We estimate healthcare costs of exposure to ACEs in California, by combining prevalence data from four years of the California Behavioral Risk Factor Survey (BRFS) with estimates of healthcare costs during 2013 from Dieleman et al.[12] for those chronic conditions identified as plausibly linked to ACEs in Waehrer et al.[13] The California BRFS is part of the Behavioral Risk Factor Surveillance System (BRFSS), the US system of cross-sectional state-based telephone surveys of residents’ health-related behaviors. State health departments queried residents aged 18 and older in all 50 states and 3 US territories about demographics, risk behaviors, chronic health conditions, and use of preventive services using a standardized questionnaire developed in collaboration with the Centers for Disease Control (CDC). Additional topical modules developed by the CDC are administered in some years by some states, probing topics such as arthritis, depression, and ACEs, with state-developed questions of local interest also allowed. This study is based on 2008, 2009, 2011, and 2013 California BRFS data on ACEs from the California Department of Public Health (CDPH)[14].

Seminal research using data from a California health plan has associated ACEs with increased odds of adverse health outcomes including heart disease, liver disease, and lung disease, with a graded association between the number of ACEs and risk of some outcomes[15–19]. Subsequent studies found support for these results using data from several states as well as global data[20–23]. At the same time, these studies varied in the demographic controls used, and how extensively the ACEs-disease relationship was mediated by health risk factors was unclear. A companion study by our research team used parallel logit models with a common set of socioeconomic controls to examine whether the direct ACEs-health relationship for seven health outcomes persisted net of health risk factors such as smoking, heavy drinking, and obesity[13, 24]. We applied the models to BRFSS data from California and 13 other states. That study also estimated the extent to which ACEs association with health outcomes ran indirectly through smoking, heavy drinking, and obesity by comparing the results from models with and without the three health risk factors. Health outcomes were deemed to be ACEs-related if, (1) net of the three health risk factors, they had a graded, dose-response relationship with the number of ACEs, (2) they were significantly associated with exposure to 4 or more ACEs, and (3) these relationships were consistently observed across the different BRFSS states[13, 24]. Among the seven health outcomes analyzed, five—asthma, arthritis, depression, COPD, and cardiovascular disease–were plausibly linked to ACEs under the criteria listed above. Two other conditions, diabetes and cancer, did not meet our criteria and therefore were excluded from this cost study.

To estimate the excess risk of disease associated with ACEs exposure, we combined results from logit models in Waehrer et al.[13] with California BRFS data on (1) the prevalence of ACEs-related health conditions for different ACEs exposures and (2) the proportion of CA population exposed to ACEs. Specifically, Table 1 presents the average change in the probability of the health condition for those with ACEs relative to those with zero ACEs using estimates from Waehrer et al.[13]. S1 Table presents the full set of estimates and confidence intervals from that study. To be consistent with the healthcare cost estimates from Dieleman et al.[12] (explained below), we used estimates in columns (1) to (3) of S1 Table, which present the increased probability of disease associated with ACEs exposure from models that do not control for risk factors such as lifetime smoking, heavy drinking, or obesity.

For each disease and category of ACEs (n = 1, 2–3, 4+), we calculated the excess risk of disease due to n ACEs (hereafter referred to as ACEs-Associated Fraction or AAF) as the percentage of the adult population with n ACEs times the predicted change in probability of the problem for adults with n ACEs versus 0 ACEs divided by the prevalence of the problem among adults with n ACEs.

Thus far, we focused broadly on the excess costs attributable to exposure to ACEs. However, 4+ ACEs have been particularly implicated in epidemiological studies of chronic disease. Our estimates show that exposure to 4+ ACEs was associated with $3.8 billion in healthcare expenses in California during 2013 (Table 3), an estimated 36% of total ACEs-attributable healthcare spending, and approximately 155,000 DALYs.

