was not helpful and so could not be adequately followed in the correct preparation of maize
porridge and beans for the children. Preparation of kitoobero, a highly nutritious complementary dish for weaning children,
was taught during the nutrition education sessions. Significantly more respondents who received nutrition education identified
at least two combinations of foods used for kitoobero (65%) compared to about 5% of those who did not receive the
intervention. The study found that majority of the respondents, 65% and 80% respectively in Kawempe and Rubaga compared
to over 85% in Nakawa and Makindye had not prepared kitoobero seven days prior to the study. The results show that nutrition
education significantly impacts on knowledge of meal preparation for weaning children, attitudes, and practices of child
feeding. It is recommended that nutrition education as a strategy of improving child feeding knowledge and practices should be
taken up by the public health sector so as to boost the fight against malnutrition which is a major problem in Uganda.