Comorbid Depression, Chronic Pain, and Disability in Primary Care
BRUCE A. ARNOW, PHD, ENID M. HUNKELER, MA, CHRISTINE M. BLASEY, PHD, JANELLE LEE, DRPH,
MICHAEL J. CONSTANTINO, PHD, BRUCE FIREMAN, MA, HELENA C. KRAEMER, PHD, ROBIN DEA, MD,
REBECCA ROBINSON, MS, AND CHRIS HAYWARD, MD, MPH
Objectives: The objectives of this study were to provide estimates of the prevalence and strength of association between major
depression and chronic pain in a primary care population and to examine the clinical burden associated with the two conditions,
singly and together. Methods: A random sample of Kaiser Permanente patients who visited a primary care clinic was mailed a
questionnaire assessing major depressive disorder (MDD), chronic pain, pain-related disability, somatic symptom severity, panic
disorder, other anxiety, probable alcohol abuse, and health-related quality of life (HRQL). Instruments included the Patient Health
Questionnaire, SF-8, and Graded Chronic Pain Questionnaire. A total of 5808 patients responded (54% of those eligible to
participate). Results: Among those with MDD, a significantly higher proportion reported chronic (i.e., nondisabling or disabling)
pain than those without MDD (66% versus 43%, respectively). Disabling chronic pain was present in 41% of those with MDD
versus 10% of those without MDD. Respondents with comorbid depression and disabling chronic pain had significantly poorer
HRQL, greater somatic symptom severity, and higher prevalence of panic disorder than other respondents. The prevalence of
probable alcohol abuse/dependence was significantly higher among persons with MDD compared with individuals without MDD
regardless of pain or disability level. Compared with participants without MDD, the prevalence of other anxiety among those with
MDD was more than sixfold greater regardless of pain or disability level. Conclusions: Chronic pain is common among those with
MDD. Comorbid MDD and disabling chronic pain are associated with greater clinical burden than MDD alone. Key words: major
depression, chronic pain, comorbidity, disability, epidemiologic comorbidity.
MDD ? major depressive disorder; HRQL ? health-related quality of
life; HMO ? health maintenance organization; PHQ ? Patient Health
Questionnaire; GCPS ? Graded Chronic Pain Scale; CP ? chronic
pain; DCP ? disabling chronic pain; GAD ? generalized anxiety
disorder; SCID ? Structured Clinical Interview for DSM-III-R;
PRIME-MD ? Primary Care Evaluation of Mental Disorders; CI ?
confidence interval; DSM-IV ? Diagnostic and Statistical Manual of
Mental Disorders, Fourth Edition; DSM-III-R ? Diagnostic and Sta-
tistical Manual of Mental Disorders, Third Edition Revised.
Findings from the World Health Organization Collaborative
and the National Comorbidity Studies revealed that a substan-
tial majority of patients with MDD also met criteria for at least
one or more concurrent psychiatric disorders (1,2). There is
also growing evidence that MDD and chronic pain frequently
However, the strength of the relationship between chronic
pain and MDD is unclear. This is in part because of variability
in case definitions for both conditions (5). Other factors lead-
ing to inconsistency of research results on the association
between depression and pain include small sample sizes and
sampling from tertiary care settings restricted to select groups
of patients with chronic pain (6). In a recent review of de-
pression and pain comorbidity, Bair et al. (7) reported highly
omorbid conditions are the norm rather than the exception
among patients with major depressive disorder (MDD).
variable estimates of the association between the two condi-
tions; the prevalence of pain in patients presenting with de-
pression ranged from 15% to 100%, whereas the prevalence of
depression among patients with pain was 1.5% to 100%.
Most cases of depression are treated in primary care set-
tings (8). Recognition of depression in primary care is com-
plicated by the presentation of somatic complaints (9), a
majority of which are pain-related (10). A substantial majority
of patients with chronic pain are also treated in primary care
settings (11), yet few studies investigating the prevalence of
comorbid chronic pain and depression have been conducted in
primary care settings.
Furthermore, the clinical burden of comorbid depression
and chronic pain remains largely unexplored. Although both
depression (12) and chronic pain (13) are known to be inde-
pendently associated with decrements in quality of life and
increased somatic preoccupation (10), the impact of the pres-
ence of both conditions on these outcomes has rarely been
investigated. Similarly, little is known about the association of
comorbid depression and pain with other psychiatric condi-
This study involved two main objectives. First, we pro-
vided estimates in a primary care patient population of the
strength of the association between MDD and chronic pain.
