Impact of Major Depressive Disorder on Comorbidities: A Systematic Literature Review

J Clin Psychiatry 2022;83(6):21r14328

To cite: Arnaud AM, Brister TS, Duckworth K, et al. Impact of major depressive disorder on comorbidities: a systematic literature review. J Clin Psychiatry. 2022;83(6):21r14328.
To share: https://doi.org/10.4088/JCP.21r14328

© 2022 Physicians Postgraduate Press, Inc.

^aSage Therapeutics, Inc., Cambridge, Massachusetts
^bNational Alliance on Mental Illness (NAMI), Arlington, Virginia
^cDepression and Bipolar Support Alliance, Chicago, Illinois
^dMental Health America of Ohio, Columbus, Ohio
^eTantalus Medical Communications Ltd., Victoria, BC, Canada
*Corresponding author: Ellison D. Suthoff, MBA, 215 First St, Cambridge, MA 02142 (Ellison.Suthoff@sagerx.com).

See commentary by Weir and Rapaport

Major depressive disorder (MDD) is defined as a depressed mood and/or loss of interest or pleasure in daily activities along with symptoms such as weight loss or gain, insomnia or hypersomnia, psychomotor agitation or retardation, fatigue, feelings of worthlessness, diminished ability to think or concentrate, and thoughts of death or suicidal ideation (minimum 5 symptoms total), which have collectively been present during the same 2-week period and represent a change from previous functioning.¹ The episode is considered MDD if it is not attributable to the physiologic effects of a substance or not better explained by another medical condition (such as schizophrenia) and a manic episode or a hypomanic episode has never been observed.¹ MDD affects over 160 million people globally² and is one of the most common mental health disorders in several countries. Notably, depression is recognized by the World Health Organization (WHO) as a major contributor to the overall global burden of disease.³

Complicating our understanding of MDD and its treatment is the presence of several physical and psychological comorbidities among individuals with depression, demonstrating an apparent relationship between physical and mental health. The interaction between depression and many of these comorbidities is not fully understood but, in several cases, appears to be complex and potentially bidirectional. The presence of comorbidities adds to both the humanistic and the economic burden associated with depression; notably, 63% of total MDD-associated costs in the United States in 2018 were attributed to an increased cost of treating comorbid conditions rather than to MDD itself.^4,5 Furthermore, individuals with MDD have been shown to have a decreased life expectancy compared with those without depression⁶; it is therefore possible that the worsening of associated comorbidities could be a contributing factor to earlier mortality.

Although the associations between depression and several comorbidities have been studied individually, not all have been explored systematically. The objective of this review was to qualitatively identify and summarize the breadth of observational data that examine the potential causality of MDD on multiple comorbidities, including the risk of developing new diseases and the impact on the disease course of preexisting comorbidities. Of note, this review explores several comorbidity categories simultaneously to provide a broader illustration of the collective comorbidity risks among people with depression.

METHODS

A systematic literature search for studies that examined the association between MDD and comorbidities was conducted with methods consistent with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guidelines.⁷

Categories of comorbidities were first determined from preliminary literature research, review of national agency and patients’ advocacy group reports, review of US medical claims data, and expert opinion. The final list of comorbidity categories included cancer, central nervous system (CNS) disorders, cardiovascular disease (CVD), metabolic and endocrine diseases, autoimmune and gastrointestinal (GI) diseases, pain-related conditions, respiratory disorders, and substance abuse disorders. This review focused on comorbidities related to physical and substance abuse disorders; anxiety and other psychiatric disorders were not included.

Databases searched included Embase, MEDLINE (including MEDLINE In-Process), Cochrane Database of Systematic Reviews, and PsycINFO, and the search used terms for MDD, comorbidity categories, and observational studies (the full search strategy is provided in Supplementary Table 1). In addition, abstracts from several relevant congresses were reviewed and hand searches of referenced publications were undertaken. Searches were conducted in November 2019 and were restricted to the previous 15 years for database searches and the previous 2 years for conference proceedings.

