Psychiatric Annals

CME Article 

Impact of Medical Comorbidity in Biomarker Discovery for Major Depressive Disorder

Andrew Czysz, MD, PhD


The interaction between major depressive disorder (MDD) and other medical conditions is increasingly recognized as an important factor for clinical management and for biomarker development. Many common medical conditions have a bidirectional epidemiologic relationship with MDD that may have roots in shared pathophysiologic, environmental, and psychosocial factors. Furthermore, depression is increasingly recognized as a systemic disease. Understanding this relationship is challenging because of the immense variability between patients with MDD and also within other medical disorders themselves. This includes not only specific symptoms, but also disease severity, duration, disability, and treatment resistance. [Psychiatr Ann. 2020;50(6):239–243.]


The interaction between major depressive disorder (MDD) and other medical conditions is increasingly recognized as an important factor for clinical management and for biomarker development. Many common medical conditions have a bidirectional epidemiologic relationship with MDD that may have roots in shared pathophysiologic, environmental, and psychosocial factors. Furthermore, depression is increasingly recognized as a systemic disease. Understanding this relationship is challenging because of the immense variability between patients with MDD and also within other medical disorders themselves. This includes not only specific symptoms, but also disease severity, duration, disability, and treatment resistance. [Psychiatr Ann. 2020;50(6):239–243.]

Understanding the relationship between major depressive disorder (MDD) and other medical conditions is critical for developing biomarkers that aid in the diagnosis, treatment, and prevention of MDD. Candidate biomarkers for depression have expanded beyond monoamine targets and now include pathways across the entire body.1 Particular attention, however, has focused on biomarkers related to metabolic, cardiovascular, neurologic, and immunologic processes, although many large clinical trials in depression contributing to this knowledge base exclude patients with significant medical comorbidity. These candidate biomarkers include laboratory tests that are already readily available like high-sensitivity C-reactive protein and hemoglobin A1c, as well as molecular and imaging markers that are not yet used at all in clinical care. It is also increasingly recognized that diseases related to these systems are disproportionately prevalent in patients with MDD. Together, these observations raise two important questions. First, which medical conditions are more prevalent in patients with depression? Second, when comorbidity exists, does one disease cause the other, or are there common risk factors? Biomarkers for both depression and these medical comorbidities across the trajectory of the disease courses are necessary to elucidate answers to these questions.

Epidemiology of Medical Comorbidity With MDD

In a longitudinal analysis from the Canadian National Population Health Survey, the presence of any long-term medical condition was associated with nearly double the risk of developing a depressive episode at 2-year follow-up.2 Similar results have been found in specific disorders, such as diabetes, myocardial infarcts, cerebrovascular disease, HIV, cancer, and multiple sclerosis.3 Other studies have examined this association in the opposite direction—the onset of medical conditions in patients with MDD. For example, patients with a history of depression had about twice the risk of developing obesity.4

These relationships become more complex when one considers how much the presentation of the same illness can vary between patients. Depression is a heterogenous syndrome, as are the most common medical comorbidities. The Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5),5 requires 5 of 9 possible symptoms to make a diagnosis of MDD; for any two patients, the symptom overlap may be minimal. Additionally, because symptoms like sleep and appetite can be either a deficit or excess, overlap may be nonexistent. These criteria only address core symptoms of depression, whereas many measures of depression like the Inventory of Depressive Symptomatology assess 30 different items.6 This includes several somatic symptoms, such as gastrointestinal discomfort and libido. In addition to the presence of symptoms, depression can also vary based on intensity, duration of illness, extent of disability and treatment resistance. Finally, the DSM-5 defines several depressive disorders, including bipolar disorders, persistent depressive disorder, substance/medication-induced depressive disorder, depressive disorder due to another medical condition, other specified depressive disorder, unspecified depressive disorder, and adjustment disorder with depressed mood.5