We hypothesized nonrandom co-occurrence, or epidemiologic
comorbidity (14), i.e., that those with chronic pain would be
significantly more likely to meet criteria for MDD than those
without pain and, conversely, that those with MDD would be
significantly more likely than those without MDD to report
chronic pain. Second, we investigated the impact of MDD and
chronic pain, singly and together, on quality of life, severity of
somatic symptoms, and three psychiatric conditions: panic
disorder, other anxiety disorder, and probable alcohol abuse/
dependence. We included two levels of pain-related disability
in our analyses to capture differences in the extent to which
pain interferes with everyday activity (15). We hypothesized
that respondents reporting MDD and disabling (i.e., moderate
From the Department of Psychiatry and Behavioral Sciences (B.A.A.,
C.M.B., M.J.C., H.C.K., C.H.), Stanford University School of Medicine,
Department of Psychiatry and Behavioral Sciences, Stanford, California;
Kaiser Permanente Division of Research (E.M.H., J.L., B.F., R.D.), Oakland,
California; and Eli Lilly & Company, Indianapolis, Indiana (R.R.).
Address correspondence and reprint requests to Bruce Arnow, PhD, Stan-
ford University, Department of Psychiatry and Behavioral Sciences, 401
Quarry Road, Stanford, CA 94305-5722. E-mail: email@example.com
Received for publication April 21, 2005; revision received September 29,
This study was supported by a grant from Eli Lilly and Company to the first
Copyright © 2006 by the American Psychosomatic Society
Psychosomatic Medicine 68:262–268 (2006)
to high activity interference) chronic pain would report a
significantly lower quality of life, a significantly higher level
of somatic symptom severity, and a significantly higher prev-
alence of panic disorder, other anxiety, and probable alcohol
abuse/dependence than all other groups.
Over a 12-week period in 2002, a total of 12,000 members of Kaiser
Permanente, a health maintenance organization (HMO) in Northern Califor-
nia, were randomly selected within 1 week of a primary care visit to receive
a mailing that included a cover letter, a consent form with an option to decline
participation, and self-report questionnaires. Participants were drawn from 31
outpatient clinics in Northern California representing urban, suburban, and
rural areas. In addition to being a member of the HMO and visiting an internal
medicine or family practice clinic, inclusion criteria included being between
the ages of 21 to 75, being sufficiently literate in English to complete the
self-report questionnaires, and having an available current home address and
phone number. Exclusion criteria included a diagnosis of schizophrenia or
other psychotic disorder, a diagnosis of bipolar disorder, dementia, pervasive
developmental disorder, or pregnancy-related diagnosis over the prior 1-year
period. The Institutional Review Boards of both Stanford University Medical
Center and Kaiser Foundation Research Institute approved the study.
If there was no response to the first mailing, a second full packet was
mailed out after 21 days. At 35 days, if there was no response, potential
respondents were contacted by phone and offered the opportunity to decline
participation, complete the questionnaire by telephone, or fill it out and return
it by mail.
Of the individuals sampled, a total of 1290 (11%) were ineligible as a
result of literacy barriers (n ? 428) or incorrect contact information (n ?
862). Of the 10,710 eligible participants, 3808 (35.6%) refused to participate
and 1094 (10.2%) failed to respond, leaving 5808 members (54% of those
eligible) as study respondents.
The primary mental health measure we used was the Patient Health
Questionnaire (PHQ) (16). The PHQ is a self-administered version of the
Primary Care Evaluation of Mental Disorders (PRIME-MD) (17) and pro-
vides Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition
(DSM-IV) diagnoses of MDD as well as diagnoses of several other psychi-
atric disorders (e.g., panic disorder), subthreshold disorders (e.g., probable
alcohol abuse), and somatic symptom severity. Its diagnostic validity is
comparable to the PRIME-MD and it has been validated in primary care
We included the eight-item version of the depression module (PHQ-8)
that omits the original item of suicide, which may be more appropriate in
mailed questionnaires that are not reviewed by a clinician for purposes of
treatment (18). For each item, respondents were asked how often, over the last
2 weeks, they had been bothered (“not at all,” “several days,” “more than half
the days,” or “nearly every day.” Items were scored zero (“not at all”) to 3
(“nearly everyday”). To qualify for a diagnosis of MDD in this study,
respondents had to endorse being bothered by five or more of eight symptoms
more than half the days (i.e., a score ?10) for the previous 2 weeks with at
least one of those five items, including depressed mood or anhedonia. Scores
?10 on the PHQ-8 have ?99% sensitivity and 92% specificity for identifying
patients with major depression when compared with interviews with health
professionals using questions from the Structured Clinical Interview for
DSM-III-R (SCID) and the PRIME-MD (19).
