Pharmacoepidemiology of comorbid chronic pain and depression: Insights from a case-control study

Background: The bidirectional relationship between depression and chronic pain is well recognized but their clinical management remains challenging. Here we characterize the shared aetiology and risk factors for their comorbidity using large population cohorts to advance understanding of pharmacological treatment outcomes. Methods: Participants completed online questionnaires about chronic pain, psychiatric symptoms, comorbidities, treatment response and general health (N=13,839). Logistic regression models were used to examine the relationship between chronic pain and clinical and demographic factors. Cumulative linked logistic regressions assessed the effect of chronic pain on treatment response for 10 different antidepressants. Findings: Chronic pain was associated with an increased risk of depression (OR=1.86 [1.37-2.54]), recent suicide attempt (OR=1.88[1.14-3.09]), higher use of alcohol, tobacco and painkiller misuse. Participants with comorbid chronic pain and depression reported fewer functional benefits from antidepressant use and lower benefits from sertraline (OR=0.75[0.68-0.83]), escitalopram (OR=0.75[0.67-0.85]) and venlafaxine (OR=0.78[0.68-0.88]) when compared to participants without chronic pain. Furthermore, participants taking sertraline (OR=0.45[0.30-0.67]), escitalopram (OR=0.45[0.27-0.74]) and citalopram (OR=0.32[0.15-0.67]) specifically for chronic pain reported lower benefits compared to other participants taking these same medications but not for chronic pain. Interpretation: The findings reveal novel insights into the complex relationship between chronic pain and depression. Treatment response analyses indicate differential effectiveness between particular antidepressants and poorer functional outcomes for these comorbid conditions. Further assessment is warranted in targeted interventional trials. This approach will advance precision psychiatry and assist in clinical management by choosing the most suitable treatment for patients, informed by specific symptoms and comorbidities.


