Testing allostatic load factor structures among adolescents: A structural equation modeling approach.

OBJECTIVES: Allostatic load (AL) represents cumulative biological "wear and tear" that results from chronic stress exposure over time, ultimately increasing risk for chronic disease. A consensus is lacking regarding the best operationalization of AL, particularly for younger, less studied populations. The purpose of this study was to test multiple hypothesized factor structures for AL to determine the best measurement approach for adolescents.

METHODS: We analyzed biologic data for 1900 adolescents aged 12-18 from four waves (2003-2010) of the National Health and Nutrition Examination Survey. AL indicator variables included cardiovascular (systolic BP, creatinine), metabolic (HDL, LDL, triglycerides, insulin, fasting glucose, HA1C, body mass index [BMI], waist circumference), and immune (albumin, CRP, WBC, EBV) biomarkers. Structural equation modeling was used to test the fit of five hypothesized AL factor structures.

RESULTS: The data best supported a unidimensional factor structure, where the AL construct directly influenced each of the indicator variables. All but two of the indicators (HDL and albumin) had positive factor loadings, thus, as AL increases the values for those indicators also increase. The best indicators for AL were those measuring metabolic dysregulation, with BMI and waist circumference having the highest factor loadings (0.95 and 0.982, respectively).

CONCLUSIONS: BMI and waist circumference may be some of the earliest clinical signs of elevated AL that manifest among adolescents. Future research should aim to include neuroendocrine biomarkers in their AL measures to have a more robust estimation of AL in younger populations.