Somatic Symptom Severity
The severity of somatic complaints often encountered in primary care was
assessed using the PHQ-15 (20). Respondents were asked how much they
were bothered (“not bothered,” “bothered a little,” or “bothered a lot”) in the
last month by 13 of the original 15 symptoms (two items queried in the
depression section were omitted). We assessed the number of physical symp-
toms by summing the number among those 13 physical symptoms to which
each respondent endorsed as “bothered a lot.” Although the measure does not
permit discriminating between explained and unexplained symptoms, the total
number of physical complaints is a predictor of somatoform disorders (21).
The PHQ scales were used to determine whether respondents met thresh-
old criteria for panic disorder, as well as another anxiety disorder, within the
last 4 weeks.
Probable Alcohol Abuse or Dependence
To meet PHQ criteria for probable alcohol abuse/dependence, the respon-
dent had to report one of a number of problematic behaviors over the past 6
months (e.g., “you had a problem getting along with other people while you
Respondents were asked whether they were “currently troubled by pain or
discomfort, either all the time or on and off.” A separate question assessed
whether the pain had persisted for more than 6 months. Participants who
endorsed both of these were defined as suffering from chronic pain. This
definition is generally consistent with the one advanced by The International
Association for the Study of Pain (22), with two exceptions. First, following
the recommendations of Elliott et al. (23), we included respondents with
intermittent pain. Second, consistent with other recent studies (3), we used the
more conservative 6-month threshold rather than 3 months.
Pain-related severity and disability were assessed with the Graded
Chronic Pain Scale (15), a seven-item questionnaire that measures both pain
intensity and interference with daily activities. The Graded Chronic Pain
Scale (GCPS) has a hierarchical structure allowing classification of respon-
dents into one of four classes: grade 1 (low intensity, low interference), grade
II (high intensity, low interference), grade III (moderate interference), or
grade IV (severe interference). In the current study, those who met criteria for
grades III and IV on the GCPS were defined as having disabling pain.
Health-Related Quality of Life
Health-related quality of life (HRQL) was assessed with the SF-8 (24).
This eight-item instrument uses a single item to query participants about each
of the following: general health, physical functioning, physical role, bodily
pain, vitality, social functioning, mental health, and emotional role. Although
validation studies of the SF-8 suggested the presence of two summary factors,
a factor analysis for the current sample indicated a unitary factor structure
(eigenvalue ? 4.6, 57% of variance explained). Thus, we report a single
summary score. Higher scores denote a lower quality of life.
A logistic regression model was used to test for the comorbidity of
depression, chronic pain, and disabling chronic pain while accounting for the
effects of age, gender, and all two-way interactions. Gender was coded
?1/2 ? females and ?1/2 ? males, and all ordinal predictors were centered
at their means. Disabling chronic pain was nested within chronic pain. Thus,
simple effects represent the association of the predictor with the outcome at
the centered values of all other predictors.
For comparison, six clinically defined groups were then formed: no pain
or major depression (neither), nondisabling chronic pain only (CP only),
disabling chronic pain only (DCP only), major depression only (MDD only),
major depression and comorbid nondisabling chronic pain (MDD plus CP),
and major depression and comorbid disabling chronic pain (MDD plus DCP).
Analyses of covariance, adjusting for age and gender, were used to compare
the six groups on HRQL and somatic symptom severity. Three logistic
COMORBID DEPRESSION IN PRIMARY CARE
263Psychosomatic Medicine 68:262–268 (2006)
regression analyses tested the effects of group membership, adjusting for age
and gender, on the presence/absence of other psychiatric conditions, including
probable alcohol abuse/dependence, other anxiety, and panic disorder. Group
comparisons were tested using deviation contrasts, which compare the mean
of a specified group with the grand mean of other specified groups. All
hypothesis testing was conducted at the .05 significance level (two-tailed).
Demographics and Response Bias
Survey respondents (n ? 5808) ranged in age from 21 to 75
years (mean ? 53.3, standard deviation ? 13.6). Fifty-seven
percent (57.6%) were female. Descriptive statistics are pre-
sented in Table 1. Compared with those who were ineligible
and nonrespondents (n ? 6192), study respondents were older
(mean age 53.3 years [13.6] versus 49.5 years [14.1], p ?
