Adherence to dietary guidelines associated with lower medical service utilization in preschoolers: a longitudinal study

Early childhood is a rapid growth period critical for physiological, cognitive, and immune development, as children transition from liquid to solid foods and establish eating behaviors. Healthy diets providing adequate and diverse nutrients support growth and reduce later-life chronic disease risk, whereas excessive calories and unhealthy patterns increase obesity and future noncommunicable diseases. Prior evidence indicates diet quality declines from 6 months to 4 years, coinciding with increased consumption of processed foods. Food-based dietary guidelines aim to improve population diet quality. Assessing overall diet quality is important for health promotion; fruits and vegetables may reduce oxidative stress and inflammation, while Westernized diets have been linked to higher asthma and respiratory infections in preschoolers. Few studies have evaluated whether adherence to dietary guidelines in preschoolers reduces medical service utilization or costs. Taiwan’s universal NHI enables comprehensive tracking of medical utilization, with high outpatient visit rates in children and particularly for respiratory diseases. This study developed the Taiwanese Children Healthy Eating Index (TCHEI) aligned with Taiwanese Food-Based Dietary Guidelines, validated it against nutrient intake, and investigated whether higher TCHEI scores are associated with lower overall and respiratory-related medical service utilization and expenditures in Taiwanese preschoolers.

Evidence suggests diet quality influences immune and respiratory health in children: higher intake of fruits and vegetables with antioxidant and anti-inflammatory components is protective, whereas Western dietary patterns are associated with increased asthma and respiratory infections. Few studies have examined cost-related outcomes of diet quality in young children. A Canadian study reported trends toward lower medical expenditures and visits among children with higher healthy eating scores, though findings were not statistically significant, possibly due to low visit rates or differing healthcare access and care-seeking behaviors. Prior research also links socioeconomic status and maternal education to children’s diet quality, food security, and obesity risk, which in turn relate to higher healthcare costs. Studies from various countries associate better dietary diversity and feeding indices with lower stunting and undernutrition risk, supporting the premise that higher-quality diets improve growth and may reduce healthcare needs.

Design and data sources: Prospective longitudinal analysis linking the Nutrition and Health Survey in Taiwan (NAHSIT) 2013–2016 to the National Health Insurance Research Database (NHIRD) 2013–2018 using scrambled IDs. Population and sampling: NAHSIT employed multistage, stratified, clustered probability sampling across 20 strata (counties/cities). Of 1295 children aged 2–6 years (2–3 y: n=370; 4–6 y: n=925), 614 consented to data linkage and formed the analytic cohort; no significant demographic differences versus the full NAHSIT pediatric sample. Ethics: Informed consent from parents; approvals from Academia Sinica IRB and China Medical University Central Regional Research Ethics Committee (CRREC-108-157). Dietary assessment: One 24-h dietary recall via trained interviewers with portion aids; all foods and beverages recorded. Foods classified into six TFBDG groups with standardized servings: grains; soy/fish/eggs/meat; dairy; vegetables; fruits; oils and nuts. Nutrient intakes derived from Taiwan Food Nutrient Composition Database and USDA FoodData Central; nutrient density calculated per 1000 kcal. TCHEI development: Total score 0–100. Dietary adequacy components (0–10 points each for six food groups; extra 0–5 points each for unrefined grains, soy products, dark vegetables) based on TFBDG recommended servings for ages 2–6 y with low physical activity; adequacy subtotal max 75. Dietary behavior components (0–5 points each; max 25): frequency of eating breakfast, sugar-sweetened beverage intake, fried food, snack/biscuit intake, seasoning use, adapted from TFBDG and Taiwanese Youth Healthy Eating Index. Higher scores reflect better adherence. Medical service utilization and expenditures: From NHIRD, annualized counts and expenditures for outpatient and emergency visits were computed; multiple visits on the same day counted as one. Respiratory disease visits defined by ICD-9 (460–466, 470–478, 480–487, 490, 493) and ICD-10 (J00–J06, J30–J39, J10–J18, J20–J21, J40, J45) codes. Follow-up from interview date to December 31, 2018. Annual expenditures summed across services and medications, discounted at 3% per year and adjusted by Taiwan CPI (2013–2018). Covariables: Child sex, age, region (north, middle, south, east/island), mother’s education (junior high or below; high school; college or above), household financial status (enough; just enough; some difficulties; very difficult), total energy intake. Nutritional status: height-for-age z-score (HAZ) and weight-for-age z-score (WAZ). Statistical analysis: TCHEI categorized into tertiles (age-specific cut-points: 2–3 y: 46.3/46.0; 4–6 y: 55.1/54.9). Group differences assessed with ANOVA (fat %) and Kruskal–Wallis (others). Associations between TCHEI tertiles and utilization modeled using multivariable generalized linear models: log link with Poisson distribution for visit counts; log link with gamma distribution for expenditures. Models adjusted sequentially: Model 1 (demographics, SES, region), Model 2 (Model 1 + total energy). Age-stratified analyses for 2–3 y and 4–6 y. Two-sided tests, P<0.05. Analyses unweighted due to distribution characteristics of utilization/costs. Sensitivity analyses excluded accidental injuries and participants with congenital conditions.