Included studies assessed the relationship between MDD and comorbidities and were required to use a covariate-adjusted analysis to identify the association of MDD with downstream comorbidities (affecting either the risk of developing a comorbidity or the disease course of an existing comorbidity). Geographically, included studies were restricted to those undertaken in Europe and North America. Covariate-adjusted analyses included those adjusted for at least one relevant covariate to minimize their impact on the association with depression and the comorbidity; examples include demographics such as age and sex, other diseases, and behaviors such as smoking and alcohol consumption. Studies that assessed only associations in the opposite direction (ie, the impact of comorbidities on the development or disease course of depression) were excluded. Included studies were required to be observational in design (including meta-analyses of observational data), and only English-language records were searched; a complete list of criteria is provided in Supplementary Table 2. Although the aim of this review was to identify studies of people with MDD, the heterogeneous way that these individuals are described and identified in scientific literature (such as through symptom rating scales or retrospectively from medical records) did not allow for a strict criterion of physician diagnosis of MDD. Indeed, scales and questionnaires are often used in clinical practice to screen, diagnose, and monitor depression. This review therefore included studies of people with MDD or depression otherwise defined by the authors that was not clearly a diagnosis for a different type or severity level of depression (such as dysthymia).

Search results were screened by 2 separate reviewers initially by titles and abstracts, followed by a review of the full text; any disputes were resolved through discussion between reviewers or consultation with a third reviewer. Data from included studies were extracted by 2 independent reviewers, and any discrepancies between extractions were verified for accuracy by an independent third reviewer. Data describing the study methodology, participant demographic and clinical characteristics, and associations between MDD and comorbidities were extracted. The quality of included studies was assessed by reviewers using the Newcastle-Ottawa Scale for observational studies and the checklist recommended by the National Institute for Health and Care Excellence (NICE) for meta-analyses (see Supplementary Tables 3–6).^8,9

RESULTS

A total of 6,763 articles were identified by database searches for initial screening, and these were combined with another 48 relevant articles identified from conference abstracts and hand searches of references lists from other reviews. Overall, 199 articles met the inclusion criteria (Figure 1); these included 142 cohort studies, 15 cross-sectional studies, 15 case-control studies, 26 meta-analyses of observational data, and 1 ancillary study from a clinical trial. Studies ranged broadly in size from fewer than 50 to nearly 5 million participants (see Supplementary Table 7 for a complete list). A summary of the strength of associations found between depression and the incidence and severity of several comorbidities is provided in Figure 2. Findings from all studies are summarized in the following sections; for brevity, comorbidities with findings of note are reported in tables with all other depression-associated outcomes presented in Supplementary Tables 8–31.

Cancer

A total of 12 observational studies (13 publications) and 2 meta-analyses were identified through this systematic review. One meta-analysis¹⁰ of 32 studies showed that although an analysis of studies that included patients with cancer and depressive symptoms demonstrated a significant association between depression and mortality (hazard ratio [HR] = 1.09; 95% confidence interval [CI]: 1.03 to 1.15; P = .003), when a clinical depression diagnosis was required this association had a wider confidence interval and statistical significance was no longer demonstrated (HR = 1.67; 95% CI, 0.96 to 2.90; P = .07). Furthermore, the meta-analysis also showed no significant association between depressive symptoms and cancer recurrence (risk ratio [RR] = 1.23; 95% CI, 0.85 to 1.77; P = .275).¹⁰ Among individual studies identified by this systematic review, we did not find a consistent and significant impact of depression on the new incidence of cancer nor on assessments of cancer severity/mortality of patients with preexisting cancer (see Supplementary Tables 8 and 9 for complete information). For example, in the Baltimore Epidemiologic Catchment Area study¹¹ that had the longest duration of follow-up of all identified studies (24 years), a lifetime history of major depressive episodes (MDE) was associated with the risk of any incident cancer (adjusted HR = 1.87; 95% CI, 1.16 to 3.01); however, the adjusted analysis lost significance in a subgroup that excluded 145 respondents who rated their health status as poor at baseline.

CNS Disorders

This review identified a total of 29 observational studies (in 30 publications) and 3 meta-analyses assessing the relationship between depression and CNS disorders. Consistently, depression was demonstrated to be significantly associated with an increased incidence of new dementia and Alzheimer’s disease or cognitive decline in people with existing disease, and with the incidence of Parkinson’s disease.

