Background: Dietary guidelines recommend consumption of 1% or skim (i.e. non-fat) milk for children >2 years due to concerns about the role of saturated fat in raising low-density lipoprotein cholesterol (LDL) levels and cardiovascular disease risk. Low- or non-fat milk is also recommended to promote energy balance and reduce obesity risk. While more recent evidence in adults suggests no association between milkfat and cardiovascular disease risk, this association has not been well studied among children who tend to be the highest consumers of milk. No known studies have demonstrated an association between milkfat consumption and lipid levels in children. Our objective was to determine if there is an association between milk fat consumption and risk of dyslipidemia and obesity among U.S. children.
Methods: We used cross-sectional data from children 2-19 years from the National Health and Nutrition Examination Survey (NHANES) 2011-2014 (n=7261). Usual milkfat intake (i.e. never/rare milk consumption, non-fat, 1%, 2%, or whole) was self-reported by parent or child. Lipid levels (12-19 years only) and adiposity (BMI z-score and weight category [underweight, normal, overweight, or obese] in all ages) were assessed by trained staff using standardized methods. Covariates included child age, sex, household income, race/ethnicity, physical activity, dieting, and total energy intake. We used multivariate linear and logistic regression models to examine the association between milkfat consumption and our outcomes and conducted pairwise comparisons to assess differences between different milkfat consumers.
Results: Most children consumed 2% milk (42.3%), followed by whole (26.1%), 1% milk (11.7%), never/rare consumption (10.2%), and nonfat (9.7%). We found no significant trends in mean cholesterol (total, HDL, LDL) with consumption of milk with greater fat content. There was a significant trend towards higher triglycerides as milkfat intake increased (p=0.01) though this association was stronger among 2% milk compared to whole fat consumers. We found no significant trend in prevalence of each BMI category across milk fat intake. Pairwise comparisons demonstrated that usual consumption of 1% and 2% milkfat was associated with both higher BMI Z-score and odds of obesity compared to whole milk.
Conclusion: We found no significant linear trend between usual %milk fat consumed and indicators of cardiovascular disease (BMI-Z and lipid levels) in children, except for triglycerides which rose as milkfat intake increased. However, triglycerides were higher among the 2% compared to whole milkfat consumers. Overall, whole milk was not associated with increased indicators of cardiovascular disease risk in children.