INTRODUCTION
Depression is estimated to affect over 264 million people worldwide and is a leading cause of global disability. 1 Its clinical manifestations and outcomes are highly heterogeneous, with multiple factors underlying susceptibility, progression and treatment response. 2 One important factor that frequently complicates the diagnosis of depression is comorbid chronic pain, 3 as patients presenting with pain are more likely to be investigated medically rather than as part of a broader biopsychosocial framework. 4 Depression and chronic pain frequently coexist, with up to 61% of chronic pain patients also presenting with depression. 5 Importantly, the combination of chronic pain and depression is more disabling and more costly to treat than either condition in isolation. 3 Patients with comorbid chronic pain and depression are also at increased risk of substance use disorder. 6 These three conditions synergistically elevate the risk of premature death, cardiovascular disease and cancer. 6 Chronic pain has been defined by the International Association for the Study of Pain (IASP) as pain persisting or recurring for longer than three months. 7 In contrast to acute pain, which alerts individuals to potential or real tissue damage, chronic pain serves no apparent physiological purpose and persists beyond normal healing time. 7 In 2016, 1.6 million Australians over the age of 45 had chronic pain, 8 while globally, lower back and neck pain were identified as the leading cause of disability. 1 Furthermore, estimates suggest chronic pain cost the Australian economy $139 billion in 2018, mostly due to lost productivity and impaired quality of life. 8 Predictions suggest this cost will almost triple by 2050. 9 In the United States, the cost is already over $500 billion per year. 10,11 The relationship between chronic pain and depression is bidirectional, as having either chronic pain or depression increases the risk of developing the other condition. 12 Furthermore, the relationship is thought to be dose-dependent, as more severe pain is associated with greater severity of depression. 13 That is especially true for elderly populations, 12 who also report the highest prevalence (13%) 12 of comorbid chronic pain and depression out of all age groups. 3 Although the precise causal mechanisms between chronic pain and depression remain elusive, recent large-scale human genetic 14 and animal model studies 15,16 have made important inroads. For example, an imbalance of neurotransmitters and hormones has been observed in both depression and chronic pain. 17 Consequently, antidepressants, which increase monoamine levels in the central nervous system, are used in the pharmacological treatment of both conditions. 18 However, antidepressants are generally prescribed in lower doses for the treatment of chronic pain, compared to the higher doses required for antidepressant effects. 19 Serotonin noradrenaline reuptake inhibitors (SNRIs; e.g., duloxetine) and selective serotonin reuptake inhibitors (SSRIs; e.g., paroxetine, sertraline) are commonly used antidepressants for the treatment of comorbid chronic pain and depression. 18 Other antidepressant options include tricyclic antidepressants (TCAs) such as amitriptyline. 18 These medications have been found to partially reduce the symptoms of both depression and pain. 18 No significant differences in efficacy between these medications have been identified so far, 20 however further research is required. 18 For example, it has not been possible to compare the efficacy of TCAs against other antidepressants for the treatment of comorbid chronic pain and depression, due to a lack of rigorous studies. 20 Despite the high prevalence and cost of comorbid chronic pain and depression 3 , research efforts have yet to deliver clinically useful findings and recommendations. 6,21 For example, a recent review highlighted that it was unclear which specific antidepressant should be prescribed as the first-line pharmacological treatment for comorbid chronic pain and depression. 18 Pharmacoepidemiological studies -which examine the use and effect of medications in large population cohorts -have been proposed as a cost-effective method for reviewing pharmaceutical safety and effectiveness, as well as helping to inform clinical guideline development. 22 In the current study, we examined the pharmacoepidemiology of comorbid chronic pain and depression in the Australian Genetics of Depression Study (AGDS) -one of the world's largest participant cohorts with a detailed history of depression and its comorbidities. 23 Here we sought to: (i) quantify the association between depression and chronic pain; (ii) assess the dependency between chronic pain severity, depression severity and recent suicidality; (iii) identify other psychiatric disorders and patterns of recent substance use that are associated with comorbid chronic pain and depression; and (iv) assess the impact of comorbid chronic pain and depression on antidepressant effectiveness. participants were recruited, included both conventional and online social media. The campaign appealed for anyone who "had been treated by a doctor, psychiatrist or psychologist for depression" to visit our website (https://www.geneticsofdepression.org.au). For the targeted mailout, invitation letters were sent by the Australian Government Department of Human Services (DHS) to individuals who, according to their records, had received at least four prescriptions for any of the ten most commonly used antidepressants in the last 4.5 years. DHS did not, at any time, share any personal information with the research team. Potential participants were directed to our website, which contained information about the study, a registration and consent form, and a comprehensive online questionnaire. The essential inclusion criteria included having been prescribed and taken antidepressants, and providing consent to donate a saliva sample for subsequent genotyping. No participant was excluded based on comorbid conditions. The online survey assessed mental health diagnoses, antidepressant response, suicidality, general health and substance use, among several other variables. A baseline description of the cohort has been published elsewhere. 23 PISA cohort. The Prospective Imaging Study of Ageing (PISA) is a longitudinal cohort of Australian adults. 24 The population-based sample recruitment pool comprised adult twins, their spouses, and first-degree relatives of twins and spouses who over previous decades, had volunteered for studies on risk factors or biomarkers for physical or psychiatric conditions and had previously been genome-wide genotyped. 25,26 The PISA protocol consisted of online questionnaires, including history of mental health diagnoses and the same pain questionnaire in AGDS. It was completed by N=2,469 PISA participants. For that reason, AGDS and PISA data were used in the present study to assess the effect of depression and demographics (e.g., age, sex) on chronic pain. All other analyses described in this manuscript were performed only in the AGDS cohort.
Depression, chronic pain and ascertainment. AGDS participants were asked to self-report whether they had ever been diagnosed with depression by a health professional, and similarly for 19 other psychiatric conditions. Individuals were classified as depression cases if they had reported both a depression diagnosis and had been prescribed antidepressants in the past five years (N=17,849). Participants were also asked to indicate whether they experienced chronic pain in their daily life and to rank its intensity on a scale from 0 -10. Only those reporting pain (rating >0) progressed to the remainder of the pain module, which included questions about the duration and location of their primary pain. Following the IASP guidelines, chronic pain was defined as pain persisting or recurring for at least three months. 18 Cases were classified as having comorbid chronic pain and depression if they fulfilled the criteria for both conditions (N=6,895), and controls were classified as those who reported depression but no chronic pain (N=4,475). We performed a complete case analysis. Thus, participants with missing data for chronic pain (i.e., did not complete the section; N=6,463) were excluded from analyses that needed data for both chronic pain and depression.