Incident dementia and Alzheimer’s disease. Two meta-analyses^12,13 (in 23 and 20 studies, respectively) that assessed the association of depression with incident dementia (pooled odds ratio [OR] = 1.96; 95% CI, 1.64 to 2.34; P < .0001; 23-study meta-analysis only) and/or Alzheimer’s disease (pooled ORs for Alzheimer’s disease = 1.85; 95% CI, 1.45 to 2.37; P < .0001 and 1.98; 95% CI, 1.76 to 2.24; P < .001, respectively) showed significant associations overall and across all subgroups. Among another 22 studies (in 23 publications) identified by the review for this comparison, most showed a significant association between depression and the incidence of dementia and many also showed a significant association between depression and the incidence of Alzheimer’s disease (see Supplementary Table 10 for complete information).

Dementia and Alzheimer’s disease severity. Five studies assessed the impact of depression on the severity of preexisting dementia and/or Alzheimer’s disease and nearly all (4 of 5) showed significant and positive associations between presence/severity of depression and cognitive decline/worsening of symptoms in these cohorts (Table 1). Additional details for associations between depression and CNS disorders, including findings in people with epilepsy and Parkinson’s disease, can be found in or subsequent to Supplementary Table 10.

Cardiovascular Disease

The association between MDD and CVD has been studied widely, and the review identified 67 observational studies and 14 meta-analyses that assessed the association between depression and incidence or worsening of CVD. Most studies identified an association between MDD and the subsequent incidence or worsening of CVD across several disease categories, including general CVD (or combined endpoints), heart failure, hypertension, ischemic heart disease/coronary artery disease, and myocardial infarction (MI; see Supplementary Tables 11–21 for complete information). In this systematic literature review, measures of disease worsening included subsequent cardiovascular events among individuals with prior events/existing chronic CVD as well as cardiac mortality in both the general populations and populations of people with preexisting CVD; the association between depression and these various measures of disease worsening was consistently significant across most analyses (see Supplementary Tables 11–23 for complete information). Among studies that took depression severity into account, these associations often became stronger as depression severity increased (see Supplementary Tables 11–23 for complete information). Furthermore, studies that assessed the duration of MDD and/or number of MDD episodes also found significant associations between greater durations or episode numbers and ischemic heart disease mortality, incidence of hypertension and heart disease, incidence of major coronary heart disease events, and incidence of acute MI (see Supplementary Tables 14–17 and 20 for complete information).^19–22

Stroke incidence and severity. Of the many CVD categories, the relationship between depression and stroke incidence had the greatest variability. The review identified 3 meta-analyses (assessing 28 studies [2 meta-analyses each] and 30 studies, respectively), which consistently demonstrated significant associations between depression and incident stroke.^23–25 Findings from the individual studies identified by the review (15 in total), however, were mixed, with some showing a significant association for this relationship and others showing mixed results across different analyses, while another group of studies showed no significant associations (see Supplementary Table 22 for complete information). Among 3 studies and 1 meta-analysis of 6 studies included in this review that assessed the association between depression and stroke severity (generally measured as poststroke function or recovery), the meta-analysis reported a significant association between depression and severe long-term disability among people with stroke (OR = 2.16; 95% CI, 1.70 to 2.77),²⁶ and results from individual studies generally reflected these findings, with 2 of 3 studies that assessed poststroke function/recovery reporting a greater likelihood of disability in patients with depression (see Supplementary Table 23 for complete information).

Metabolic and Endocrine Disorders

Incident diabetes. In total, 17 studies (including 2 analyses from metabolic syndrome studies) and 4 meta-analyses assessed the relationship between depression and a subsequent diagnosis of diabetes with follow-up times ranging from 3 to 17 years. The meta-analyses showed a consistently significant association between depression and incident diabetes in the follow-up period both overall and across several subgroup and sensitivity analyses.^27,29,30,45 Six individual study analyses focused on type 2 diabetes only, 1 study each focused on type 1 diabetes and gestational diabetes only, and the remaining studies either did not clarify diabetes type or included both types 1 and 2. Several individual studies reflected the significant associations found in the meta-analyses (either overall or in 1 or more subgroups of participants; Table 2).

