Do somatic symptoms bias depression screening? Reliability and equivalence of PHQ-8 in those with and without chronic pain: A nationally representative study of U.S. adults

Abstract

The 8-item Patient Health Questionnaire (PHQ-8) is implemented into clinical practice and scientific research worldwide. However, in the context of chronic pain (CP), depression screening could be confounded by overlapping somatic symptoms. Inaccurate attribution of CP symptoms to depression could lead to overestimation of depressive symptom severity or even artifactual inflation of depression prevalence estimates. We aimed to determine the reliability and equivalence of PHQ-8 in adults with and without CP. This nationally representative survey study examined data from the 2019 National Health Interview Survey. The final analytic sample contained (n=30,983) U.S. adults who completed survey items on CP and the PHQ-8. Descriptives and visualizations were generated; reliability, cross-group equivalence, and measurement invariance were assessed. The PHQ was reliable; measurement invariance and cross-group equivalence were observed at the configural, metric, and scalar levels; differential item functioning was not observed. No evidence consistent with overestimation of depression prevalence or severity in the contexts of CP or HICP was observed.

The PHQ-8 is reliable and measures depressive symptoms equivalently in the context of CP, supporting the use of PHQ-8 to screen for depression in clinical practice, scientific research, and population health surveillance of adults with and without CP. CP is regularly observed to be associated with five to ten times greater prevalence of positive screens for clinically significant depressive symptoms. This study offers compelling evidence that this disparity reflects true differences in the prevalence of depression in those with chronic pain and is not attributable to measurement artifacts.

Introduction

Depression and chronic pain (CP) commonly co-occur in general and patient populations.[22] Among adults screening positive for depression, 56% have co-occurring CP, [16] while 40% of adults with CP have co-occurring depression.[1] About half of these—20% of adults with CP—have clinically significant depressive symptoms that remain insufficiently addressed.[16] Co-occurrence of CP and depressive symptoms is associated with higher prevalence of functional limitations in work, daily activities, and social participation, compared to those with either condition alone.[16] Addressing the unmet mental health needs of those with CP is essential for improved functioning and quality of life, [13] yet those with CP have 40% lower odds of receiving needed mental health treatment.[15]

Depression screening tools have been widely implemented into general health settings; today, most cases of depression are identified and addressed in primary care.[2] However, depression screening could be systematically biased in the context of CP given that several diagnostic criteria for depression (sleep disturbance, fatigue, appetite changes, psychomotor slowing) are also typical in non-depressed people with CP. Criterion contamination is an important concern: if CP symptoms were inaccurately attributed to depression, overestimation of depression severity, inappropriate treatment decisions, and artifactual inflation of epidemiological estimates could result.[3,33] Several prior studies have explored depression screening in the context of CP;[35,37] the majority were limited by small, non-representative samples or lack of relevant comparison groups. Findings are inconsistent: some found acceptable screening validity in those with CP [4,12,20,23] while others found support for psychometric concerns.[11,21,28,40]

The Patient Health Questionnaire-8 (PHQ-8) has been implemented into routine clinical practice worldwide; epidemiological estimates based on PHQ-8 are increasingly influential in mental health policy and resource allocation.[26] While the PHQ-8 has demonstrated excellent validity and reliability across geographies, settings, and contexts,[38] its equivalence in people with CP remains unclear. If the PHQ-8 functions differently in CP, clinical decisions, scientific research, and prevalence estimates could be substantially biased. Therefore, the equivalence of PHQ-8 to screen for depression in those with CP warrants rigorous investigation. This study used representative population data to evaluate the reliability and cross-group equivalence of PHQ-8 in the context of CP, addressing a persistent knowledge gap with important implications for clinical practice, research, and public health.