- TCHEI and nutritional status: Mean TCHEI was 50.6±10.6. Higher TCHEI (T3) associated with greater HAZ (0.07 vs. −0.12 in T1), higher maternal education, and better household financial status. - Nutrient intake differences (T3 vs. T1): Higher protein intake (3.5 vs. 2.2 g/kg BW), PUFA (9.23 vs. 8.09 g/1000 kcal), dietary fiber (7.2 vs. 4.3 g/1000 kcal), vitamin C (76.4 vs. 30.8 mg/1000 kcal), vitamin B6 (0.9 vs. 0.7 mg/1000 kcal), potassium (1280 vs. 978 mg/1000 kcal), calcium (380 vs. 224 mg/1000 kcal), magnesium (139 vs. 110 mg/1000 kcal), iron (7.0 vs. 5.3 mg/1000 kcal); lower saturated fat (12.1 vs. 13.5 g/1000 kcal; P=0.085). All P<0.0001 unless noted. - All-disease medical service utilization: In the total cohort, T2 had 18% fewer overall visits (95% CI 0.78–0.87) and 17% fewer outpatient visits (95% CI 0.78–0.87) vs. T1 (Model 2). Emergency visit reductions trended but were not significant. - Age 2–3 y: Both T2 and T3 had fewer visits vs. T1—overall: −25% and −16%; outpatient: −24% and −15%; emergency: −52% and −58% (Model 2). - Age 4–6 y: T2 had 15% fewer overall and outpatient visits (95% CI 0.80–0.91) vs. T1; emergency visits not significantly reduced. - Respiratory disease utilization: Total cohort T2 showed reductions—overall: −17%; outpatient: −16%; emergency: −51% vs. T1 (Model 2). Age 2–3 y: T2 and T3 reduced overall (−27%, −21%), outpatient (−26%, −20%), and emergency (−73%, −72%) respiratory visits. Age 4–6 y: T2 reduced overall and outpatient respiratory visits by 11%; emergency not significant. - Medical expenditures: All-disease total expenditures were lower in T2 vs. T1 for total cohort (expβ 0.68, 95% CI 0.57–0.80), 2–3 y (0.72, 0.53–0.96), and 4–6 y (0.68, 0.56–0.84). Outpatient expenditures: 4–6 y T2 −33% (95% CI 0.54–0.82); 2–3 y not significant. Respiratory expenditures: 2–3 y T2 −28% (95% CI 0.53–0.97) for total respiratory costs; respiratory emergency expenditures markedly lower in 2–3 y T2 and T3 (−87% [95% CI 0.04–0.37] and −80% [95% CI 0.06–0.63]). - Cost examples: Children aged 2–3 in T1 had the highest mean annual expenditures for all-disease outpatient (NT$12,159; US$392) and emergency (NT$1,189; US$38), and for respiratory outpatient (NT$6,121; US$197) and emergency (NT$793; US$25). - Sensitivity analyses: Excluding accidental injuries yielded consistent results. Excluding congenital conditions attenuated significance for all-disease expenditures in 2–3 y, while utilization findings remained consistent.

Higher adherence to the Taiwanese Food-Based Dietary Guidelines, as quantified by the TCHEI, was associated with improved nutrient profiles and growth indicators and with reduced medical service utilization and expenditures among preschoolers, particularly pronounced in toddlers aged 2–3 years. These findings address the research question by demonstrating that better diet quality corresponds to fewer outpatient and emergency visits, including for respiratory illnesses, which are common drivers of pediatric healthcare use in Taiwan. Potential mechanisms include higher intakes of dietary fiber, polyunsaturated fats, vitamins, and minerals that may enhance immune function and reduce inflammation, contributing to fewer infections and medical visits. Socioeconomic gradients in diet quality were evident—higher maternal education and better household finances correlated with higher TCHEI—suggesting part of the association may reflect broader resource and knowledge advantages, though models adjusted for these factors. Among 4–6-year-olds, a U-shaped pattern (lowest utilization in T2) suggests complex influences of behavior, environment, and care-seeking on healthcare use, beyond diet alone. Findings align with prior work showing trends toward lower healthcare use with better diet quality in school-aged children, though previous studies lacked power or comprehensive utilization data. Given Taiwan’s high pediatric outpatient visit rates, diet quality improvements may be a feasible strategy to reduce healthcare burden and expenditures in early childhood.

The study developed and validated the Taiwanese Children Healthy Eating Index (TCHEI) based on TFBDG and showed that preschoolers with better adherence—particularly toddlers—had superior nutrient intake, healthier growth indicators, and lower overall and respiratory-related medical service utilization and expenditures. These results support promoting guideline-adherent diets in early childhood as a public health strategy to improve health and reduce healthcare use. Future research should: (1) employ multi-day dietary assessments to capture usual intake; (2) incorporate objective measures of physical activity, environmental exposures, and psychosocial factors; (3) investigate causal pathways and potential non-linear associations; and (4) evaluate generalizability and cost-effectiveness in diverse populations and healthcare systems.

- Dietary assessment relied on a single 24-h recall, which may not reflect usual intake, although preschool diets may be relatively stable short term. - The TCHEI was constructed per TFBDG and analyses were unweighted; findings are most applicable to the study sample and may not generalize to other populations. - Key covariates such as physical activity, environmental exposures (e.g., air pollution, sanitation, healthcare access), and psychological stress were not collected, leaving potential residual confounding. - Indirect costs (e.g., caregiver productivity loss) were not assessed. - Sensitivity analyses indicated that after excluding congenital conditions, the inverse association with all-disease expenditures in 2–3-year-olds was no longer significant, suggesting results for expenditures may be sensitive to underlying health conditions.