.001) and more likely to be female than male (57.6% versus
54.6%, p ? .01), respectively.
Depression, Pain, and Epidemiologic Comorbidity
Of the 5808 respondents, 413 (7.1%; 95% confidence in-
terval [CI]: 6.4–7.7%) met criteria for MDD, 1887 (32.5%;
95% CI: 31.3–33.7%) reported experiencing nondisabling
chronic pain, and 731 (12.6%; 95% CI: 11.8–13.4%) reported
disabling chronic pain. The six groups differed significantly in
mean age (F ? 20.4, df ? 5,5808, p ? .0001). Scheffe post
hoc tests indicated that the two groups with pain and no
depression (i.e., CP only and DCP only) were significantly
older than the other four groups. The three groups meeting
criteria for MDD had higher proportions of females than the
other groups (X  ? 62.9, df ? 5, p ? .0001).
Compared with those without chronic pain, respondents
who reported pain were more likely to report depression (DCP
only: odds ratio [OR] ? 5.4, p ? .0001; CP only: OR ? 3.0,
p ? .001). Younger patients and females were also more
likely to be depressed (age: OR per year ? 0.9, p ? .0001,
gender: OR ? 0.6, p ? .0001). No two-way interactions were
significant indicating that the degree of comorbidity between
pain and depression did not differ according to age or gender.
Figure 1 shows the prevalence of chronic pain among those
with and without MDD.
The Clinical Burden Associated With Depression and
Health-Related Quality of Life and Somatic Symptoms
Respondents with MDD plus DCP had significantly worse
HRQL than all other groups (F ? 485.9, df ? 7,5558, p ?
.0001; p ? .0001 all five contrasts).1A similar pattern was
detected regarding somatic symptom severity (F ? 359.6,
df ? 7,5808, p ? .0001; all five contrasts p ? .0001).2Scheffe
post hoc tests further elucidated group differences. Group
means and standard deviations for the HRQL and somatic
symptoms measures, as well as the Scheffe test results, are
presented in Table 2.
Criteria for panic disorder were met by 3.6% (211) of the
sample, whereas 6.2% (359) met criteria for other anxiety and
6.0% (348) met criteria for probable alcohol abuse/depen-
dence. Table 3 shows the prevalence of these psychiatric
conditions across the six groups.
In the logistic regression model for panic, group membership
was a significant predictor (Wald ? 302.2, df ? 5, p ? .0001).
Those most likely to report panic were younger (B ? ?0.017,
1Because the HRQL contains two items that query pain and emotional
problems, the group comparison was also done without these items. Similar to
the eight-item version, the six-item version had a unitary factor structure. The
same pattern of results was observed, with the MDD plus DCP group having
the poorest HRQL (all p’s ? .01).
2Because the Somatic Symptom Severity scale contained several items that
query pain (e.g., back pain, stomach pain), the group comparison was also
done without these items. The same pattern of results was observed, with the
MDD plus DCP group having the highest somatization scores (all p’s ? .001).
TABLE 1. Demographic Characteristics of Study Respondents
(n ? 5808)
(n ? 3048)
(n ? 1786)
(n ? 561)
(n ? 142)
MDD ? Nondisabling
(n ? 101)
MDD ? Disabling
(n ? 170)
Female gender (%)
Native American (%)
East Indian (%)
53.3 ? 13.652.1 ? 14.355.6 ? 12.334.3 ? 13.049.6 ? 13.551.8 ? 13.250.9 ? 11.9
MDD ? major depressive disorder.
B. A. ARNOW et al.
264Psychosomatic Medicine 68:262–268 (2006)
df ? 1, p ? .002, OR ? 0.98) and female (B ? ?0.53, df ? 1,
p ? .002). A deviation contrast comparing the depressed groups
versus nondepressed groups showed the impact of depression
(B ? 2.4, df ? 1, p ? .001, OR ? 11.9). A second deviation
contrast showed that the prevalence of panic was significantly
higher in the group with MDD plus DCP (B ? 0.85, df ? 1, p ?
When predicting other anxiety, group membership (Wald ?
718.9, df ? 5, p ? .0001) was again a statistically significant
predictor when covarying for age and gender (age: B ?