Although this study helps to quantify the long-term impact of childhood exposure to ACEs in California, DALYs account for more than 90% of total estimated costs. Therefore, our cost estimate is quite sensitive to, and closely tracks the cost per DALY. Popular approaches to valuing DALYs are to survey how much a person would be willing to pay to avoid loss of a year of healthy life [e.g.,60, 61] or simply to divide the value of statistical life by the number of years of life that would be lost if the average person in the life valuation studies died (discounted to present value) [e.g.,31, 62, 63]. Hirth et al.[64] reviewed 42 studies on DALY/QALY values and found a wide range depending on methodology. Our use of $235,855 per DALY is in line with that study’s “intermediate median value” of $249,872 per DALY (inflated to 2017 dollars using the Employment Cost Index) from surveys, and lower than the $350,000 per DALY implied by a more recent meta-analytic study on the value of a statistical life[65]. At the same time, our use of a uniform value per DALY ignores evidence from disease- or country-specific studies which suggest the value per DALY is situational. For example, an intermediate value per DALY in surveys of asthma patients is $31,890 (in 2017 dollars), much lower than that for dialysis patients[66] and lower than the value used in our study. Given the uncertainty surrounding valuation of DALYs, we present the values separately from health expenditures and also present non-monetized estimates of the DALYs associated with ACEs.

Another source of variation in ACEs costs derives from our estimates of the ACEs-associated fractions (AAF), which describe associations but not necessarily causation. Our estimates are specific to California and may not carry over to other localities with different policies toward child abuse and mental health. Prior estimates of ACE associations with health outcomes across 13 BRFSS states[13, 24] yielded larger estimates of ACEs-associated fractions in those other states relative to California for obesity, asthma, and depression, but smaller estimates for heavy drinking and CVD (results available upon request). However, applying the 13-state estimates to California leaves the cost estimate relatively unchanged–ACEs exposure is associated with an estimated $9.7 billion in healthcare expenses and 378,000 DALYs valued at $89 billion.

This study relies on prior estimates of ACEs association with disease, healthcare expenses, and DALYs which have their own underlying limitations. The ACEs-associated fraction estimates are based on our prior study[13, 24] using retrospective ACEs data which are self-reported and subject to recall bias[67, 68]. Total associations with 4+ ACEs in our prior study are generally consistent with prior BRFSS studies for asthma and depression (e.g. [21][23]) and with results from a meta-analysis by Hughes et al.[69]; however, the odds of cardiovascular disease and COPD for 4+ ACEs are smaller in our study than in the seminal ACEs study which focused on a sample of mostly middle-class white adults in California [e.g., 18]. To the degree that ACEs exposure is under- or over-reported and because our analysis was unable to control for childhood background or the timing and intensity of ACEs, our estimates may misstate ACEs associations with outcomes and have a limited causal interpretation. Self-reports of health conditions in BRFSS may also be subject to recall bias and the underreporting of socially undesirable information. Bowlin et al.[70] report that the BRFSS may understate the prevalence of cardiovascular disease risk factors such as obesity and smoking. At the same time, this and other studies have found moderate sensitivity in BRFSS self-reports of diabetes and arthritis[71, 72] and BRFSS prevalence has been treated as a benchmark to test the possibility of using data on filled heart disease prescriptions as a proxy for prevalence[73]. However, to the extent that we underestimate the prevalence of the health conditions and risk factors, we may underestimate the AAF and costs of ACEs-associated health outcomes. For health costs, we relied on published healthcare expenses from Dieleman et al.[12] who used incomplete survey data requiring statistical modeling and scaling up to reflect national estimates and compensate for the lack of a census of US healthcare expenditures. The DALYs estimates share the limitations of the GBD estimates underlying them, notably the use of disability weights that are not tailored to high-income countries, the garbage codes in 15% of US cause-of-death records, and the use of models rather than California data sets to decompose nonfatal estimates by state. Finally, while our study estimates ACEs-related costs for many of the leading causes of death in the US, we did not examine a comprehensive list of all possible health outcomes. Thus, our estimates of health burden and costs may understate the true impact of ACEs exposure.

Despite these limitations, this study provides the first look at the adult health burden and healthcare costs to Californians associated with exposure to ACEs. Our estimates can be used to inform efforts to reduce the impact of ACEs and toxic stress in California. Future research could extend these results to other states or provide nationwide estimates.