Gebhardt et al.: Pain in depressive disorders

Supplemental On-line Content

references / study design / treatment/ dosage/ number of patients / duration / mean age/ gender/
further characte-ristics / rating scales D/P=
depressive/
pain symptomatology / results
D/P = depressive/pain symptomatology
R = data on relations between depressive and pain symptomatology
SSNRI:
Detke et al. 2002[39] / randomized, multicenter, double-blind, placebo-controlled / duloxetine (60 mg/d, n=123),
placebo (n=122) / 9 weeks / 42 yrs (67% females) / D: HAMD
P: VAS / D: Duloxetine superior to placebo (mean HAMD change -10.9 vs. 6.1; p<0.001).
P: Duloxetine superior to placebo (mean VAS change -8.5 vs. -1.3; p=0.02).

Detke et al. 2002a[40]

/

randomized, multicenter, double-blind, placebo-controlled

/

duloxetine (60 mg/d, n=128), placebo (n=139)

/

9 weeks

/

41 yrs (69% females)

/

D: HAMD

P: VAS

/

D: Duloxetine superior to placebo (p=0.024) (estimated probabilities of response/remission 65/43% vs. 42/28%; HAMD change date not given).

P: Duloxetine superior to placebo (mean VAS change -11.0 vs. -6.4; p=0.037).