4
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 24, 2020. ; Recent suicidality and substance use. Suicidal ideation over the last month was measured on a 10-point scale: 0 indicated having no suicidal ideation in the past month (never) and 10 denoted persistent suicidal ideation. Participants with score>0 were classified as positive cases for suicidal ideation. Suicide attempt was measured using a similar 10-point scale in regard to how close a participant had come to making an attempt. Only those with a score of 10 (labelled as "I have made an attempt") were considered cases for suicide attempt. Participants also reported their frequency in using a range of substances over the last three months. Alcohol consumption frequency was measured as the number of days the participant drank three or more standard drinks. For all other substances, the response options were: "never" (0), "once or twice" (1), "monthly" (2), "weekly" (3) or "daily" (4). These responses were modelled as continuous variables when assessing their correlation with chronic pain.
Antidepressant use and response. Participants were asked whether they had ever been prescribed any of the ten most commonly used antidepressants in Australia for any indication. These are sertraline, escitalopram, venlafaxine, amitriptyline, mirtazapine, desvenlafaxine, citalopram, fluoxetine, duloxetine and paroxetine. Information regarding the reason(s) for prescription was collected using a checklist of 17 possible responses including depression, chronic pain, and anxiety (among others). Multiple selections were possible. Participants were asked to report on the best aspects of taking antidepressants using the following item: "What were the best aspects of taking the antidepressant(s)? Include any antidepressant you have taken." Participants were then able to select all that apply out of a list including: relief of depressive symptoms, relief of other symptoms e.g. sleep disturbance, reduction in suicidal ideation, return of normal emotion, improved relationships, returning to normal activities and restored control over mood . Moreover, participants rated the effectiveness of each antidepressant they had taken, using a scale ranging from 0 (e.g., "sertraline works not at all well for me") to 2 (e.g., "sertraline works very well for me"). Two analyses were performed: (i) first, antidepressant effectiveness was compared between participants who reported taking an antidepressant prescribed for chronic pain against the rest of the participants (i.e., not prescribed for chronic pain); and (ii) we compared antidepressant effectiveness between participants reporting chronic pain and those reporting no chronic pain (regardless of explicit indication).

Statistical analyses.
The relationship between chronic pain and several other variables of interest was assessed using multivariable logistic regression. This approach enabled us to quantify the associations while adjusting for age, sex and all other relevant factors (e.g., the correlation between alcohol and chronic pain while keeping usage of all other substances equal ). Fully adjusted odds ratios were calculated from effect sizes on the logit scale, and p-values were estimated using Wald-tests. For all analyses, the presence of chronic pain was modelled as a binary variable, while chronic pain severity was modelled as a quantitative score from zero to 10. The relationship between chronic pain and antidepressant response was examined using cumulative link logistic regressions to accurately model treatment response, which was coded on an ordinal scale. Furthermore, to assess the effect of chronic pain across all antidepressants, a random effect was 5 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020.
included to account for repeated responses from participants. This analysis was performed in R using the ordinal package and the clm and clmm functions, adjusting for the effects of sex and age when antidepressant treatment started. All other statistical analyses were performed and figures generated in python using these modules: statsmodels , scipy , numpy , pandas , matplotlib and seaborn .

Sample demographics and association between chronic pain and depression
Demographics and chronic pain prevalence for both AGDS (enriched for depression) and PISA (not enriched for depression) cohorts are shown in Table 1 and Supplementary  Figures 1 and 2 . Figure 1 shows the prevalence of chronic pain by age, stratified by cohort. A significant cohort effect is evident. Despite the PISA cohort being older on average ( Supplementary Figure 1 ), the AGDS cohort showed a higher prevalence of chronic pain. This cohort effect is partly attributable to depression status (Depression OR=

Chronic pain is associated with severity of depression and recent suicidality
Results presented here are from AGDS where all participants reported depression. Higher pain severity (intensity) was found to be associated with longer durations of pain ( Supplementary Figure 3 ). Increased pain severity was also associated with an increased number of depressive episodes ( Supplementary Figure 4 ). The prevalence of suicidal ideation was higher in the comorbid chronic pain group (OR=1.49 [1.38 -1.61]). Likewise, recent suicide attempt was associated with chronic pain (OR=1.88 [1.14 -3.09]). Within the chronic pain group, recent suicidal thoughts and suicide attempt scores were also positively correlated with chronic pain severity scores ( Supplementary Figure 4 ).

