Using chronic pain to predict depressive morbidity in the general population
Ohayon MM and Schatzberg AF;
Commented by , 24 Feb 2003
Background
While the association between pain and depression is known, few studies have controlled for the presence of comorbid nonpainful medical conditions.
Purpose
In a large community sample, examine:
1) The prevalence of chronic painful physical conditions (CPPCs, > 6 months) (joint, limb, or back pain, headache, or gastrointestinal diseases), depressive symptoms and major depressive disorder (MDD).
2) The association between CPPCs, depressive symptoms and MDD; and whether comorbid nonpainful medical conditions modify the association.
Methods
A random sample of 18980 subjects (80.4% participation rate) from 15-100 years old, representative of the general population of 5 European countries participated in a cross sectional telephone survey conducted by lay interviewers. Answers from the Sleep-EVAL system, validated in various psychiatric settings, were the main outcome measures.
The questionnaire obtains sociodemographic, physical health, sleep and psychiatric symptom (as per DSM-IV) data. All rates were given 95% confidence intervals or standard errors. Chi-square and logistic regression were used for bivariate and odds ratio analyses, respectively.
Results
17.1% of subjects reported having at least 1 CPPC, increasing linearly with age. 16.5% reported at least 1 of 3 key depressive symptoms (sadness, anhedonia, hopelessness); 27.6% of these subjects had at least 1 CPPC. These subjects (with a comorbid CPPC) reported a longer duration of depressed mood (19 months) than those without a CPPC (13.3 months).
MDD was diagnosed in 4.0%; 43.4% of these subjects had at least 1 CPPC, 4 times more often than in subjects without MDD. Conversely, the prevalence of MDD was 10.2% in those with a CPPC compared with 2.7% in those without. Most subjects with MDD (61.6%) reported having either a CPPC or a nonpainful medical condition.
In a logistic regression model, CPPC was strongly associated with MDD (OR: CPPC alone, 3.6; CPPC + nonpainful medical condition, 5.2; nonpainful medical condition alone, 2.2). Additional sociodemographic factors associated with MDD included
- female gender (OR, 1.4)
- ages 45-54 (OR, 1.5)
- night shift work (OR, 1.5)
- unemployment (OR, 2.2)
- being a homemaker (OR, 1.6)
- having moderate (OR, 1.7) and high (OR, 3.0) stress levels
- smoking < 20 cigarettes/day (OR, 1.3)
Discussion
The authors provide a noteworthy contribution to our understanding of the pain-depression interface, particularly in its prominence as a major public health issue. Community subjects with MDD have a 4-fold prevalence of a chronic painful condition, with a similar reciprocal relationship.
Of all sociodemographic and clinical factors, a CPPC made the strongest independent contribution to MDD. This strong association has also been shown by Stewart et al (1), who suggest pain may be the best indicator of depression, particularly in the elderly. A CPPC also prolonged the course of depressive symptoms.
This treatment-resilient nature of comorbid pain and depression has been noted by others (2,3), who submit treating only one or the other may underlie such treatment-resistance.
As the authors concede, the pain-depression interaction is elusive; the cross-sectional design limits causal inferences as to what extent pain causes depression or vice versa.
This former dualistic approach is in fact too simplistic in the face of evidence supporting a complex reciprocal relationship between neurobiological, psychological and behavioral models (4).
While each model has empiric support, none appear to be adequate in isolation, thus informing integrated treatment approaches (2,3) which readers are encouraged to review.
References
1. Stewart RB, Blashfield R, Hale WE, et al. Correlates of Beck Depression Inventory scores in an ambulatory elderly population: symptoms, disease, laboratory values and medications. Journal of Family Practice 1991; 32; 497-502
2. Turk DC. Combining somatic and psychosocial treatment for chronic pain patients: perhaps 1+1 does = 3. The Clinical Journal of Pain 2001; 17; 281-283
3. Gallagher RM, Verma S. Managing pain and comorbid depression: A public health challenge. Seminars in Clinical Neuropsychiatry 1999; 4; 203-220
4. Von Korff M, Simon G. The relationship between pain and depression. British Journal of Psychiatry 1996; S101-S108