A school-based assessment was conducted across ten primary schools in Kisarawe District to document the presence and completeness of the school feeding program relative to Tanzania’s National Guideline. All ten schools reported ongoing feeding activities, but none fully adhered to national implementation criteria. Key infrastructural gaps included the absence of designated eating areas or cafeterias in all schools. Only half (50%) had food storage facilities, and 70% reported adequate cooking facilities. Seventy percent of schools provided meals consistently over the preceding academic year, yet only 40% supplied meals containing at least four food groups as recommended. Overall student coverage of the program was 65%, with non-participation linked in part to economic barriers.
The study used a cross-sectional quantitative design conducted in Kisarawe District, a predominantly rural agricultural district in Pwani Region. Data collection occurred from 22–25 April 2025. A multi-stage procedure selected four wards (two urban, two rural) and ten primary schools (five urban, five rural) by convenience sampling. From these schools, simple random sampling selected 300 students aged 10–14 years, enrolled in standards five and six, to minimize classroom disruption and recall bias.
Trained enumerators administered structured questionnaires in Swahili and used a 24-hour dietary recall to record all foods and beverages consumed from waking to bedtime. Data capture used electronic tablets with KOBO Collect. Dietary adequacy was operationalized by two measures:
Dietary diversity: attainment of at least four of seven predefined food groups in the previous 24 hours (grains; legumes and nuts; dairy; vitamin A-rich fruits and vegetables; other vegetables; meat and fish; eggs).
Meal frequency: number of eating occasions in the previous 24 hours, with regular frequency defined as three or more meals.
Data quality measures included pre-testing, daily supervision, and routine completeness checks. Analysis used Stata v15 with descriptive statistics for implementation status and multilevel modified Poisson regression to estimate adjusted prevalence ratios (APR).
Three hundred students participated (mean age 11.5 ± 1.1 years). Parental education was low: 68% of fathers and 72.7% of mothers had no formal education or primary education only. Household food insecurity in the previous month was reported by 14% of households.
Dietary patterns in the 24-hour recall showed near-universal consumption of grains, roots, and tubers (98%), frequent intake of vitamin A-rich fruits and vegetables (71%), and low egg consumption (7.3%). Overall, 49.3% of students met the minimum dietary diversity threshold, while only 15.3% reported consuming at least three meals in the prior day.
Multivariable analysis evaluated the association between participation in the school feeding program and attainment of minimum dietary diversity while adjusting for potential confounders. Participation was associated with a 45% higher likelihood of meeting minimum dietary diversity (APR = 1.45; 95% CI: 1.00–2.12), with a p-value of 0.051, marginally above conventional thresholds for statistical significance. Household food insecurity significantly reduced the likelihood of achieving minimum dietary diversity (APR = 0.44; 95% CI: 0.23–0.85; p = 0.014). After adjustment, age, parental education, and household size were not significantly associated with dietary diversity.
Participation in the feeding program showed a strong positive association with meeting recommended meal frequency. Students who participated were nearly 12 times more likely to report consuming three or more meals in the prior 24 hours compared with non-participants (APR = 11.68; 95% CI: 2.81–48.53; p = 0.001). Gender was associated with meal frequency: males were 48% less likely than females to consume regular meals (APR = 0.52; 95% CI: 0.28–0.99; p = 0.047). Other sociodemographic variables did not show significant adjusted associations with meal frequency.
The study used convenience selection of wards and schools, which may limit generalizability. Data derive from a single 24-hour recall and the survey period fell in April, the pre-harvest lean season for the district, which may influence dietary diversity and food availability. The analysis and interpretation followed ethical approvals obtained from the Muhimbili University of Health and Allied Sciences Research Ethics Committee (IRB: MUHAS-REC-03-2025-2693). Data are not publicly shared due to ethical agreements, but may be requested from the issuing IRB per the authors’ report.
The evaluation indicates that while school feeding programs are present in Kisarawe, implementation is incomplete and constrained by infrastructural and logistical weaknesses. Program participation was associated with markedly higher meal frequency and a marginal increase in dietary diversity, suggesting potential benefits if programs were fully implemented and scaled. Persistent household food insecurity remained an important barrier to dietary diversity, underscoring that school-based feeding alone may not fully address nutritional deficits.
The authors recommend strengthening infrastructure (designated dining areas, storage, and kitchen capacity), ensuring reliable and diverse food supplies, and expanding universal coverage to improve program effectiveness and sustainability. Linking school feeding to local food systems and addressing household food insecurity may further enhance dietary adequacy among school-aged children in Kisarawe and similar settings.