Methods

Data

The National Health Interview Survey (NHIS) data (n=31,997) is sampled and weighted to represent 244.6 million adults in the U.S. population.[31] The 2019 dataset contains all items of PHQ-8 and the chronic pain module. NHIS data are routinely used to estimate prevalence, incidence, trends, and disparities related to depressive symptoms and CP.[19,30,32]

Measures

Chronic pain (CP) was identified using the item “In the past 3 months, how often did you have pain? Would you say never, some days, most days, or every day?” – responses of “Most days” or “Every day” indicated CP. Among those with CP, high-impact chronic pain (HICP) was identified using the item: “Over the past 3 months, how often did your pain limit your life or work activities? Would you say never, some days, most days, or every day?” Responses of “Most days” or “Every day” indicated HICP while responses of “Never” or “Some days” to this item indicated Lower-Impact Chronic Pain (LICP). Definitions of CP and HICP align with the IASP classification of chronic pain for the International Classification of Diseases (ICD-11).[39] CP and HICP constitute this study’s primary analytic focus; however, their logically nested relationship could be confusing. LICP is included in select reporting to facilitate contextualized interpretation of mean estimates and to support representative, non-redundant visualizations of the U.S. adult population.

The National Health Interview Survey includes the PHQ-8 and not the PHQ-9; it is worth noting that scores and cutpoints are nearly equivalent for these measures.[41] PHQ-8 score was measured by summing the 8-item Patient Health Questionnaire (PHQ-8). The PHQ-8 asks respondents: “Over the last 2 weeks, how often have you been bothered by any of the following problems?” PHQ8–1 Anhedonia: “Little interest or pleasure in doing things”; PHQ8–2 Sadness/Blues: “Feeling down, depressed, or hopeless”; PHQ8–3 Sleep: “Trouble falling or staying asleep, or sleeping too much”; PHQ8–4 Fatigue: “Feeling tired or having little energy”; PHQ8–5 Appetite: “Poor appetite or overeating”; PHQ8–6 Self-Blame: “Feeling bad about yourself—or that you are a failure or have let yourself or your family down”; PHQ8–7 Concentration: “Trouble concentrating on things, such as reading the newspaper or watching television”; and PHQ8–8 Psychomotor Function: “Moving or speaking so slowly that other people could have noticed? Or the opposite—–being so fidgety or restless that you have been moving around a lot more than usual.” Responses are ordinal: (0) “Not at all,” (1) “Several Days,” (2) “More than Half the Days,” and (3) “Nearly Every Day.” Positive screens were determined using the standard cut point for PHQ-8 (≥10).[27] Negative screens included those with absent/minimal symptoms (PHQ-8 scores 0–4) or subclinical/mild symptoms (PHQ-8 scores 5–9); positive screens include those with moderate symptoms (PHQ-8 scores 10–14) or severe symptoms (PHQ-8 scores 15–24).[26]

For respondents with non-zero PHQ-8 scores, somatic symptoms proportion was calculated as the somatic symptom score divided by the PHQ-8 total score; those with zero PHQ-8 scores were coded as zero for the proportion. Somatic score was calculated by summing items 3, 4, 5, and 8— sleep, fatigue, appetite, and psychomotor functioning; cognitive/affective score was calculated by summing items 1, 2, 6, and 7—anhedonia, sadness, self-blame, and concentration. The maximum somatic score and the maximum cognitive/affective score are both 12 points each; thus, each has an equivalent maximum potential contribution to the total PHQ-8 score. This equivalence facilitates interpretation of the proportional contribution of somatic or cognitive/affective symptoms to determine whether either of the two scores are dominant.

Missing data

A total of 1,009 observations (3.15% of data) were missing (responses of “Refused,” “Not Ascertained,” or “Don’t Know”) from PHQ-8, CP, or both. Considering the study data characteristics, including a large sample size, minimal missing data, and minimal expectation of precision and bias loss in the context of observational data, which suggests that deviations from the missing at random (MAR) assumption are unlikely to affect the validity of estimates, complete case analysis was used.[29] We evaluated means for key study variables comparing observations with and without missing data; consistent means provided additional support that the MAR assumption is justifiable.

