Liver Fat and Cardiometabolic Risk Factors Among School-Age Children

Abstract

Background and aims: Nonalcoholic fatty liver disease is a major risk factor for cardiometabolic disease in adults. The burden of liver fat and associated cardiometabolic risk factors in healthy children is unknown. In a population-based prospective cohort study among 3,170 10-year-old children, we assessed whether both liver fat accumulation across the full range and nonalcoholic fatty liver disease are associated with cardiometabolic risk factors already in childhood.

Approach and results: Liver fat fraction was measured by magnetic resonance imaging, and nonalcoholic fatty liver disease was defined as liver fat fraction ≥5.0%. We measured body mass index, blood pressure, and insulin, glucose, lipids, and C-reactive protein concentrations. Cardiometabolic clustering was defined as having three or more risk factors out of high visceral fat mass, high blood pressure, low high-density-lipoprotein cholesterol or high triglycerides, and high insulin concentrations. Nonalcoholic fatty liver disease prevalences were 1.0%, 9.1%, and 25.0% among children who were normal weight, overweight, and obese, respectively. Both higher liver fat within the normal range (<5.0% liver fat) and nonalcoholic fatty liver disease were associated with higher blood pressure, insulin resistance, total cholesterol, triglycerides, and C-reactive protein concentrations (P values < 0.05). As compared with children with <2.0% liver fat, children with ≥5.0% liver fat had the highest odds of cardiometabolic clustering (odds ratio 24.43 [95% confidence interval 12.25, 48.60]). The associations remained similar after adjustment for body mass index and tended to be stronger in children who were overweight and obese.

Conclusions: Higher liver fat is, across the full range and independently of body mass index, associated with an adverse cardiometabolic risk profile already in childhood. Future preventive strategies focused on improving cardiometabolic outcomes in later life may need to target liver fat development in childhood.

Figure 1

Associations of liver fat fraction and nonalcoholic fatty liver disease with cardiometabolic risk factors at school age. Values are regression coefficients (95% CI) from linear regression models that reflect differences in childhood cardiometabolic risk factors in SDS per SDS change in childhood liver fat fraction as compared with the reference group (children with <2.0% of liver fat; left side of each graph) or for children with nonalcoholic fatty liver disease as compared with the reference group (children with <5.0% of liver fat; right side of each graph). Associations are adjusted for child’s age, sex, ethnicity, maternal prepregnancy BMI, and maternal education. Trend lines are given only when P value for linear trend <0.05.

Figure 2

Associations of liver fat fraction and nonalcoholic fatty liver disease with odds of clustering of cardiometabolic risk factors at school age. Values are ORs (95% CI) analyzed in a subgroup of cases with complete data for all cardiometabolic variables (n = 1,906) that reflect the risk of cardiometabolic clustering per increase in liver fat fraction as compared with the reference group (<2.0%; left side of the figure) or for children with nonalcoholic fatty liver disease as compared with the reference group (children with <5.0% of liver fat; right side of the figure). Bars represent the percentage of cardiometabolic clustering per liver fat fraction group. Cardiometabolic clustering was defined as having three or more risk factors (high [greater than seventy‐fifth percentile] visceral fat mass, high [greater than seventy‐fifth percentile] systolic or diastolic blood pressure, low [less than twenty‐fifth percentile] HDL cholesterol or high [greater than seventy‐fifth percentile] triglycerides, and high [greater than seventy‐fifth percentile] insulin. Associations are adjusted for child age, sex, ethnicity, maternal prepregnancy BMI, and maternal education. Trend lines are given only when P value for linear trend <0.05.