Among studies that presented stratified analyses by age and sex, no clear and consistent trends were observed,^32,38,39 although it was notable that a meta-analysis showed a significant association between incident type 2 diabetes in men and depression (RR = 1.57; 95% CI, 1.24 to 1.99), whereas this association was not significant among women (RR = 1.26; 95% CI, 0.95 to 1.67).⁴⁵ Three studies evaluated the impact of depression severity on the increased risk of incident diabetes. In one analysis, people with severe depression had greater odds of developing type 2 diabetes compared with those who had no depression (OR = 1.42; 95% CI, 1.01 to 1.99), and at the same time antidepressant use (regardless of depression severity) was also significantly associated with incident diabetes (OR = 2.76; 95% CI, 1.93 to 3.94), suggesting that medication use could be indicative of a unique depression-diabetes association.³⁷ By contrast, the Zaragoza Dementia and Depression (ZARADEMP) study, which adjusted for antidepressant use as a covariate, found an association between nonsevere depression and incident diabetes (HR = 1.66; 95% CI, 1.01 to 2.75; P = .048), but this did not remain significant in people with severe depression.³³ A separate analysis of the Nurses’ Health Study restricted to participants without depression at baseline found that women had a higher risk of developing type 2 diabetes as their depressive symptoms, measured by the 5-item Mental Health Index (MHI-5) score (categorized into 4 groups in order from best to worst mental health: 86–100 [reference case], 76–85, 53–75, and 52 or lower), worsened over time to 53–75 compared with those who had scores that remained between 86 and 100 (RR = 1.13; 95% CI, 1.02 to 1.26), but this association was no longer significant when scores decreased to a threshold of 52 and below, which was considered “depressed mood.”⁴¹

Diabetes severity. Six studies, including 3 separate analyses of the Pathways Epidemiologic Study, assessed the association between depression and diabetes severity (type 2 diabetes only in 5 of 6 studies; 1 study included both type 1 and type 2 diabetes). A consistent association between presence/severity of depression with diabetes severity, measured by glycemic control, macrovascular and microvascular events, and diabetic retinopathy, was demonstrated across most analyses, with all 6 studies reporting at least one positive association between depression and diabetes severity (see Supplementary Table 25 for complete information).

Incident metabolic syndrome. The association between depression and a subsequent diagnosis of metabolic syndrome was reported in 3 studies. In general, most analyses did not show a significant association between depression presence or recurrence and the incidence of metabolic syndrome. In one exception, a subgroup of men in the cross-sectional Study of Health In Pomerania (SHIP-0; but not the similar SHIP-TREND-0) study had a significant association between a history of depression at the syndromal level and metabolic syndrome (OR = 1.53; 95% CI, 1.06 to 2.21; P ≤ .05), but other subgroups of men (those with lifetime MDD and those from the SHIP-TREND-0 study) and all parallel subgroups of women did not report similar findings.⁴³ People with recurrent depression in the Study of Women’s Health Across the Nation (SWAN) study also demonstrated nonsignificant associations between history of or current MDE as a predictor of metabolic syndrome (HR = 1.83; 95% CI, 0.99 to 4.76), and when this was no longer restricted to recurrent depression only the association weakened further (HR = 1.54; 95% CI, 0.93 to 3.40).⁴⁴ In the prospective Cohorte Lausannoise (CoLaus)/Psychiatric arm of the CoLaus Study (PsyCoLaus) population cohort that assessed the relationship by MDD subtypes, only atypical MDD (OR = 2.49; 95% CI, 1.30 to 4.77; P < .01), and not melancholic (OR = 1.45; 95% CI, 0.78 to 2.69) or unspecified MDD (OR = 1.44; 95% CI, 0.83 to 2.49), was associated with incident metabolic syndrome over approximately 5 years.⁴⁶

Incident obesity. The review identified 10 studies (including 2 analyses of abdominal obesity from metabolic syndrome studies) and 2 meta-analyses that assessed the association between depression and incident obesity (see Supplementary Table 26 for complete details). Findings varied across studies; however, there was a notable trend of a stronger association between depression and incident obesity among women compared with men. In particular, both meta-analyses demonstrated a significant association between depression and incident obesity among all participants and among women-only subgroups but not among men-only subgroups.^47,48 The use of antidepressants has a known association with obesity; the Canadian National Population Health Survey (NPHS) demonstrated that both use of venlafaxine (a serotonin and norepinephrine reuptake inhibitor; HR = 4.9; 95% CI, 1.8 to 13.0; P < .001) and use of selective serotonin reuptake inhibitors (SSRIs; HR = 1.9; 95% CI, 1.2 to 3.2; P < .01) were significantly associated with incident obesity, whereas a diagnosis of MDE was not (with or without covariate adjustment for sex, age, overweight body mass index at baseline, level of activity, and antidepressant use).⁴⁹ The meta-analyses did not indicate any adjustment for medication use; thus, the potential impact of antidepressants on the findings among men and women subgroups should be taken into account.⁴⁷ Of the 10 individual studies identified, only 3 mentioned an adjustment for antidepressant medication; these studies still reported significant associations between depression and obesity in at least some subgroups analyzed (see Supplementary Table 26 for complete details).