?0.02, df ? 1, p ? .0001, OR ? 0.98; gender: B ? ?0.25,
df ? 1, p ? .07, OR ? 0.78). The deviation contrast indicated
a substantially higher prevalence of other anxiety among the
three groups with MDD relative to those without depression
(B ? 3.5, df ? 1, p ? .0001, OR ? 33.6). Unlike the model
predicting panic, the deviation contrast comparing those with
MDD plus DCP versus the other respondents with depression
was not statistically significant (p ? .90).
Again, group membership (Wald ? 16.8, df ? 1, p ? .005)
was a statistically significant predictor when covarying for the
effects of age and gender (age: B ? ?0.03, df ? 1, p ? .0001,
OR ? 0.97; gender: B ? 1.4, df ? 1, p ? .0001, OR ? 3.99).
The deviation contrast indicated that the three depressed
groups were more likely to report alcohol abuse than the
TABLE 3. Prevalence of Psychiatric Syndromes by Group
(n ? 3048)
(n ? 1786)
(n ? 561)
(n ? 142)
MDD ? Nondisabling
(n ? 101)
MDD ? Disabling
(n ? 170)
MDD ? major depressive disorder.
Figure 1. Comorbidity of pain and depression: pain status among respondents with and without major depressive disorder.
TABLE 2.Group Means (?standard deviation) on Measures of Quality of Life and Somatization
Chronic Pain Only
Pain OnlyMDD Alone
MDD ? Nondisabling
MDD ? Disabling
(N ? 3048; 52.5%) (N ? 1786; 30.8%)(N ? 561; 9.7%)(N ? 142; 2.4%)(N ? 101; 1.7%)(N ? 170; 2.9%)FpContrasts
16.8 ? 6.119.4 ? 5.126.0 ? 5.2 26.7 ? 6.427.9 ? 4.931.3 ? 4.3 493b
?.0001 6 ? 5 ? 4 ?
3 ? 2 ? 1
Som0.6 ? 1.1 1.1 ? 1.3 2.4 ? 1.72.5 ? 2.22.8 ? 2.13.8 ? 2.3 376c
?.0001 6 ? 5 ? 4 ?
3 ? 2 ? 1
aHigher scores denote lower quality of life.
bANCOVA test statistic, main effect of group (df ? 7,5558) adjusting for age and gender, data missing at random.
cANCOVA test statistic, main effect of group (df ? 7,5808) adjusting for age and gender.
MDD ? major depressive disorder; HRQL ? health-related quality of life; ANCOVA ? analysis of covariance.
COMORBID DEPRESSION IN PRIMARY CARE
265Psychosomatic Medicine 68:262–268 (2006)
nondepressed groups (B ? 0.71, df ? 1, p ? .0001, OR ?
2.03). Respondents with MDD plus DCP did not have a higher
prevalence of alcohol abuse relative to other respondents with
depression (p ? .74).
Our study had two aims: 1) examining the prevalence and
strength of association between chronic pain and MDD in
primary care patients, and 2) describing the clinical burden of
comorbid MDD and chronic pain. Regarding aim 1, we found
evidence of a nonrandom association between depression and
chronic pain, i.e., epidemiologic comorbidity (14). Approxi-
mately two thirds of those meeting criteria for MDD also
reported chronic pain and, among those with chronic pain, the
prevalence of those with MDD was significantly higher than
was observed among those without chronic pain. Regarding
aim 2, evidence that the clinical burden associated with MDD
and chronic pain is significantly greater than for those with
either condition alone was mixed. Specifically, respondents
with MDD plus DCP had significantly poorer HRQL and
greater somatic symptom severity than all other groups. Ad-
ditionally, the prevalence of panic disorder among those with
MDD plus DCP was approximately twice as high as in the
other two depressed groups and four times as high as the
prevalence among those with disabling pain who did not meet
criteria for MDD. However, the prevalence of probable alco-
hol abuse/dependence and other anxiety was substantially
higher among those with MDD compared with those without
MDD regardless of pain or disability level.
The prevalence of MDD in our sample was 7.1%. Our
findings are similar to those (i.e., 7.3%) reported by Olfson
and colleagues (25) whose sample was drawn from the same
HMO and used a modified version of the Structured Clinical
Interview for DSM-III-R (26). In general, the prevalence of
MDD in primary care samples is 5% to 10% (27).
Chronic pain was present in 45% of our sample. In an
epidemiologic study in the United Kingdom (23) that used a
similar, although less restrictive (i.e., 3 months versus 6
months), definition of pain, the prevalence of chronic pain was
50%. Also, similar to Elliott and colleagues, we found that
nearly 28% of those with chronic pain reported moderate to
severe disability on the Graded Chronic Pain Scale (15).