Study design and analysis plan

Preliminary descriptive analyses examined prevalence of positive PHQ-8 screens by CP status, distributions of PHQ-8 scores and item-level responses grouped by CP and depression status, and whether CP was associated with differences in the proportional contribution of somatic symptoms in PHQ-8 scores.

Reliability was assessed using Cronbach’s alphas and McDonald’s Omega coefficients; values greater than 0.80 were indicative of acceptable reliability.[36] Fit statistics were evaluated; Comparative Fit Index (CFI) and Tucker-Lewis Index (TLI) values greater than 0.95 were considered indicative of good model fit; Root Mean Square Error of Approximation (RMSEA) and Standardized Root Mean Square Residual (SRMR) values less than 0.05 indicate close fit and values less than 0.08 indicate reasonable model fit.[10,24]

Cross-group equivalence and measurement invariance were assessed using a stepwise approach. First, configural invariance was evaluated to determine whether the factor structure was equivalent across those with and without CP. Next, metric invariance was tested by constraining factor loadings to be equal across groups. Scalar (threshold) invariance was then examined by constraining both factor loadings and item thresholds to equality. Finally, residual invariance was assessed by additionally constraining residual variances to equality across groups. The process was repeated for those with and without HICP.

Standardized factor loadings and item-discrimination parameters using graded response models were assessed. Standardized factor loadings greater than 0.7 were indicative of “excellent” item performance; those greater than 0.5 and less than or equal to 0.7 were indicative of “good” item performance.[5] Discrimination parameters greater than or equal to 1.70 were considered “quite satisfactory;” those greater than or equal to 1.35 and less than 1.70 were considered “good.” [8] Item Response Theory (IRT) Test Information Functions (TIFs) and Item Information Functions (IIFs) were evaluated. Analyses were completed using R version 4.4.2.; several packages were used: ggplot2, survey, lavaan, and mirt.

Results

Preliminary Descriptive Analyses

Comparing Prevalence of Positive PHQ-8 screens

The prevalence of positive PHQ-8 screens was 6.9% in this NHIS sample representative of the U.S. adult general population. U.S. adults with CP or HICP had significantly higher prevalence of positive screens than those without CP (NoCP). Prevalence of positive screens was 3.5% in NoCP, compared to 20.1% in CP—reflecting prevalence of 11.5% in LICP and 34.8% in HICP (all p-values <0.001).

Comparing PHQ-8 scores

In the overall U.S. adult general population sample, the estimated mean PHQ-8 score was 2.48. Those with CP or HICP had significantly higher scores compared to those without CP. The mean score was 1.77 in NoCP, compared to 5.25 in CP, reflecting means of 3.85 in LICP and 7.71 in HICP (all p-values <0.001).

Next, we visualized distributions of PHQ-8 scores and item-level responses grouped by six mutually-exclusive and exhaustive combinations of CP status and depression screening status (Figure 1). Among those who screened negative, the mean PHQ-8 score was 1.61; the mean score was 1.33 in NoCP compared to 2.91 in CP—reflecting means of 2.58 in those with LICP and 3.72 in those with HICP (all p-values <0.001). Among those with negative screens, the distributions of PHQ-8 scores differed qualitatively across NoCP, CP, and HICP. The prevalence of absent/minimal symptoms was 90.9% in NoCP compared to 72.2% in CP—reflecting prevalence of 77.0% in LICP and 60.6% in HICP (all p-values <.001) (Figure 1, lower right). Among those who screened positive, the mean PHQ-8 score was 14.29; the mean was 13.83 in NoCP compared to 14.61 in CP — reflecting means of 13.65 in LICP and 15.12 in HICP (all p-values <.001). Among those with positive screens, the distributions of PHQ-8 scores were qualitatively similar across NoCP, CP, and HICP (Figure 1, upper right). Statistically significant CP-associated differences in mean PHQ-8 score existed among those with positive screens, yet the magnitude of the difference was attenuated in this group compared tothose who screened negative. The prevalence of severe symptoms was 34.6% in NoCP, and 43.7% in those with CP: reflecting prevalence of 33.9% in LICP and 49.3% in HICP (all p-values <0.001).