Autoimmune/GI Disorders and Musculoskeletal/Pain Conditions

Incident autoimmune diseases. Six studies assessed the impact of depression on the incidence of various autoimmune diseases (see Supplementary Table 27 for complete details). The most comprehensive study was a prospective, population-based analysis of “any” autoimmune disease (including ankylosing spondylitis, autoimmune thyroiditis, celiac disease, Crohn’s disease, idiopathic thrombocytopenic purpura, iridocyclitis, multiple sclerosis, primary adrenocortical insufficiency, psoriasis vulgaris, rheumatoid arthritis, Sjögren’s syndrome, systemic lupus erythematosus, thyrotoxicosis, type 1 diabetes mellitus, and ulcerative colitis) in a series of linked databases assessing a cohort of over 1 million people in Denmark over a 17-year period.⁵⁰ In this analysis, a history of depression was significantly associated with the incidence of any autoimmune disease (IRR = 1.25; 95% CI, 1.19 to 1.31; P < .01).⁵⁰ The incidence of certain autoimmune disorders in the same population-based cohort was also shown to be significantly associated with a history of depression, including Crohn’s disease (IRR = 1.36; 95% CI, 1.16 to 1.60; P < .01), systemic lupus erythematosus (IRR = 1.38; 95% CI, 1.00 to 1.91; P <  0.01), and psoriasis (IRR = 1.45; 95% CI, 1.13 to 1.85; P < .01).⁵⁰ By contrast, a history of depression was not significantly associated with the incidence of ulcerative colitis, celiac disease, or ankylosing spondylitis.⁵⁰

For the multiple sclerosis comorbidity, the Danish population-based cohort demonstrated a significant association between a history of depression and incident multiple sclerosis (IRR = 1.46; 95% CI, 1.26 to 1.69; P < .01).⁵⁰ In addition, incident multiple sclerosis was significantly associated with a single episode of depression (IRR = 1.48; 95% CI, 1.27 to 1.74; P < .01) but not multiple episodes of depression (IRR = 1.30; 95% CI, 0.88 to 1.92).⁵⁰ In alignment with these findings, a study of the Swedish National Patient Register also demonstrated a significant association between depression (overall [HR = 1.86; 95% CI, 1.73 to 2.00; P < .001] and severe depression [HR = 1.46; 95% CI, 1.27 to 1.68; P < .0001]) and multiple sclerosis.⁵¹

The impact of depression on incident rheumatoid arthritis was mixed. The Danish population cohort study reported no significant associations between a history of depression and incident rheumatoid arthritis (IRR = 1.01; 95% CI, 0.90 to 1.44),⁵⁰ whereas an analysis of The Health Improvement Network (THIN) database found a strong significant association between depression and incident rheumatoid arthritis over a nearly 7-year follow-up (HR = 1.38; 95% CI, 1.31 to 1.46; P < .0001).⁵² A UK-based case-control study showed that, among a group of people with recurrent depression, there were elevated odds of rheumatoid arthritis (OR = 2.72; 95% CI, 1.31 to 5.63).³⁵ However, when corrected for multiple testing, this association was not considered to be significant (P = .10). The Canadian NPHS study reported that MDD at baseline led to a significant association with arthritis/rheumatism over the 8-year follow-up (HR = 1.7; 95% CI, 1.3 to 2.2); when assessed based on duration of past-year depressive episodes, this was significant only for those with a depression episode of 13–52+ weeks (HR = 2.2; 95% CI, 1.5 to 3.3) and not for those with a duration of 2–12 weeks (HR = 1.2; 95% CI, 0.8 to 1.7).¹⁹