Detke et al. 2004[41] / randomized, multicenter, double-blind, placebo- and active-comparator-controlled, outpatients / duloxetine (80 mg/d, n=95), duloxetine (120 mg/d, n=93), paroxetine (20 mg/d, n=86), placebo (n=93) / 8 weeks; continuation phase of 6-months in responders / 45 yrs (73% females) / D: HAMD, MADRS
P: VAS, SSI Pain Item Average / D: During acute phase duloxetine 80/120 mg/d and paroxetine superior to placebo (week 8: mean HAMD change -11.0/-12.1/-11.7 vs. -8.8; p=0.001/p<0.001/ p<0.001). Comparable results for MADRS. Continuation phase: data on HAMD change not given. During continuation phase, duloxetine 80/120 mg/d and paroxetine treatment groups showed significant within-group improvement in HAMD total score and MADRS (no HAMD/MADRS change data and the according significance data given).
P: During acute phase duloxetine 80/120 mg/d and paroxetine superior to placebo (p≤0.05); no difference between duloxetine and paroxetine. During continuation phase duloxetine 120 mg/d improvement on the SSI Pain Item Average (within-group; p=0.034). No significant within-group VAS change from baseline to endpoint.
Goldstein et al. 2002[42] / randomized, multicenter, double-blind, placebo- and active-comparator-controlled, outpatients / duloxetine (40 mg/d, n=86), duloxetine (80 mg/d, n=91), paroxetine (20 mg/d, n=87), placebo (n=89) / 8 weeks / 41 yrs (61.5% females) / D: HAMD
P: VAS / D: Duloxetine 40/80 mg/d superior to placebo (HAMD change by 2.43/3.62 points; p=0.034/0.002), while paroxetine was not (1.51 points; p=0.150); duloxetine 80 mg/d was superior to paroxetine (2.39 points; p=0.037).
P: Duloxetine 80 mg/d was superior to placebo (mean VAS change -10.3 vs. -3.2 mm; p=0.048); paroxetine 20 mg/d showed a trend (-8.1; p=0.07), duloxetine 40 mg/d showed no differences to other treatments.
Brannan et al. 2005[43] / randomized, multicenter, double-blind, placebo-controlled / duloxetine (60 mg/d, n=141), placebo (n=141) / 7 weeks / 41 yrs (65% females) / D: HAMD, BPI
P: VAS, BPI, SQSS / D: Duloxetine not superior to placebo.
P: Duloxetine not superior to placebo in mean baseline-to-endpoint VAS changes; trend towards a lower BPI average pain score under duloxetine compared to placebo (mean BPI change -2.3 vs. -1.8; p=0.066).
Kroenke et al. 2006 [44] / randomized, multicenter, double-blind, placebo-controlled / venlafaxine (mean dosage 177 mg/d, n=55), placebo (n=57) / 12 weeks / 47 yrs, (80% females), multi-somatoform disorder, inclusion if HAMD-17 ≥ 14 / D: HAMD
P: PHQ-15 pain subscale, MOS SF-36 bodily pain / D: Venlafaxine superior to placebo as a non-significant trend (HAMD change -13.2 vs. -9.6; p=0.070).
P: Venlafaxine superior to placebo (PHQ-15 change -3.3 vs. -2.6; p=0.030; MOS SF-36 bodily pain change +25.1 vs. +14.5; p=0.030).
Perahia et al. 2006[45] / randomized, multicenter, double-blind, placebo- and active-comparator-controlled / acute/continuation phase: duloxetine (80 mg/d, n=93/58), duloxetine (120 mg/d, n=103/62), paroxetine (20 mg/d, n=97/61), placebo (n=99/62) / acute phase: 8 weeks; continuation phase: 6 months / 45 yrs (66% females); outpatients / D: HAMD
P: VAS / Acute phase:
D: Duloxetine 80/120 mg/d superior to placebo (HAMD-change -12.1/-12.4 vs. -10.8; p=0.045/p=0.014); paroxetine not superior to placebo (HAMD-change -11.9; p=0.089).
P: Duloxetine 80 mg/d superior to placebo (VAS-change not shown; p=0.044).
Continuation phase:
D: No significant difference in time to loss of response between any of the treatment groups.
P: No improvement from week 8 to endpoint for any treatment except for paroxetine on pain while awake (p≤0.05). Duloxetine 80 mg/d showed significant worsening (p=0.045) on severity of overall pain.
Brecht et al. 2007[46] / randomized, multicenter, double-blind, placebo-controlled / duloxetine (60 mg/d, n=162), placebo (n=165) / 8 weeks / 50 yrs (74% females) / D: MADRS
P: BPI-SF / D: Duloxetine superior to placebo (mean changes in MADRS total scores -16.69 vs. -11.31, p<0.0001).
P: Duloxetine superior to placebo (mean changes in BPI-SF average pain scores -2.57 vs. -1.64, p<0.001).
Wohlreich et al. 2009[47] / randomized, multicenter, double-blind, placebo-controlled / duloxetine (60 mg/d, n=207), placebo (n=104) / 8 weeks / 73 yrs (59% females / D: HAMD
P: VAS / D: Duloxetine superior to placebo (mean changes in HAMD total scores, arthritis: -6.38 vs. -3.47, p≤0.05; no arthritis: -7.97 vs. -4.45, p≤0.005)
P: Duloxetine superior to placebo (mean changes in VAS overall pain, arthritis: -4.19 vs. -6.61, , p≤0.05; no arthritis: -5.39 vs. 8.11; n.s.).
Gaynor et al. 2011[48] / randomized, multicenter, double-blind, placebo-controlled, outpatients / duloxetine (60 mg/d, n=192), placebo (n=204) / 8 weeks / 46 yrs (69% females) / D: MADRS