Depressive Disorder Due to Medical Conditions or Treatments

Because medical conditions and medical treatments can initiate symptoms of depression, it can be challenging to determine where best to attribute individual symptoms. The DSM-5 distinguishes MDD from depressive disorder due to another medical condition when the onset of depressive symptoms is best explained by a medical illness.5 A detailed symptom history coupled with appropriate laboratory testing is necessary to make this diagnosis. Onset of depressive symptoms should follow the onset of the medical problem and respond well to medical treatment. In some cases, however, a medical condition may be contributing to the severity of an otherwise independent depressive episode, or symptoms may not fully resolve with medical treatment. In these cases, true comorbidity should be considered rather than a medically induced depressive disorder.

Anemia, sleep apnea, and thyroid dysfunction are among the most common medical conditions that directly cause depressive symptoms.7 Like MDD, hypothyroidism alone can present with decreased energy, anhedonia, and sleep disturbance. Hyperthyroidism can also present with sleep disturbance but is more so associated with more anxious and/or irritable mood. Anemia can also present with symptoms of depression, notably fatigue. These conditions all have readily available laboratory tests (thyroid-stimulating hormone, hemoglobin) that can act as biomarkers for the medical condition; although they may correlate with depressive symptoms, they are not necessarily concordant. Additionally, it remains unclear how to incorporate these tests alongside other depression biomarkers. Hemoglobin may be problematic with functional magnetic resonance imaging because this test relies on blood flow. Finally, sleep apnea is a medical cause of depressive symptoms. Hypoxic events during sleep disrupt normal sleep architecture, which can result in debilitating fatigue. This fatigue, much like with anemia and hypothyroidism, can lead to other depressive symptoms, such as decreased interest in pleasurable activities and excessive guilt related to functional impairment from fatigue.

Treatments for certain medical problems have also been associated with onset or worsening of depressive symptoms.5 Depressive symptoms are a well-described side effect of interferon therapy.8 Fortunately, newer therapies for hepatitis C and multiple sclerosis have largely, but not completely replaced the use of interferon. Corticosteroid use is also associated with changes in mood. Acute treatment is associated with manic symptoms, whereas withdrawal from acute corticosteroid treatment is associated with onset of depressive symptoms. Chronic corticosteroid use on the other hand is primarily associated with depressive symptoms.9 According to DSM-5, depressive symptoms that are best explained by a medical treatment are categorized as a substance/medication-induced depressive disorder, rather than MDD.

Trajectory of Depression and Chronic Medical Conditions

It is much more difficult to tease out the relationship between other medical conditions, particularly chronic illnesses, with depression. Conditions like type 2 diabetes are framed as gradually controlled over time rather than cured,10 and hence linking improvements with depressive symptoms is challenging. As mentioned previously, most of these conditions have some objective biologic marker of current disease state (eg, hemoglobin A1c for diabetes) but factors like severity, duration, disability, and treatment resistance are not necessarily captured in these markers. Hence biomarkers for these conditions do not always provide the complete history of disease burden. A patient with type 2 diabetes may currently have a well-controlled hemoglobin A1c, but still suffer from grave diabetic neuropathy due to years of poorly controlled glucose. Additionally, acute events like stroke may have very specific anatomical lesions that vary among patients. Two patients may have a similar amount of damaged tissue, but depressive sequalae may vary depending on the brain circuits that are affected.11

So, when considering the impact of a medical illness on depression, time course of disease can be relevant. Conditions like myocardial infarcts or cerebrovascular events can have rapid and intense impacts on human physiology with prolonged sequelae. In contrast, diabetes or hypertension have more insidious onset with culminative worsening over time. Finally, conditions like multiple sclerosis or asthma can manifest with relapsing and remitting patterns. On top of this, depression itself can manifest in all these patterns.12 Major life stressors can trigger the onset of major depressive episodes, but for many, onset is gradual and persistent. For most patients, however, depressive symptoms have clear episodic worsening and improvement. Connecting these variable patterns of depressive illness with the heterogenous course of medical illness makes interpreting cross-sectional biomarkers exceptionally challenging.