Comorbid psychiatric diagnoses and recent substance use
In this subsection, the results presented are from AGDS where all participants reported depression. Out of the nineteen mental health conditions examined, social anxiety disorder was found to have the strongest association with chronic pain (p<0.01). Anorexia nervosa was found to be negatively associated with the likelihood of developing chronic pain (p<0.05). Although both of these results were nominally significant, no association survived correction for multiple testing ( Figure 2 , Supplementary Table 2 ). Notably, chronic pain was significantly associated with increased use of alcohol and tobacco, as well as painkiller misuse (e.g., opioids). Nominal associations were observed for other drugs such as cocaine (negative relationship) and opioids ( Figure 2 , Supplementary  Table 3 ).

Chronic pain and antidepressant response
6 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. ; In this subsection, the results presented are from AGDS where all participants reported depression. The three most commonly prescribed antidepressants for the indication of chronic pain were amitriptyline (N=606), duloxetine (N=288) and sertraline (N=160; Supplementary Figure 5 ). Overall, the large majority of participants with chronic pain did not report an antidepressant prescription for chronic pain ( Supplementary Table 4 ). Participants with chronic pain were less likely to report positive benefits from antidepressants such as relief of depressive symptoms ( Supplementary Figure 6 ). A trend -whereby participants with chronic pain were more likely to report a reduction in suicidal symptoms as a positive aspect of antidepressant treatment was noted -but this result did not survive correction for multiple testing ( Supplementary Table 5 ). Furthermore, in the chronic pain group, the average self-reported benefits from taking antidepressants was significantly lower compared to the group without chronic pain (OR=0.75 [0.71 -0.80]; p<2×10 -16 ). A similar but non-significant finding was observed between the average response of participants prescribed antidepressants for chronic pain versus those without a chronic pain indication (OR=0.94 [0.80 -1.1]; Supplementary  Figure 7 ). Next we examined whether these findings held true for each antidepressant under study. For most antidepressants, no statistically significant difference in effectiveness was found between participants with chronic pain (or an indication for chronic pain) and participants without chronic pain. Participants with chronic pain who had taken sertraline, escitalopram or venlafaxine, reported significantly lower effectiveness than participants without chronic pain ( Figure 3a ). Furthermore, lower effectiveness of sertraline, escitalopram and citalopram was observed depending on whether they were prescribed for chronic pain or a different clinical indication ( Figure 3b ). The only antidepressants with a positive effect (i.e., greater effectiveness) were duloxetine, venlafaxine and amitriptyline, but only when prescribed for comorbid chronic pain and depression. However, none of the positive associations reached statistical significance ( Figure 3b , Supplementary Table 6 , Supplementary Table 7 ).