P: BPI / D: Duloxetine superior to placebo (mean changes in MADRS total scores -16.77 vs. -12.73, p<0.001).
P: Duloxetine superior to placebo (mean changes in BPI average pain scores -1.93 vs. -1.31, p≤0.001).
R: Patients with BPI response had higher depression remission rates (p<0.001).
Gaynor et al. 2011a[49] / randomized, multicenter, double-blind, placebo-controlled, outpatients / duloxetine (60 mg/d, n=209), placebo (n=217) / 8 weeks / 45 yrs (69% females) / D: MADRS
P: BPI / D: Duloxetine superior to placebo (mean changes in MADRS total scores -14.96 vs. -10.77, p<0.001).
P: Duloxetine superior to placebo (mean changes in BPI average pain scores -1.66 vs. -1.71, p≤0.001).
R: Patients with BPI response had higher depression remission rates (p<0.001).
Robinson et al. 2013[50] / randomized, parallel-group, multicenter, double-blind, placebo-controlled / duloxetine 60 mg/d (n=523), placebo (n=532) / two 8-week trials / 46/45 yrs (69/68% females) / D: MADRS
P: BPI / D: Duloxetine resulted in a greater improvement in depression than placebo (-15.8 vs -11.7; p<0.001).
P: Duloxetine resulted in a greater improvement in pain than placebo (p<0.001).
R: Duloxetine-treated patients with reduction of BPI score had higher rates of remission. Path analysis indicated that 16% of likelihood of remission in depression symptoms was because of the direct effect of treatment, 41% because of pain reduction, and 43% because of functional improvement.
Clayton et al. 2013[51] / randomized, parallel-group, multicenter, double-blind, placebo-controlled / desvenlafaxine 50 mg/d (n = 217), placebo (n = 217) / 8 weeks / 53/53 yrs (100% females) / D: , HAMD MADRS
P: VAS / D: Desvenlafaxineresulted in a greater improvement in depression than placebo (HAMD: -9.9 vs -8.1; p=0.004).
P: Desvenlafaxineresulted in a greater improvement in pain than placebo (-1.5 vs. -0.5; p<0.001).
Robinson et al. 2014[52] / randomized, multicenter, double-blind, placebo-controlled / duloxetine 60 or 120 mg/d (n=204), placebo (n=95) / 24 weeks / 73/73 yrs (66/59% females) / D: HAMD, MADRS
P: BPI / D: Duloxetine showed no greater improvement of depression at week 12, but at weeks 4,8,16, and 20 (p<0.05).
P: Duloxetine showed greater improvement of pain at week 12 (p=0.009).
SSRI:
Heiligenstein et al. 1995 [53] / randomized, multicenter, double-blind, placebo-controlled / fluoxetine (20 mg/d, n=261), placebo (n=271) / 6 weeks / 68 yrs (54% females) / D: HAMD
P: MOS SF-36 / D: Data not given.
P: Fluoxetine superior to placebo (mean change 7.1 vs. 2.9; p<0.05).
R: Changes in SF-36 scores correlated with changes in HAMD scores.
Dickens et al. 2000[54] / randomized, placebo-controlled / paroxetine (20 mg/d, n=44),
placebo (n=48) / 8 weeks / 45 yrs (54% females) / D: MADRS
P: VAS, MPQ, Oswestry Disability Questionnaire / D: Paroxetine not superior to placebo.
P: Paroxetine not superior to placebo.
Tsui et al. 2011 [55] / randomized, double-blind, placebo-controlled / escitalopram (10 mg/d; n=75), placebo (n=75) / 12 weeks / 38 yrs (76% men) / D: BDI
P: VAS, BPI / D: No significant difference between escitalopram and placebo.
P: Escitalopram superior to placebo (VAS/BPI-change -36.4/-3.2 vs. -12.5/-1.1; p<0.01/p<0.05) from baseline to follow-up.
Tricyclic AD:
Ward et al. 1979extra [56]
Ward et al. 1982extra [57] / prospective, open label / doxepin (196 mg/d; n=16), nontreatment (n=25) / 4 weeks / age and gender not described / D: HAMD
P: MPQ / D: No data given.
P: In 87% some pain relief, in 56% complete pain relief.
R: Correlation between HAMD reduction and reduction of pain (p<0.005).
Ward et al.
1984extra [58] / prospective, double blind / total n=26,
doxepin (188 mg/d), desipramine (173 mg/d) / 4 weeks / 43 yrs (of completers 12 males, 14 females) / D: HAMD, BDI
P: MPQ, 10-point-scale / D: Both drugs produced significant decreases in depression (HAMD change -16.7); no significant difference was seen between groups.
P: Pain reduction of -2.8 (10-point-scale); results on MPQ not explicitly displayed; pain severity showed a better response to doxepin than to desipramine.
Ward 1986extra [59] / prospective, double blind / total n=35,
doxepin (188 mg/d), desipramine (173 mg/d) / 4 weeks / 40 yrs (18males) / D: HAMD, POMS
P: MPQ, 10-point-scale, VAS / D: Both drugs produced significant decreases in depression (HAMD change -17.1); no data given on differences between groups.
P: Pain reduction of – 2.6 (10-point-scale); results on MPQ not explicitly displayed; no significant differences in pain response between doxepin and desipramine.
R: Significant association of depression relief and relief von pain severity and pain frequency (p<0.001, p<0.003).