Incident headache and other chronic pain. Four studies (5 publications) described the association between depression and migraine or headache; most (3 of 4) did not show a significant association in the fully adjusted models for baseline depression and risk of “any” migraine, although some significant associations were reported between depression presence/severity and migraine with aura (see Supplementary Table 28 for complete details). Three studies that assessed other pain-related outcomes (spinal pain, temporomandibular pain, and back problems) generally did not show an association between depression presence or severity and incident pain, with the exception of back problems reported in 1 study (see Supplementary Table 28 for complete details).¹⁹ In parallel with the single-study finding, a meta-analysis of 11 studies assessing the association between depression and low-back pain reported a pooled OR of 1.59 (95% CI, 1.26 to 2.01), demonstrating a significant association, and this relationship remained significant across all subgroup and sensitivity analyses (method of depression diagnosis, whether studies adjusted for confounders, and studies restricted to older participants).⁵³ The meta-analysis also showed that this association was affected by depression severity, with a significant association observed for people with the most-severe level of depression (OR = 2.51; 95% CI, 1.58 to 3.99) but not for the lowest level of depression severity (OR = 1.51; 95% CI, 0.89 to 2.56).⁵³

Infectious Diseases

Searches for infectious diseases identified 3 studies assessing the relationship between MDD and HIV infection. A US-based Medicaid-eligible cohort analysis of over 4 million people demonstrated that baseline MDD was associated with increased odds of incident HIV/AIDS in people both with (OR = 3.04; 95% CI, 2.75 to 3.36; P < .001) and without (OR = 1.12; 95% CI, 1.04 to 1.21; P < .01) concurrent substance abuse disorder.⁵⁴ Furthermore, among HIV-infected individuals, the odds of low HIV RNA levels (ie, maintaining viral suppression) at 12 months were significantly decreased in those with depression who did not use SSRIs compared with those who did not have depression (OR = 0.77; 95% CI, 0.62 to 0.95; P = .02).⁵⁵ Similar findings were observed in a study that used generalized linear mixed models to show that individuals classified as having “moderate-increasing” depression severity led to significantly greater odds of a low CD4 count (ie, disease worsening) compared with “low-chronic” depression (OR = 1.53; 95% CI, 1.08 to 2.19).⁵⁶

Respiratory Disorders

The review identified 3 studies assessing the relationship between MDD and respiratory disorders, including asthma and bronchitis. Although several significant associations were identified, directionality was often unclear. For example, in a case-control analysis based in the United Kingdom, people with recurrent depression reported significantly higher lifetime asthma (OR = 2.19; 95% CI, 1.53 to 3.13; P = .00022), but the methods did not permit clear confirmation of whether depression preceded asthma.³⁵ Similarly, an analysis from the cross-sectional National Health and Nutrition Examination Survey (NHANES) study demonstrated a significantly increased prevalence of asthma according to both presence and severity of depression (OR = 3.18; 95% CI, 2.37 to 4.26; P < .01 for moderately severe; OR = 3.95; 95% CI, 2.38 to 6.56; P < .01 for severe depression).⁵⁷ However, the directionality was again unclear because the cross-sectional study design did not account for the order in which these disorders occurred in participants.⁵⁷ The Canadian NPHS study showed significant associations with incident asthma and bronchitis in both people who had MDD at baseline (HR = 1.8; 95% CI, 1.3 to 2.5 for asthma; HR = 2.1; 95% CI, 1.5 to 2.9 for bronchitis) and those who had MDD assessed as a time-varying characteristic (HR = 1.7; 95% CI, 1.2 to 2.4 for asthma; HR = 2.6; 95% CI, 1.9 to 3.7 for bronchitis) throughout the 8-year follow-up, regardless of duration of depression and covariates added to the model.¹⁹

Substance Abuse

A total of 23 observational study publications that described the association between depression and substance abuse were identified by the review, including multiple analyses from several large studies such as the Canadian NPHS cohort, the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) cohort, and the cross-sectional annual National Survey on Drug Use and Health (NSDUH) survey in the United States. Results varied widely among studies that assessed the impact of depression on incident alcohol or cannabis abuse or related disorders. In general, a greater proportion of studies did not show a significant association with incident drug or alcohol abuse, although there appeared to be a consistent relationship between MDD severity and these incidences (see Supplementary Table 30 for complete information). Furthermore, studies showed a consistent impact of depression on the severity of alcohol abuse, particularly among the young adult cohorts (see Supplementary Table 31 for complete information).