The significance of time course depends on the type of biomarker. Genetic markers like single nucleotide polymorphisms (SNPs) are stable across the lifespan; therefore, their assay is not dependent on when in the disease course they are analyzed.13 Thus, they can help define disease risk, but offer little insight into predicting disease burden at a given moment. Furthermore, they do not incorporate environmental factors that can influence disease course, such as early life trauma. In contrast, cytokines and metabolomic markers are much more fluid and state dependent.1 Unlike SNPs, they are influenced by environmental factors like chronic stress. These markers are well suited for clinical use, as they can change over time and inform clinicians about disease progression and improvement. Both types of markers may be useful tools for predicting treatment response, but none are currently validated for use in clinical care. SNP markers are already commercially available as predictors of antidepressant metabolism, but their utility remains debated.14

To date there are no single genetic markers that are sensitive or specific for predicting onset of depression. Instead, there are numerous markers that are enriched in depressed populations.15 Interestingly, some of these are independently associated with other medical conditions. Replication of SNP findings itself is challenging but can be further complicated when looking at patients with specific medical comorbidity. For example, several SNPs previously associated with depression were analyzed in a large cohort of patients with type 2 diabetes.16 Three of these a priori SNPs were enriched with depressive symptoms, but in the opposite direction as previously described.

Shared Pathophysiology and Mechanisms

Although there is a bidirectional epidemiologic relationship between medical illnesses and depression, causality remains unclear. That is, does depression lead to other conditions, do these conditions lead to depression, or are there common risk factors?

Metabolic Mechanisms

Insulin resistance and obesity are associated with each other but also independently with depression. Much of this work has focused on dysregulation of the hypothalamic pituitary axis, including its role as a mediator between early life adversity, depressive symptoms, and metabolic disease.17 Chronic hyperactivity of this axis leads to cortisol resistance and increased inflammatory cytokines signaling that negatively impact both insulin resistance and depressive symptoms. In addition, molecular pathways first recognized in diabetes are being studied in the context of depression, and cognitive dysfunction related to depression is increasingly recognized as a sequalae of diabetes. Recent work has implicated both sphingolipids and adipokines as relevant to both depressive symptoms18 and treatment response.19 Selective attention to negative stimuli is a common cognitive symptom of MDD.2,20 Interestingly, both body mass index and waist circumference have also been negatively associated with accuracy and processing time on selective attention tasks.21,22

Cardiovascular Mechanisms

Serotonin signaling may explain some of the overlap between depression and cardiovascular disease due to the role of serotonin in platelet activation.23 In a study that compared to both healthy controls and people who are not depressed with ischemic heart disease, people who are depressed had higher levels of platelet factor 4 and beta-thromboglobulin, suggesting increased platelet activation.24 Additionally, polymorphisms in the serotonin transporter gene have also been associated with additional cardiac events after an acute myocardial infarction.25

Neurologic Mechanisms

Mechanisms of depression associated with neurologic conditions may be more direct. This includes direct injury from stroke that impairs monoamine signaling.26 Other pathways may be relevant too, especially given the emerging role of glutamate in depression.27 In support of this, glutamate-mediated excitotoxicity after stroke has been associated with onset of depressive symptoms.28 Finally, the neuroinflammatory response after stroke,29 but also in other neurologic conditions, such as multiple sclerosis,30 may mediate shared pathophysiology. In addition to direct neuroinflammatory mechanisms, peripheral inflammation has been extensively studied with depression. Because of its depth, this topic is covered in a companion piece to this article.31 Potentially, biomarkers that establish excitotoxicity versus inflammatory pathophysiology could predict whether a patient would benefit more from a glutamate-based therapy or treatment that targets inflammation.