Fig 1. PHQ-8 item-level responses and total scores of U.S. adults, grouped by screening outcome and chronic pain status.

Mutually exclusive and exhaustive combinations of depression screening status and chronic pain status was used to facilitate visualization of the complete U.S. adult general population. Waffle plots: each unit square in 10 by 10 grids represents 1% of subgroup; color fill represents item-level severity (No fill = “Not at all”, Lightest color = “Several days”, Medium color = “More than half the days”, Darkest color = “Nearly every day”). Abbreviations: NoCP= without chronic pain, LICP = lower impact chronic pain; HICP= high-impact chronic pain. Data Source: National Center for Health Statistics, National Health Interview Survey, 2019.

Comparing individual PHQ-8 items

We visualized response distributions for each individual item of the PHQ-8 grouped by screening and CP status. (Figure 1, panels at left).

Among those who screened negative CP was associated with statistically and substantively significant increases in mean severity of the 8 individual items (scores ranging from 0 to 3): anhedonia (0.12 vs. 0.28), sadness (0.11 vs. 0.26), sleep (0.33 vs. 0.72), fatigue (0.41 vs. 0.89), appetite (0.14 vs 0.33) self-blame (0.09 vs. 0.18), concentration (0.08 vs. 0.18) and psychomotor functioning (0.03 vs. 0.08) (all p-values <0.001) (Figure 1, lower left).

However, among those who screened positive, CP-associated differences in symptom severity were statistically and substantively attenuated (Figure 1, upper left). CP was associated with statistically significantly (p<0.001) increases in mean severity of sadness (1.65 vs. 1.78), sleep problems (2.24 vs. 2.37), and fatigue (2.33 vs. 2.60); mean differences in severity of anhedonia, appetite, self-blame, concentration, and psychomotor functioning did not reach statistical significance. Statistically significant increases in severity were observed in HICP for seven of the PHQ-8 items (p<0.001); only concentration did not reach significance. In all combinations of screening status (negative vs. positive) and CP status (NoCP vs. LICP vs. HICP), mean severity of sleep and fatigue are at least twice the mean severity of other items (all p-values <0.001).

Comparing contribution of somatic symptoms

The proportional contribution of somatic symptoms to PHQ-8 total scores was calculated for all respondents. The mean proportion of somatic symptoms was highest in those with PHQ-8 total scores of 1, decreasing linearly as PHQ-8 total score increased.

Smoothed distributions of the somatic symptoms proportion were visualized within PHQ-8 total scores, and compared across NoCP, LICP, and HICP (Figure 2, panels at left). The individual observations utilized in calculating the smoothed distributions are also visualized—each dot represents the somatic symptoms proportion of an individual survey respondent (Figure 2, panels at right). The inverse linear relationship between the proportion of somatic symptoms and the PHQ-8 total score was observed regardless of chronic pain status.

Fig 2. Proportional contribution of somatic symptoms to PHQ-8 total scores, grouped by chronic pain status.

Mutually exclusive and exhaustive chronic pain status grouping was used to facilitate visualization of the complete U.S. adult general population. Color fill represents somatic symptoms proportion.

Abbreviations: NoCP= without chronic pain, LICP = lower impact chronic pain; HICP = high-impact chronic pain. Scatterplots, Panel at Right: each point represents a single survey respondent. Data Source: National Center for Health Statistics, National Health Interview Survey, 2019.

Among those who screened negative, observations with clear dominance of either somatic symptoms or cognitive/affective symptoms were regularly noted; however, among those who screened positive, observations with balanced proportions of somatic and cognitive/affective symptoms were most prevalent. This finding was uniform across NoCP, LICP, and HICP.