Quality Assessment

Quality assessment was conducted for 198 of the 199 included studies; 1 study did not undergo quality assessment because it lacked a validated checklist. Most studies were assessed to be truly or somewhat representative of the average population in the community and selected their comparator group (cohort, controls) using sufficient methods from either the same population or the general community. In addition, most studies were deemed to adequately control for the most important factors and often for additional important factors; this was likely a reflection of the covariate adjustment required by the review inclusion criteria. However, several cohort studies lost > 20% of participants to follow-up or provided insufficient information about follow-up rates; approximately half of the case-control studies provided no description of the nonresponse rate (ie, dropouts) among both groups; and all cross-sectional studies either did not describe their statistical test or used one that was incomplete or not appropriate. Further details of the quality assessment for individual studies are reported in Supplementary Tables 4–7.

DISCUSSION

In a systematic review of nearly 200 studies, the presence of MDD was identified as a risk factor for both the development and the worsening of a range of comorbidities in several categories, including CNS disorders (eg, dementia/Alzheimer’s disease and Parkinson’s disease), CVD (eg, general CVD, ischemic heart/coronary artery disease, MI, and heart failure), metabolic and endocrine disorders (particularly diabetes in men and obesity in women), certain autoimmune disorders (such as Crohn’s disease, psoriasis, and multiple sclerosis), and substance use disorders. These associations were observed consistently despite the notable variability in methodology across studies, including population sampled, criteria for depression, length of follow-up, and covariates included in the analysis. Although the association was less consistent between depression and other comorbidity categories, significant associations were often observed for certain subgroups or specific relationships. For other associations, such as those in the cancer category, wide confidence intervals were often observed, which was likely influenced to some extent by the heterogeneity of these comorbid diseases. It should also be noted that within multiple studies, certain associations strengthened with increasing depression severity or with increasing number of depressive episodes.

The effects of depression on comorbidities can arise both directly, through biological pathways, and indirectly, through a reduced ability to care for oneself or other risky health behaviors. Several biological mechanisms have been implicated in the potential relationship between depression and physical comorbidities; many of these involve dysfunction in the hypothalamic-pituitary-adrenal (HPA) axis and its impact on cortisol levels and the immune system. In particular, it has been suggested that elevated cortisol is responsible for activating cancer cell growth pathways in vitro⁵⁸ and for hippocampal atrophy, accumulation of amyloid-β plaques, inflammatory processes, blood flow alterations, and lack of nerve growth factors in people with dementia and Alzheimer’s disease.^59,60 Inflammation, whether activated by HPA axis dysfunction or other mechanisms such as the sympathetic nervous system, has been suggested as a contributing factor to the biological mechanisms underpinning depression, and the association between depression and several inflammation-related disorders (such as Crohn’s disease and coronary heart disease) could be thus explained to some degree by inflammation.⁶¹ Elevated cortisol may also be responsible for visceral fat accumulation, insulin resistance (type 2 diabetes), and disturbances in lipid metabolism in people with metabolic and endocrine disorders.^62,63 It also remains possible, however, that depression could be a prodromal symptom of certain neurologic disorders, which was acknowledged by authors of included studies that identified associations between depression and dementia,⁶⁴ Parkinson’s disease,^65,66 and stroke.⁶⁷ However, in a sensitivity analysis of the Pathways Epidemiologic Study that excluded people with an International Classification of Diseases, 9th revision (ICD-9) diagnosis of dementia in the 2 years after baseline, depression remained associated with an increased risk of dementia, suggesting that depression was less likely to be a prodromal symptom or secondary to dementia.⁶⁸ Depression could be both a causal factor and a prodromal symptom for these neurologic disorders, and additional research could further clarify the directionality of this relationship.

