Consumption of Foods Derived from Subsidized Crops Remains Associated with Cardiometabolic Risk: An Update on the Evidence Using the National Health and Nutrition Examination Survey 2009-2014

Abstract

In this study, we examined the associations between the consumption of foods derived from crops subsidized under the 2008 United States (US) Farm Bill and cardiometabolic risk factors and whether the magnitude of these associations has changed since the 2002 US Farm Bill. Four federal databases were used to estimate daily consumption of the top seven subsidized commodities (corn, soybeans, wheat, rice, sorghum, dairy, and livestock) and to calculate a subsidy score (0-1 scale) for Americans' daily dietary intake during 2009-2014, with a higher score indicative of a higher proportion of the diet derived from subsidized commodities. The cardiometabolic risk factors included obesity, abdominal adiposity, hypertension, dyslipidemia, and dysglycemia. Linear and logistic regression models were adjusted for age, sex, race/ethnicity, the poverty-income ratio, the smoking status, educational attainment, physical activity, and daily calorie intake. During 2009-2014, adults with the highest subsidy score had higher probabilities of obesity, abdominal adiposity, and dysglycemia compared to the lowest subsidy score. After the 2002 Farm Bill (measured using data from 2001-2006), the subsidy score decreased from 56% to 50% and associations between consuming a highly-subsidized diet and dysglycemia did not change (p = 0.54), whereas associations with obesity (p = 0.004) and abdominal adiposity (p = 0.002) significantly attenuated by more than half. The proportion of calories derived from subsidized food commodities continues to be associated with adverse cardiometabolic risk factors, though the relationship with obesity and abdominal adiposity has weakened in recent years.

Keywords: cardiometabolic disease; farm subsidies; food policy; obesity.

Figure 1

Adjusted prevalence ratio of cardiometabolic risk factors by subsidy score quartile in the National Health and Nutrition Examination Survey from 2009–2014. Subsidy score quartiles were defined as follows: Q1 is 0.00–0.41; Q2 is 0.42–0.53; Q3 is 0.54–0.63; and Q4 is 0.64–1.00. Q1 was used as a comparison group. Abdominal adiposity was defined as a ratio of waist circumference to height of at least 0.59. Dyslipidemia was defined as diagnosed (self-reported) or undiagnosed (no self-reported diagnosis and non-HDL cholesterol level ≥ 160 mg/dL) or currently taking anticholesterolemia medication. Dysglycemia was defined as a self-reported diabetes diagnosis or hemoglobin A1c level of at least 5.7%. Hypertension was defined as diagnosed (self-reported) or undiagnosed (no self-reported diagnosis and systolic blood pressure ≥ 140 mm Hg or diastolic blood pressure ≥ 90 mm Hg) or currently taking antihypertensive medication. Obesity was defined as a body mass index of at least 30 kg/m². Individuals with missing data were excluded from the models. Model adjusted for sex, age, race/ethnicity, the highest education level, the poverty–income ratio, the smoking status, participation in moderate to vigorous physical activity, and total daily caloric intake. * p < 0.05.

Figure 2

Adjusted prevalence ratio of cardiometabolic risk factors by subsidy score quartile in the National Health and Nutrition Examination Survey from 2009–2014 compared to the National Health and Nutrition Examination Survey 2001–2006. Subsidy score quartiles were defined as follows: Q1 is 0.00–0.41; Q2 is 0.42–0.53; Q3 is 0.54–0.63; and Q4 is 0.64–1.00. Obesity was defined as a body mass index of at least 30 kg/m². Abdominal adiposity was defined as a ratio of waist circumference to height of at least 0.59. Dysglycemia was defined as a self-reported diabetes diagnosis or hemoglobin A1c level of at least 5.7%. Individuals with missing data were excluded from the models. Model adjusted for sex, age race/ethnicity, the highest education level, the poverty–income ratio, the smoking status, participation in moderate to vigorous physical activity, and total daily caloric intake. Difference in the prevalence ratio between time periods was compared using interaction with the time period; * p < 0.05.

Figure 3

Adjusted prevalence of obesity (A), abdominal adiposity (B), and dysglycemia (C) by subsidy score quartile in the National Health and Nutrition Examination Survey between 2001–2006 and 2009–2014. Subsidy score quartiles were defined as follows: Q1 is 0.00–0.41; Q2 is 0.42–0.53; Q3 is 0.54–0.63; and Q4 is 0.64–1.00. Obesity was defined as a body mass index of at least 30 kg/m². Abdominal adiposity was defined as a ratio of waist circumference to height of at least 0.59. Dysglycemia was defined as a self-reported diabetes diagnosis or hemoglobin A1c level of at least 5.7%. Individuals with missing data were excluded from the models. Model adjusted for sex, age, race/ethnicity, the highest education level, the poverty–income ratio, the smoking status, participation in moderate to vigorous physical activity, and total daily caloric intake. Significant differences in the adjusted prevalence within a quartile between the two time periods are denoted by * p < 0.05 and ** p < 0.0001.