DISCUSSION
We have reported the largest study on comorbid chronic pain and depression treatment outcomes through comprehensive population-based sampling. A key finding is that in participants with comorbid chronic pain and depression, self-reported benefits from taking particular SSRI & SNRI antidepressants (i.e., sertraline, escitalopram, venlafaxine) were significantly lower compared to participants without chronic pain. Indeed, this result is corroborated by participants with chronic pain reporting fewer functional benefits from taking antidepressants, such as relief of depressive symptoms and getting back to normal daily activities. Furthermore, we found that participants prescribed particular SSRI antidepressants (i.e., sertraline, escitalopram, citalopram) for chronic pain reported significantly lower benefits compared to those taking the same medications but for a different indication. These results suggest that although SSRI and SNRI antidepressants as a class may be equally efficacious in the treatment of comorbid chronic pain and depression, 20,27 specific antidepressants have differential effectiveness. By utilizing large 7 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. ; independent population cohorts, we have shown evidence for differential effectiveness between several specific antidepressants in comorbid chronic pain and depression. We consider these findings to be robust because our methodological approach took into account the inherent clinical heterogeneity, high comorbidity and wide individual variation commonly observed in psychiatric disorders, in contrast to highly selective subject sampling in clinical trial studies.
The current study's main findings of differential antidepressant effectiveness and fewer functional benefits from antidepressant use in comorbid chronic pain and depression is further underlined by demonstrating several results consistent with previous research. These include: (i) a strong association between depression and chronic pain; 3,28,29 (ii) increasing severity of chronic pain was associated with a higher number of depressive episodes experienced by participants 13 and (iii) older age, lower educational attainment and female sex were associated with higher chronic pain prevalence. [30][31][32] While there have been conflicting reports regarding the link between chronic pain severity and suicidal behaviours, 33 , 34-36 we provide evidence that supports an association between comorbid chronic pain and depression with both an increased risk for suicidal ideation and suicide attempt. Considering suicide is a leading cause of death -particularly for young people 37 -and depression and chronic pain are both treatable conditions, assessing their comorbidity in at-risk youth populations may help to reduce suicide rates. 33 Consistent with previous observations, 29 we also found comorbid chronic pain and depression was associated with recent increased use of alcohol, tobacco and painkillers (e.g., opioids). However, we did not observe a significant association between comorbid chronic pain and depression with a self-reported substance use disorder. Previous reports suggest chronic pain, depression and substance use disorder are often comorbid. 38 For example, rates of alcohol use disorder in individuals with comorbid chronic pain and depression are estimated to be between 8.9% to 10%. 6 It is possible that screening and diagnosis of substance use disorders in Australia may be lacking in those with comorbid chronic pain and depression. As such, clinicians need to consider substance use disorder in patients presenting with this comorbidity, as all three conditions increase the risk for other chronic diseases such as cardiovascular disease and cancer, while also increasing the risk of premature death. 6 In the current study, amitriptyline was the most commonly prescribed antidepressant to individuals with comorbid chronic pain and depression. Indeed, it was prescribed over two times more frequently than the next most commonly prescribed medication -duloxetine. This finding differs from previous research and some recommendations, 18 which found that duloxetine was the most commonly used antidepressant in comorbid chronic pain and depression. Amitriptyline is a first line treatment for chronic pain, however, its side-effect profile and mortality risk in overdose often limits its use. 39 It is possible that the severity of illness in the AGDS cohort, which is enriched for severe and recurrent depression, 23 has necessitated the extensive use of amitriptyline. Alternatively, this finding may reflect a lack 8 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. ; of clarity regarding treatment recommendations for comorbid chronic pain and depression, which has been identified as an area requiring further research. 18 Further research can also address certain limitations in the current study. Information on dosages of the prescribed antidepressants, as well as adjunct therapies such as psychotherapy and pain-specific pharmacotherapies, was not gathered. We also did not assess whether participants were taking combinations of antidepressants. These measures can be included in future work. As the data were based on self-reported responses, they may also be subject to a degree of participant recall bias, and the antidepressant response data may include non-specific (placebo) effects.
The current study suggests pharmacoepidemiological approaches in psychiatry will be increasingly valuable as a cost-effective, first-line strategy to enhance the design, feasibility and clinical utility of randomised controlled trials. 40 Identifying the differential effectiveness of specific medications and their functional benefits, along with key contributing comorbidities for assessment in targeted interventional trials, will help to accelerate their clinical translation into improved treatment guidelines and patient outcomes.

AUTHOR CONTRIBUTIONS
WHR, AIC and MER designed this study and wrote the first version of the manuscript. AIC performed the analyses with input from MER and WHR. NGM, SEM, NRW and IBH designed and directed the AGDS data collection efforts. NGM and MKL led the PISA study data collection efforts. TTN designed the pain module in both the AGDS and PISA online surveys and conceived the pharmacogenetic investigation of comorbid pain and depression. LMG-M and GC-P contributed to data analyses. All authors contributed to the interpretation of the results and provided feedback on the preliminary versions of the manuscript.

ROLE OF THE FUNDING SOURCE
The research funders did not participate in the study design, data analyses, results interpretation or writing of the manuscript. 10 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020.  Table 1. 14 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. ;

Figure 2. Comorbidities and substance use associations with chronic pain in AGDS cohort
Forest plots depict the chronic pain odds ratios (ORs) for (a) comorbid disorders and (b) recent substance use during the past three months. Chronic pain was significantly associated with increased use of alcohol, tobacco and painkiller misuse including opioids. Diamonds represent ORs and horizontal lines depict 95% CI. ORs were estimated from a multivariate logistic regression accounting for all relevant covariates (see Methods). *p<0.05; **p<0.05 after Bonferroni correction for multiple testing (AGDS data only).

