The characteristics of the noninstitutionalized US population are shown by year in . Some demographic characteristics have shifted over time. Insurance status also has shifted, such that the proportion with private insurance declined while publicly funded insurance and lack of insurance increased. Of note, the prevalence of some categories of diagnosed mental health conditions (specifically, affective disorders as well as anxiety, somatoform, dissociative, and personality disorders) trended upward over time.
Sociodemographic Characteristics of US Adults, by Year of the National Health Interview Surveya
In , we show characteristics of the population across the strata of psychological distress. In our sample, 86.2% of the population was classified as having no/low psychological distress, 9.7% had mild-moderate psychological distress, and 4.1% had severe psychological distress. Every characteristic we examined differed significantly across the strata of psychological distress. Those with higher levels of distress were more likely to be older, female, unmarried, minority, poorer, uninsured, sicker both medically and psychiatrically, less educated, and living in a rural area; these characteristics are consistent with previous reports.28
All of these factors were therefore selected as covariates in our models adjusting for factors associated with psychological distress. Note that the frequencies for the K6 strata in the subset population of those with linked NHIS and MEPS data were similar to those in the NHIS population who were not in the linked MEPS data: in the NHIS-MEPS data, there were 86.2% in the no/low distress group (K6 score 0–6), 9.7% in the mild-moderate group (K6 score 7–12), and 4.1% in the severe group (K6 score >12) compared with 87.9% in the no/low group, 8.7% in the mild-moderate group, and 3.4% in the severe group among the NHIS-only population.
Sociodemographic Characteristics of US Adults, by Level of Psychological Distressa
Time was not significantly related to psychological distress in any of our analyses. It remained stable over the period from 1997 to 2004 (). In addition, the mean psychological distress score was flat over this time period. This indicates that despite an increase in diagnosed mental health conditions over time, the symptom burden did not appear to increase.
Unadjusted Trends in Psychological Distress Among US Adults, 1997 to 2004
The shows how expenditures and utilization changed over time within each stratum of psychological distress (unadjusted values are shown). Note that we present mean values, as these could be weighted using the survey sample. As expected, the fully adjusted model for total per person overall healthcare expenditures showed a significant upward trend over time (0.05, 95% confidence interval [CI] 0.04, 0.06; P <.001). Outpatient expenditures also trended upward in the fully adjusted model (0.04, 95% CI 0.03, 0.05; P <.001). However, we did not observe any change over time in the number of outpatient visits (in the fully adjusted model, 1.01, 95% CI 0.99, 1.02; P = .26).
Figure Strata of Psychological Distress (No/Low, Mild-Moderate, Severe) Versus Subsequent Total Expenditures (Panel A), Outpatient and Office-Based Expenditures (Panel B), and Outpatient and Office Visits (Panel C)a
There was a gradient in total per person expenditures across strata of psychological distress. The fully adjusted mean annualized per person total healthcare expenditure (in 2006 US dollars) was $4820 (95% CI $3830, $4986) for those in the group with no/low distress. Compared with the no/low-distress group, these mean expenditures were significantly higher among the group with mild-moderate distress ($8549, 95% CI $5645, $9261; P <.001) and the group with severe distress ($11,954, 95% CI $7704, $13,646; P <.001). The mean total expenditures for the groups with mild-moderate distress and severe distress also differed significantly (P <.0001).
Our findings for outpatient expenditures were similar to our findings for total expenditures. The mean annualized outpatient expenditures in 2006 US dollars were $1396 (95% CI $1354, $1434) for the group with no/low distress, $2090 (95% CI $1924, $2263) for the group with mild-moderate distress, and $2766 (95% CI $2445, $3105) for the group with severe distress. These expenditures differed significantly between each group (P <.0001).
For outpatient visits, the strata of psychological distress also differed. With the no/low-distress group as the referent, the incremental relative risk for an office-based or outpatient visit for those in the group with mild-moderate distress was 1.27 (95% CI 1.19, 1.37; P <.001) and that for the group with severe distress was 1.54 (95% CI 1.37, 1.72; P <.001). Of note, the groups with mild-moderate and severe distress also differed significantly (P = .003).
We examined the interaction of psychological distress strata and year for total expenditures, outpatient expenditures, and outpatient visits to test whether changes in psychological distress over time were related to expenditures and utilization. Given that we found that psychological distress scores were relatively stable over time, it was not surprising to find that the interaction between psychological distress strata and year was not significant for total expenditures, outpatient expenditures, or office-based and outpatient visits.