Among those who screened positive, the mean proportion of somatic symptoms was 0.55. The mean proportion was 0.53 in NoCP, compared to 0.55 in CP: reflecting proportions of 0.56 in LICP and 0.55 in HICP. The difference in somatic symptom proportions (0.019) was statistically significant (p < 0.008), given the large sample size, but the magnitude of the difference is trivial.

Reliability and Measurement Invariance

The PHQ-8 shows excellent internal consistency reliability; cross-group equivalence and measurement invariance are observed. The test of measurement invariance determines whether the CFA factor loading matrix is invariant across groups, i.e. whether the PHQ-8 items relate to the single latent trait of depression in the same way. Model fit indices supported cross-group equivalence and measurement invariance in CP (Table 1) and HICP (Supplemental Table 1) at the configural (factor structure), metric (factor loading), and scalar (item intercept) levels. The residual invariance model did not worsen fit compared with the scalar model, indicating that constraining residual variances across groups is reasonable. These findings confirm that the PHQ-8 measures depression consistently among U.S. adults regardless of CP status, including those with HICP.

Table 1.

Internal structure, reliability, cross-group equivalence and measurement invariance of the Patient Health Questionnaire (PHQ-8) comparing those with and without chronic pain (CP).

Internal Structure						Internal Consistency Reliability
	χ2	CFI	TLI	RMSEA (90% CI)	SRMR	Cronbach’s α	McDonald’s ω
No CP n=23,935	885.710	0.994	0.991	0.043 (0.040, 0.045)	0.041	0.82	0.86
CP n=7,083	560.092	0.993	0.991	0.062 (0.057, 0.066)	0.043	0.85	0.89
*Cross-Group Equivalence and Measurement Invariance: Comparing CP vs No CP
	χ2	CFI	TLI	RMSEA (90% CI)	SRMR
Configural	1445.802	0.994	0.991	0.048 (0.046, 0.050)	0.042
Metric	1471.063	0.994	0.992	0.044 (0.042, 0.046)	0.042
Scalar	1818.334	0.992	0.993	0.043 (0.041, 0.044)	0.044

Fit statistics: χ²: chi-square test; CFI: comparative fit index; TLI: Tucker–Lewis index; RMSEA: Root Mean Square Error of approximation; SRMR: Standardized Root Mean Square Residual; χ² with 40 degrees of freedom and p < 0.001 for configural invariance model; χ² with 47 degrees of freedom and p < 0.001 for metric invariance model; χ² with 54 degrees of freedom and p < 0.001 for scalar invariance model.

^*Cross-group equivalence tests use multi-group confirmatory factor analysis with the full sample (n=31,018), applying progressively constrained nested models. Configural invariance: same factor structure across groups. Metric invariance: equal factor loadings. Scalar invariance: equal item intercepts. Higher degrees of freedom reflect simultaneous analysis of both groups with increasing constraints. Excellent fit across all three levels supports that the PHQ-8 measures depression equivalently across chronic pain groups, enabling valid between-group comparisons.

Data Source: National Center for Health Statistics, National Health Interview Survey, 2019

CFA Standard Factor Loadings and IRT Discrimination Parameters

CFA and IRT models are conceptually equivalent, yet the interpretation of standard factor loadings and discrimination parameters is complementary;[34] therefore, we report both (Table 2). Higher standard factor loadings indicate a greater shared variance between the item and the underlying factor. All standardized factor loadings (λ) exceeded 0.660 across groups, indicating “very good” to “excellent” item performance with substantial shared variance between each PHQ-8 item and the underlying depression factor.

Table 2.

Confirmatory Factor Analysis (CFA) standardized item loadings and Item Response Theory (IRT) discrimination parameters of Patient Health Questionnaire (PHQ-8) in U.S. adults, grouped by chronic pain status.