Lindsay Wyckoff 1981extra[60]

/ open label, prospective / n=300; first and second choice: amitriptyline/ imipramine/ desipramine/ doxepin (150-300 mg/d), third choice: monoamine oxidase inhibitor / “several weeks” / approx. 44 yrs (gender not given) / D: criteria of Feighner et al. 1972
P: subjective analog scale / D: Data not given.
P: Approximately 60% had complete pain relief, approximately 40% had partial relief.
Data on methods and results are imprecise.
Hameroff et al. 1984 [61] / randomized, double-blind, placebo-controlled, outpatients / doxepin (mean dose 200 mg/d, n=30), placebo (n=30) / 6 weeks / 49 yrs (53% males) / D: HAMD, POMS
P: VAS / D: Doxepin superior to placebo (p<0.05).
P: Doxepin superior to placebo (p<0.05).
Jaracz et al. 2015extra [62] / open-label, part-randomized / nortriptyline 50-150 mg/d (n=30), escitalopram 10-30 mg/d (n=30) / 8 weeks / 39/41 yrs (73% males) / D: MADRS, HAMD, BDI
P: VAS / D: Significant depression symptom reduction (50% reduction) between 4 and 6 weeks and throughout the study.
P: No difference between nortriptyline and escitalopram in pain reduction. Significant pain reduction (50% reduction) after 2 weeks of treatment and throughout the study.
R: A 50% reduction in pain intensity preceded the 50% reduction of depression severity.
Tetracyclic AD:
Onghena et al. 1993 [63] / cross-over, double-blind, placebo-controlled / mianserin (90mg/d, n=20), placebo / 12 weeks / 45 yrs (59% females) / D: HAMD, BDI
P: MPQ, VAS / D: Significant reduction (p<0.05), depression scores were lower under mianserin compared to placebo (p<0.05; in 8 depressive patients without pain).
P: No significant reduction in any group.
Freynhagen et al. 2006extra [64] / prospective, non-randomized, uncontrolled, outpatients / mirtazapine (mean dose 34.5 mg/d, n=594) / 6 weeks / 55 yrs (66% females) / D: a four-point scale
P: a five-point scale (self-assess-ment) and a four-point scale (phy-sicians) / D: Improvement of depression, not statistically tested.
P: Reduction of pain (mean change 2.1; p<0.0001).

1 / 13

Gebhardt et al.: Pain in depressive disorders

Table legend

Table S1: The pharmacotherapy studies on patients with depression and pain symptomatology included in the present meta-analysis and the “extra-meta-analysis” (see Discussion Section)

AD = antidepressant; BDI = Beck Depression Inventar; BPI (-SF)= Brief Pain Inventory (Short Form); HAMD=Hamilton Rating Scale for Depression; MADRS = Montgomery-Asberg Depression Rating Scale; MPQ = McGill Pain Questionnaire; MOS SF-36 = Medical Outcomes Study Short Form; PHQ-15 = Patient Health Questionnaire;POMS = Profile of Mood States; SQSS = Symptom Questionnaire, Somatic Subscale; SSI = Somatic Symptom Inventory; SSNRI = selective serotonin-noradrenaline reuptake inhibitors; SSRI = selective serotonin reuptake inhibitors; VAS = Visual Analog Scales for pain.

extra = non-placebo-controlled TCA studies included in the “extra-meta-analysis” (see Discussion Section)

1 / 13