Skip to main content
. 2017 Sep 14;20(18):3275–3284. doi: 10.1017/S1368980017002348

Table 3.

Associations between family meal type (breakfast, lunch, dinner, total meals) and pre-school children’s diet quality and BMI percentile among non-Hispanic and Hispanic households; baseline data from NET-Works and GROW studies among parent–child (aged 2–5 years) pairs from respectively Minnesota (n 222 non-Hispanics; n 312 Hispanics) and Tennessee (n 545 Hispanics; n 55 non-Hispanics), USA, 2012–2014

Non-Hispanic households (n 276) Hispanic households (n 857)
β sd P value β sd P value
HEI-2010
Breakfast
Model 1* 1·44 0·32 <0·0001 0·19 0·15 0·1899
Model 2 1·10 0·34 0·0013 0·15 0·15 0·3347
Model 3 1·30 0·41 0·0015 0·05 0·19 0·7948
Lunch
Model 1 0·61 0·33 0·0632 0·20 0·15 0·1799
Model 2 0·32 0·34 0·3345 0·13 0·15 0·3807
Model 3 −0·46 0·40 0·2515 0·04 0·19 0·8497
Dinner
Model 1 0·50 0·39 0·1901 0·29 0·16 0·0728
Model 2 0·51 0·37 0·1721 0·29 0·16 0·0726
Model 3 0·37 0·38 0·3373 0·26 0·17 0·1204
Total
Model 1 0·52 0·15 0·0005 0·12 0·06 0·0578
Model 2 0·38 0·15 0·0137 0·10 0·07 0·1211
Model 3
BMI percentile
Breakfast
Model 1 −0·27 0·37 0·4638 0·00 0·18 0·9793
Model 2 −0·19 0·40 0·6433 −0·12 0·19 0·5173
Model 3 0·12 0·48 0·8089 −0·26 0·24 0·2723
Lunch
Model 1 −0·31 0·37 0·3948 0·13 0·18 0·4780
Model 2 −0·47 0·39 0·2293 0·06 0·19 0·7299
Model 3 −0·54 0·47 0·2580 0·23 0·24 0·3453
Dinner
Model 1 −0·26 0·43 0·5413 −0·07 0·20 0·7326
Model 2 −0·10 0·43 0·8193 −0·01 0·20 0·9775
Model 3 0·04 0·45 0·9349 0·02 0·21 0·9413
Total
Model 1 −0·17 0·17 0·3296 0·02 0·08 0·8502
Model 2 −0·15 0·18 0·3903 −0·01 0·08 0·8827
Model 3

NET-Works, Now Everybody Together for Amazing and Healthy Kids; GROW, Growing Right Onto Wellness; SNAP, Supplemental Nutrition Assistance Program; WIC, Special Supplemental Nutrition Program for Women, Infants, and Children; HEI-2010, Health Eating Index-2010.

Estimates are beta coefficients. Interpretation=for every one-day increase in eating (breakfast, lunch, dinner, total) family meals, HEI-2010 score and BMI percentile changes. Values shown in bold are significant at P<0·05.

*

Model 1: unadjusted.

Model 2: adjusted for age, sex, marital status, employment status and site. For non-Hispanic participants, also adjusted for race/ethnicity. BMI models also adjusted for mean total daily energy intake.

Model 3: adjusted for all of the covariates in Model 2 plus the frequency of the other two family meals.