CFA Standardized Item Loadings
		anhedonia	sadness	sleep	fatigue	appetite	self-blame	concentration	psychomotor
	n	λ1	λ2	λ3	λ4	λ5	λ6	λ7	λ8
NoCP	23,925	0.841	0.875	0.714	0.755	0.744	0.839	0.783	0.759
CP	7,063	0.833	0.876	0.755	0.731	0.731	0.834	0.777	0.727
HICP	2,611	0.812	0.884	0.665	0.734	0.685	0.823	0.754	0.690
IRT Discrimination Parameters
		anhedonia	sadness	sleep	fatigue	appetite	self-blame	concentration	psychomotor
	n	α1	α2	α3	α4	α5	α6	α7	α8
NoCP	23,925	2.821	3.167	1.810	2.085	2.109	2.978	2.389	2.290
CP	7,063	2.710	3.144	1.528	1.931	1.883	2.829	2.244	1.951
HICP	2,611	2.523	3.277	1.442	1.737	1.615	2.724	2.044	1.712

Abbreviations: NoCP= without chronic pain, CP= chronic pain, HICP = high-impact chronic pain.

Data Source: National Center for Health Statistics, National Health Interview Survey, 2019

Standardized item loadings: > 0.7 = “excellent” item performance; ≤ 0.7 and > 0.5 = “good” item performance.

Discrimination parameters: ≥ 1.70 = “quite satisfactory” discriminative ability; < 1.70 and ≥ to 1.35 = “good” discriminative ability.

The NoCP group demonstrated the strongest factor structure (λ: 0.714–0.875), indicating excellent performance of all items (λ ≥ 0.71). The CP group showed comparable results (λ: 0.727–0.876) with eight items reaching “excellent” performance. The HICP group exhibited the most variability (λ: 0.665–0.884) with five items “excellent” and three “good.” Sadness demonstrated the highest factor loading (λ: 0.875–0.884) across groups.

IRT discrimination parameters reflect an item’s ability to discriminate severity of depression. All items demonstrated “good” to “quite satisfactory” discrimination across groups (α :1.442 to 3.277). Most items achieved quite satisfactory performance (α ≥ 1.70).

Sleep achieved relatively lower, though still “good” (α ≥ 1.35) discrimination ability in CP and HICP, while appetite, fatigue, and psychomotor functioning achieved “good” discrimination in HICP. This suggests that these somatic symptoms may be relatively less discriminant to identify different levels of depression among those with HICP, but does not suggest problems with comparing groups on the same depression trait dimension.[34]

TIFs show that the PHQ-8 provides acceptable measurement precision across depression severity levels (theta, θ), with nearly identical curves across the three groups (Figure 3, left panel). Theta represents latent depression severity on a standardized scale (0 = average; negative = below average; positive = above average). Information peaks around θ = 0 to +1, indicating optimal precision for mild to moderate depression. The overlapping TIF curves demonstrate equivalent measurement precision regardless of chronic pain status. IIFs show consistent patterns across groups (Figure 3, right panels), with sadness demonstrating the highest discriminative ability.

Fig 3. PHQ-8 Test Information Functions (TIFs) and Item Information Functions (IIFs), grouped by chronic pain status.

TIFs (left panel) display total measurement precision across depression severity (theta, θ); IIFs (right panels) show each item’s contribution. Theta is a standardized scale where 0 = average severity, negative values = below average, positive values = above average. Higher information indicates greater precision. Overlapping curves demonstrate equivalent precision across groups.

Abbreviations: NoCP = without chronic pain, CP = chronic pain; HICP= high-impact chronic pain. Data Source: National Center for Health Statistics, National Health Interview Survey, 2019.