We found that a substantial majority of those patients who
met criteria for MDD also reported chronic pain. Specifically,
among participants with MDD, 41% reported disabling chronic
pain and 25% nondisabling chronic pain. Among those not meet-
ing criteria for MDD, 10% reported having disabling chronic
pain and 33% reported nondisabling chronic pain. Although
fewer studies have investigated the prevalence of chronic pain
symptoms among those with MDD than the converse (7), our
findings are consistent with growing evidence suggesting that
a large percentage of patients with MDD experience comorbid
symptoms of chronic pain (3). Moreover, our findings re-
vealed that pain was more likely to be disabling when MDD
is present. Among those reporting chronic pain but not MDD,
23.9% met criteria for moderate to severe disability, whereas
among those with MDD and chronic pain, 62.7% reported
disability in the moderate to severe range.
Although we found clear evidence of epidemiologic co-
morbidity (14) or nonrandom co-occurrence of MDD and
chronic pain, the magnitude of the relationship between
chronic pain and MDD differed depending on which group of
participants was examined. Among those with MDD, nearly
66% reported chronic pain. Among participants reporting any
chronic pain, the overall percentage of those who also met
criteria for MDD was 10.4%, whereas the prevalence of MDD
among those without pain was 4.5%. However, among those
with chronic pain, the prevalence of MDD was considerably
different depending on whether the individual had nondis-
abling (5.4%) or disabling (23.3%) chronic pain. Although our
findings underscore the importance of disability in the pain–
depression association, the direction of effect is unclear. It is
possible that those who are disabled by pain become de-
pressed, and it is possible that those who are depressed are
more likely to become disabled.
Our findings on HRQL are consistent with those of others
who have reported significant decrements in quality of life
associated with depressive disorders (12). However, we also
found that when depression and disabling chronic pain are
comorbid, the HRQL is even lower. In addition, those with
disabling chronic pain who were not depressed showed HRQL
impairment that was similar to those with MDD alone and
MDD plus CP.
The pattern of findings on somatic symptom severity was
similar for the MDD plus DCP group. That is, we found
significant differences in the hypothesized direction among
those with MDD plus DCP compared with all other groups.
Consistent with findings that depressed patients in primary
care settings often present somatic complaints (28), depressed
participants without pain were as high on somatic symptom
severity as nondepressed patients with disabling chronic pain.
The prevalence of panic disorder in our sample was 3.6%.
Our findings on the prevalence of panic among those with
MDD only, as well as MDD plus CP, are similar to those
reported from the National Comorbidity Study data (29).
However, the prevalence of panic among those with MDD
plus DCP (31.2%) was nearly twice as high as in the two other
One explanation for the striking comorbidity of panic dis-
order among those with MDD plus DCP is that catastrophic
thinking may be a feature of both panic disorder and disabling
chronic pain. A number of investigators have conceptualized
panic in terms of catastrophic appraisal of somatic sensations
associated with elevated autonomic arousal (30). In addition,
catastrophic thoughts regarding pain are related to increased
pain intensity and disability, often independent of physical
impairment (31). Although anxiety researchers and pain in-
vestigators have independently found catastrophic thinking to
be a factor in appraising autonomic and painful somatic cues,
respectively, it is possible that the propensity for such thinking
extends to both types of experiences.
B. A. ARNOW et al.
266Psychosomatic Medicine 68:262–268 (2006)
The prevalence of other anxiety in our sample was 6.2%.
This is similar to the 7% PHQ-derived prevalence rate in the
PHQ Primary Care Study (16). Unlike the findings for panic
disorder, those with MDD plus DCP were not different than
the other groups with MDD.
Consistent with prior studies (16,25), the prevalence of
probable alcohol abuse/dependence in our sample was 6%. In
the absence of MDD, we did not find that the presence of
chronic pain, whether disabling or nondisabling, was associ-
ated with increased prevalence of alcohol abuse. The three
depressed groups reported significantly more alcohol abuse/
dependence than the nondepressed groups, but again, neither
presence nor level of pain was associated with increased
prevalence. Thus, in our sample, MDD was associated with
increased levels of alcohol use, but chronic pain was not.
Strengths of our study included: 1) examining both the
prevalence of chronic pain among those with depression and,
conversely, the prevalence of depression among those with
chronic pain in the same sample; 2) sampling in primary care,
in which the majority of these patients are encountered; and 3)
examination of the clinical burden associated with comorbid
MDD and chronic pain, singly and together.