In addition to biological mechanisms, depression may impact a person’s ability to take care of themselves, seek care, and adhere to treatment recommendations, which could lead to an increased risk of developing comorbidities and/or a worsened disease course. For example, many studies acknowledged the role that diet and lifestyle factors could play, in addition to biological mechanisms, for the association between depression and CVD.^69–72 In addition, many substance abuse studies focused on the self-medication hypothesis as a primary factor that was likely to be mediating the association between depression and substance abuse.⁷³

The relationship between physical and mental health has several implications for the integration of care and the need for further education among physicians, patients, caregivers, and society. Certain clinical guidelines as well as various government and patient organizations have begun to address comorbidities among people with MDD to varying degrees. For example, both the Canadian Network for Mood and Anxiety Treatments (CANMAT) 2016 guidelines and the US Centers for Disease Control and Prevention (CDC) note that depression is an independent risk factor for several diseases.^74,75 The CDC recommends actively addressing mental health disorders early along with providing support to improve healthy behaviors as a strategy to decrease the risk of physical comorbidities such as cardiovascular events⁷⁶; suggested actions to promote heart disease prevention include integrating mental health into multidisciplinary teams and incorporating mental health screening in the care of other diseases. However, there remains considerable room to further acknowledge and explore the relationship between depression and comorbidities and to update guidelines and other educational tools to incorporate recommendations that are informed by the link between mental and physical health.

Several limitations should be acknowledged when interpreting the results of this review. First, the observational study search and screen were date-limited to studies published between the years 2005 and 2020, and therefore any relevant studies outside of this range were not included in the evidence summary. In addition, based on their design, some studies may not present a true causal relationship in which depression occurred before the comorbidity in all individuals. It should also be noted that some of these findings could be explained to a degree by changes in patient care following a depression diagnosis and not an actual causal relationship between depression and a comorbidity. For example, if a person is diagnosed with depression, he or she may undergo more frequent contact with health care providers, which could in turn lead to the identification of diseases that had previously gone undiagnosed.

In conclusion, the presence of MDD was identified as a statistically significant risk factor for both the development and the worsening of a range of comorbidities. Collectively, these results highlight that depression may impact many comorbidities concurrently and could thus have a considerable negative impact on a person’s whole health. Additional research assessing the combined comorbidity risks may be warranted to further our understanding of the collective burden of depression-associated comorbidities. Furthermore, addressing depression with appropriate services, treatment, and support could have a broad positive impact on physical health, thus renewing the importance of appropriately timed screening and diagnosis of depression and adequate treatment of MDD.

Submitted: November 18, 2021; accepted June 27, 2022.
Published online: October 19, 2022.
Relevant financial relationships: Mr Suthoff is a current employee and Ms Arnaud and Dr Werneburg are former employees of Sage Therapeutics, Inc., and own stock/stock options. Drs Reinhart and Aleksanderek are current or former employees of Tantalus Medical Communications, Ltd, which received financial support from Sage Therapeutics, Inc., and Biogen, Inc., for research and medical writing support. During the peer review process, both Sage Therapeutics, Inc., and their collaboration partner Biogen, Inc., had the opportunity to review and comment on this manuscript; however, the authors had full editorial control of the manuscript and provided final approval on all content. Ms Fulwider has received consulting fees from Sage Therapeutics, Inc. Drs Brister and Duckworth and Ms Foxworth have nothing to disclose.
Funding/support: This work was supported by Sage Therapeutics, Inc., and Biogen, Inc.
Role of the sponsor: Employees of Sage Therapeutics, Inc., contributed to the protocol design and manuscript drafting and review but had no role in the screening and study selection for inclusion in the systematic review.
Previous presentation: Portions of this work were previously presented as an abstract at the International Society for Pharmacoeconomics and Outcomes Research Annual Meeting; May 18–20, 2020 (virtual meeting).
Acknowledgments: The authors thank Katie Stenson, MSc, a former employee of Sage Therapeutics, Inc., for contributions to the design and interpretation of the systematic review and Shawn Eapen, MSc, and Sara Watchko, MPH, both affiliates of Tantalus Medical Communications (Victoria, BC, Canada) for assistance with the systematic review. Mr Eapen and Ms Watchko have no conflicts of interest to declare. The authors thank Tantalus Medical Communications for providing research and medical writing support, which was funded by Sage Therapeutics, Inc., and Biogen, Inc., in accordance with Good Publication Practice (GPP3) guidelines.
Supplementary material: Available at Psychiatrist.com.

Clinical Points

• Although there are many well-studied comorbidities among people with depression, a broader illustration of how the risk and severity of several comorbidities across different disease areas are simultaneously impacted by the presence of depression is less clear.
• Providers should ensure that people have appropriate access to screening, diagnosis, and adequate treatment not only for depression, but also for the comorbidities that could be impacted by depression.