Discussion

This representative population study of U.S adults (n = 30,983) found that the PHQ-8 reliably and equivalently measures depressive symptoms in those with and without CP. Study findings should enhance confidence among clinicians and researchers that the PHQ-8 is suitable to screen for depression in people with CP and that the interpretation of results is straightforward. Further, study findings offer compelling evidence that comparing PHQ-8 total scores or individual PHQ-8 items between people with and without CP is scientifically appropriate and analytically valid. Finally, these findings persuasively suggest that the very high estimated prevalence of positive PHQ-8 screens among those with CP reflect true differences in the burden of clinically significant depressive symptoms and are not attributable to measurement artifacts.

Consistent with prior studies, the prevalence of positive PHQ-8 screens was 5 to 10 times higher in U.S. adults with CP or HICP compared to others. People with CP were observed to have elevated severity of several somatic items including fatigue and sleep disturbance compared to the severity of other PHQ-8 items. However, comparisons across those with and without chronic pain revealed that this cross-item pattern is consistent regardless of CP status.

We evaluated whether CP-associated differences existed in the proportional contribution of somatic symptoms, holding PHQ-8 scores constant. Study data did not support the possibility that somatic (nor any other) symptoms may exert undue influence in the PHQ-8 screening outcomes of people with CP. The mean proportion of somatic symptoms decreased linearly as PHQ-8 score increased; substantively meaningful cross-group differences were not observed. Consistent across groups, proportional dominance of somatic symptoms was frequently observed in those who screened negative on PHQ-8, while balanced contributions of somatic and cognitive/affective symptoms were typical among those who screened positive. Overall, the results of rich descriptive analyses, evaluation of fit statistics, tests of measurement invariance and cross-group equivalence were consistent; factor loadings and item discrimination coefficients were likewise uniform, and no differential global or item functioning was observed. The National Health Interview Survey includes the PHQ-8 and not the PHQ-9; it is worth noting that scores and cutpoints are nearly equivalent for these measures.[41]

CP was associated with statistically significant increases in severity of all items and PHQ-8 total scores. However, we suggest cautious interpretation of this difference, as differences in mean severity were disproportionately driven by those who screened negative on PHQ-8. Among those who screened positive, cross-group differences in severity of individual items and total scores were attenuated or eliminated; however, in the context of HICP, significant, if substantively attenuated, differences were preserved for 7 of 8 items.

This study has limitations. First, persons with no fixed address, living on military bases, in long-term care facilities, or in correctional facilities are not reached by the NHIS. Second, though the PHQ-8 items are identical to those used in clinical practice, it is possible that screening data collected by NHIS could be less accurate than screening data collected during a clinical encounter.

For decades, the binary opposition of pain and depression has been misaligned with modern scientific understanding about the biopsychosocial nature of pain. Scientific consensus is coalescing that dualistic assignment of symptoms to either CP or depression could be clinically and scientifically counterproductive;[11,13] indeed, the International Association for the Study of Pain (IASP)’s definition of pain explicitly presumes the accompaniment of negative affective states.[39]

Addressing the unmet mental health needs of people with CP is recognized as a fundamental human right and a key priority for global population health, therefore, expanded access to pain care and mental health care is urgently needed.[9,14] Yet, those with CP face end-to-end disparities across the mental health patient journey—including higher prevalence of unmet mental health needs, lower likelihood that unmet mental health needs will be identified, and lower likelihood that mental health care will be received.[15] Those who screen positive on PHQ-8 benefit from equitable access to information, resources, and support, regardless of CP status. Broad implementation of accessible, CP-informed mental health supports, person-centered clinical communication, and interventions to address CP-associated mental health stigmas are needed.[6,7,17,18,25]

Highlights.

*Depression screening is not biased by overlapping symptoms in depression and chronic pain.

*Interpretation of screening results in those with chronic pain is straightforward.

*Those with chronic pain experience profound mental health disparities.

*Mental health stigmas specific to the lived experience of chronic pain should be addressed.

Data Availability Statement:

All data used in this study are collected by the National Health Interview Survey, de-identified, and made publicly available on the National Center for Health Statistics’ website.