There are several limitations to be considered in interpret-
ing our findings. First, our sample is restricted to HMO
members who made a recent visit to their physician. The
extent to which the findings may be generalized to uninsured
populations or HMO patients in general is unclear. Second, we
exclusively used self-report measures as opposed to interview-
administered assessment. However, our findings on the prev-
alence of the psychiatric disorders we assessed were similar to
other studies. Third, a substantial number of potential partic-
ipants did not return questionnaires. Compared with nonre-
spondents, our participants were somewhat older and more
likely to be female. We cannot rule out the possibility that
respondents and nonrespondents differed on the measures of
clinical interest. Finally, because patients with chronic pain
frequently report difficulties in areas such as concentration,
energy, and sleep, questions have been raised regarding the
probability of false-positives when assessing the prevalence of
MDD within this population (32).
The significance of our findings is underscored by previous
evidence that, compared with either condition singly, comor-
bid chronic pain and depression are associated with a more
malignant course and poorer response to treatment. The pres-
ence of depression among those with chronic pain is associ-
ated with longer pain duration (33) and nonrecovery (7).
Symptoms of depression are associated with poor treatment
response in patients with pain (34). Also, the presence of
comorbid pain in depressed patients treated in primary care
was associated with poor response to standard antidepressant
medications (35), although Lin and colleagues (36) found
improvements in pain severity and functioning among older
depressed primary care patients with arthritis receiving treat-
ment for depression that included medication, psychotherapy,
in-person and telephone follow up. Attention to both chronic
pain and depression among those presenting either condition
singly is likely to be necessary to achieve better outcomes.
1. Ustun TB, Sartorius N, eds. Mental Illness in General Health Care: An
International Study. New York, NY: John Wiley & Sons; 1995.
2. Kessler RC, Nelson CB, McGonagle KA, Liu J, Swartz M, Blazer DG.
Comorbidity of DSM-III-R major depressive disorder in the general
population: results from the US National Comorbidity Survey. Br J
3. Ohayon MM, Schatzberg AF. Using chronic pain to predict depressive
morbidity in the general population. Arch Gen Psychiatry 2003;60:
4. Magni G, Moreschi C, Rigatti-Luchini S, Merskey H. Prospective study
on the relationship between depressive symptoms and chronic musculo-
skeletal pain. Pain 1994;56:289–97.
5. Banks SM, Kerns RD. Explaining high rates of depression in chronic
pain: a diathesis–stress framework. Psychol Bull 1996;119:95–110.
6. Von Korff M, Simon G. The relationship between pain and depression.
Br J Psychiatry 1996;168:101–8.
7. Bair MJ, Robinson RL, Katon W, Kroenke K. Depression and pain
comorbidity: a literature review. Arch Intern Med 2003;163:2433–45.
8. Regier DA, Farmer ME, Rae DS, Myers JK, Kramer M, Robins LN,
George LK, Karno M, Locke BZ. One month prevalence of mental
disorders in the United States and socioeconomic characteristics: the
Epidemiologic Catchment Area Project. Acta Psychiatr Scand 1993;88:
9. Kirmayer LJ, Robbins JM, Dworkin M, Yaffe K. Somatization and the
recognition of depression and anxiety in primary care. Am J Psychiatry
10. Kroenke K, Spitzer RL, Williams JBW, Linzer M, Hahn SR, deGruy FV.
Physical symptoms in primary care: predictors of psychiatric disorders
and functional impairment. Arch Family Med 1994;3:774–9.
11. Sullivan MD, Turner JA, Romano J. Chronic pain in primary care:
identification and management of psychosocial factors. J Fam Pract
12. Wells KB, Stewart A, Hayes RD, Burnam MA, Rogers W, Daniels M,
Berry S, Greenfield S, Ware J. The functioning and well-being of de-
pressed patients. Results from the Medical Outcomes Study. JAMA
13. Becker N, Thomsen AB, Olsen AK, Sjogren P, Bech P, Eriksen J. Pain
epidemiology and health related quality of life in chronic non-malignant
pain patients referred to a Danish multidisciplinary pain center. Pain
14. Kraemer HC. Statistical issues in assessing comorbidity. Stat Med 1995;
15. Von Korff M, Ormel J, Keefe FJ, Dworkin SF. Grading the severity of
chronic pain. Pain 1992;50:133–49.
16. Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-
report version of PRIME-MD: the PHQ primary care study. Primary Care
Evaluation of Mental Disorders. Patient Health Questionnaire. JAMA
17. Spitzer RL, Williams JBW, Kroenke K, Hornyak R, McMurray J, Heart-
well S. Validity and utility of the PRIME-MD Patient Health Question-
naire in assessment of 3000 obstetrics–gynecologic patients: the
PRIME–MD Patient Health Questionnaire Obstetrics-Gynecology Study.
Am J Obstet Gynecol 2000;183:759–69.
18. Kroenke K, Spitzer R. The PHQ-9: a new depression diagnostic and
severity measure. Psychiatric Annals 2002;32:508–15.
19. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief
depression severity measure. J Gen Intern Med 2001;16:606–13.
20. Kroenke K, Spitzer RL, Williams JB. The PHQ-15: validity of a new
measure for evaluating the severity of somatic symptoms. Psychosom
21. Kroenke K, Spitzer RL, deGruy FV III, Swindle R. A symptom checklist
to screen for somatoform disorders in primary care. Psychosomatics
22. The International Association for the Study of Pain: classification of
chronic pain. Pain 1986;3:S1–S226.
23. Elliott AM, Smith BH, Penny KI, Smith WC, Chambers WA. The
epidemiology of chronic pain in the community. Lancet 1999;354:
COMORBID DEPRESSION IN PRIMARY CARE
267 Psychosomatic Medicine 68:262–268 (2006)
24. Ware JE, Kosinski M, Dewey JE, Gandek B. A Manual for Users of the Download full-text
SF-8 Health Survey. Lincoln, RI: QualityMetric Inc; 2001.
25. Olfson M, Fireman B, Weissman MM, Leon AC, Sheehan DV, Kathol
RG, Hoven C, Farber L. Mental disorders and disability among patients
in primary care group practice. Am J Psychiatry 1997;154:1734–40.
26. Spitzer RL, Williams JBW, Gibbon M, First MB. The Structured Clinical
Interview for DSM-III-R (SCID), I: history, rationale, and description.
Arch Gen Psychiatry 1992;49:624–9.
27. Katon W, Schulberg H. Epidemiology of depression in primary care. Gen
Hosp Psychiatry 1992;14:237–41.
28. Kirmayer LJ, Robbins JM, Dworkind M, Yaffee MJ. Somatization and
the recognition of depression and anxiety in primary care. Am J Psychi-
29. Roy-Byrne PP, Stang P, Wittchen HU, Ustun B, Walters EE, Kessler RC.
Lifetime panic–depression comorbidity in the National Comorbidity Sur-
vey. Br J Psychiatry 2000;176:229–35.
30. Barlow DH. Anxiety and Its Disorders, 5th ed. New York: Guilford Press;
31. Severeijn R, Vlaeyen JWS, van den Hout MA, Weber WEJ. Pain cata-
strophizing predicts pain intensity, disability, and psychological distress
independent of the level of physical impairment. Clin J Pain 2001;17:
32. Fishbain DA, Cutler R, Rosomoff HL, Rosomoff RS. Chronic pain
associated depression: antecedent or consequence of chronic pain? A
review. Clin J Pain 1997;13:116–37.
33. Burton AK, Tillotson KM, Main CJ, Hollis S. Psychosocial predictors of
outcome in acute and subchronic low back trouble. Spine 1995;20:722–8.
34. Blanchard EB, Andrasik F, Neff DF, Arena JG, Ahles TA, Jurish SE,
Pallmeyer TP, Saunders NL, Teders SJ, Barron KD, Rodichok LD.
Biofeedback and relaxation training with three kinds of headache: treat-
ment effects and their prediction. J Consult Clin Psychol 1982;50:
35. Bair MJ, Robinson RL, Eckert GJ, Stang PE, Croghan TW, Kroenke K.
Impact of pain on depression treatment response in primary care. Psy-
chosom Med 2004;66:17–22.
36. Lin EH, Katon W, Von Korff M, Tang L, Williams JW Jr, Kroenke K,
Hunkeler E, Harpole L, Hegel M, Arean P, Hoffing M, Della Penna R,
Langston C, Unutzer J; IMPACT Investigators. Effect of improving
depression care on pain and functional outcomes among older adults with
arthritis: a randomized controlled trial. JAMA 2003;290:2428–9.
B. A. ARNOW et al.
268 Psychosomatic Medicine 68:262–268 (2006)