15
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. ;

Figure 3. Effect of chronic pain on antidepressant effectiveness in AGDS cohort
Forest plots depicting the results from a cumulative link logistic regression predicting antidepressant treatment response using (a) self-reported chronic pain or (b) self-reported prescription for chronic pain as predictors, while adjusting for the effects of sex and age started taking the antidepressant. *Statistically significant after Bonferroni multiple testing correction (p<0.005; AGDS data only). Further details are in Supplementary Figure 4 and Supplementary Tables 4 and 5. 16 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. ; Age mean (sd) 60 (6.8) 60 (6.9) Cases: participants reporting chronic pain Controls: participants reporting no chronic pain *p<0.05 two-sample t-test or χ2 test 17 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Supplementary Figure 1. Age and sex composition of AGDS and PISA cohorts
Histograms with kernel density plots showing the distribution of ages for AGDS and PISA cohorts (left) and their sex-stratified versions (right). Note that the PISA cohort is on average older at the time the participants responded to the questionnaire, but present lower rates of chronic pain on average than the AGDS (depression-enriched) cohort.

Supplementary Figure 2. Significant difference in educational attainment between AGDS and PISA cohorts
Barplots showing the frequency of educational attainment for the AGDS and PISA cohorts. The difference is statistically significant (χ2 test p<E-100).

18
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Supplementary Figure 3. Positive association between pain intensity and pain duration in AGDS cohort
Bar plots illustrate average pain intensity (x-axis) stratified by pain duration, with a strong positive relationship found between them (p=5.5e-106 linear regression; AGDS data only).

19
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Supplementary Figure 4. Positive association between chronic pain severity (intensity), depression (number of episodes) and suicidality in AGDS cohort
Point plots comparing chronic pain severity with: (a) average age of onset, (b) average number of depressive episodes, (c) average suicide thoughts score or (e) suicide attempt score (e). Differences in the prevalence of (d) recent suicide thoughts and (f) recent suicide attempt are also shown. Error bars denote 95% confidence intervals (AGDS data only).

20
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Supplementary Figure 5. Antidepressant intake by chronic pain participants
Barplots showing the percentage (number) of participants with comorbid chronic pain (CP) and depression taking ten commonly used antidepressants in Australia (AGDS data only).

21
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. ; https://doi.org/10.1101/2020.05.23.20110841 doi: medRxiv preprint Supplementary Figure 6. Reduced functional benefits associated with antidepressant use in participants with comorbid chronic pain and depression. This forest plot shows the results from logistic regressions to quantify the correlation between chronic pain and self-reported aspects of antidepressant use. Diamonds represent odds ratios (ORs) and horizontal lines depict 95% confidence intervals. *p<0.05; **p<0.05 after correction for seven tests (AGDS data only).

22
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. ; https://doi.org/10.1101/2020.05.23.20110841 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020.

24
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. 25 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. Alcohol use was measured as the number of days drinking three or more standard drinks in the past three months. All other variables were measured on a scale ranging from never using to daily or almost daily over the past three months.

26
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. ; 27 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 24, 2020. 28 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 24, 2020. ; https://doi.org/10.1101/2020.05.23.20110841 doi: medRxiv preprint

RESEARCH IN CONTEXT
Evidence before this study Depression is a highly heterogeneous condition. A common cause of this heterogeneity is comorbidity with other psychiatric and physical diseases. The association between comorbid chronic pain and depression with outcomes such as suicidality and treatment response have been suggested. However, clinically useful findings and recommendations regarding treatment are lacking. Pharmacoepidemiological approaches have been proposed as a new strategy to address the current knowledge gaps and enormous unmet clinical need in comorbid chronic pain and depression populations.

Added value of this study
We examined a large cohort of Australian adults and demonstrated evidence for an increased risk between depression and chronic pain. An overall poorer prognosis including higher suicidality, substance use and lower treatment benefits was associated with comorbid chronic pain and depression. Notably, our findings indicate specific antidepressants have lower effectiveness and generally reduced functional benefits for patients with chronic pain.

Implications of all available evidence
The available evidence indicates comorbidity with chronic pain is a significant source of depression heterogeneity. The controversies underlying several failed antidepressant randomised clinical trials may therefore be due (in part) to not accounting for this common comorbidity. Our results can be immediately implemented in designing more targeted interventional trials to improve treatment guidelines and patient outcomes..

29
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 24, 2020. ; https://doi.org/10.1101/2020.05.23.20110841 doi: medRxiv preprint