Psychosocial Mechanisms

Finally, it is also important to note the psychosocial impact of medical conditions. For example, a phenomenon called “diabetes distress” has been associated with the onset of type 2 diabetes. This includes stress related to taking daily medication, tracking blood glucose, and worry over the long-term impact on a patient's health.32 Furthermore, functional impact can occur related to disability from a medical condition. Diabetic wounds, or worse, amputations, can have a profound impact on a patient's mobility and behavior. For patients with a baseline high level of activity, medical disability can cause a dramatic swing in their activity level that results in significant physiologic and psychologic changes. This complete history may be relevant to care but may not be captured using a cross-sectional biomarker.


The complex role of medical health in depression remains a major hurdle for the development of biomarkers to personalize depression management and prevention. In addition to directly impacting brain physiology, it is increasingly clear that there are systemic contributions to MDD. The immense complexity of these contributions presents many challenges, but also exciting opportunities for biomarker discovery. In the search for clinically deployable biomarkers, peripheral markers are simply more accessible than brained-based biomarkers and most readily available for integration into clinical care. To date, most of the work in this field is limited to cross-sectional assays of individual markers that do not capture the full patient medical and psychiatric history. Longitudinal biomarker studies that comprehensively capture the diversity of both depressive symptoms and medical illness, such as Texas Resilience Against Depression,33 are needed.


  1. Gadad BS, Jha MK, Czysz A, et al. Peripheral biomarkers of major depression and antidepressant treatment response: current knowledge and future outlooks. J Affect Disord. 2018;233:3–14. doi:10.1016/j.jad.2017.07.001 [CrossRef]. PMID:28709695
  2. Patten SB. Long-term medical conditions and major depression in a Canadian population study at waves 1 and 2. J Affect Disord. 2001;63(1–3):35–41. doi:10.1016/S0165-0327(00)00186-5 [CrossRef] PMID:11246078
  3. Katon WJ. Clinical and health services relationships between major depression, depressive symptoms, and general medical illness. Biol Psychiatry. 2003;54(3):216–226. doi:10.1016/S0006-3223(03)00273-7 [CrossRef] PMID:12893098
  4. Luppino FS, de Wit LM, Bouvy PF, et al. Overweight, obesity, and depression: a systematic review and meta-analysis of longitudinal studies. Arch Gen Psychiatry. 2010;67(3):220–229. doi:10.1001/archgenpsychiatry.2010.2 [CrossRef] PMID:20194822
  5. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Arlington, VA: American Psychiatric Publishing; 2013.
  6. Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH. The Inventory of Depressive Symptomatology (IDS): psychometric properties. Psychol Med. 1996;26(3):477–486. doi:10.1017/S0033291700035558 [CrossRef] PMID:8733206
  7. Cosci F, Fava GA, Sonino N. Mood and anxiety disorders as early manifestations of medical illness: a systematic review. Psychother Psychosom. 2015;84(1):22–29. doi:10.1159/000367913 [CrossRef] PMID:25547421
  8. Loftis JM, Hauser P. The phenomenology and treatment of interferon-induced depression. J Affect Disord. 2004;82(2):175–190. doi:10.1016/j.jad.2004.04.002 [CrossRef]
  9. Brown ES, Varghese FP, McEwen BS. Association of depression with medical illness: does cortisol play a role?Biol Psychiatry. 2004;55(1):1–9. doi:10.1016/S0006-3223(03)00473-6 [CrossRef] PMID:14706419
  10. Buse JB, Caprio S, Cefalu WT, et al. How do we define cure of diabetes?Diabetes Care. 2009;32(11):2133–2135. doi:10.2337/dc09-9036 [CrossRef] PMID:19875608
  11. Whyte EM, Mulsant BH. Post stroke depression: epidemiology, pathophysiology, and biological treatment. Biol Psychiatry. 2002;52(3):253–264. doi:10.1016/S0006-3223(02)01424-5 [CrossRef] PMID:12182931
  12. Angst J. Clinical course of affective disorders. In: Helgason T, Daly RJ, eds. Depressive Illness. New York, NY: Springer; 1998:1–48.
  13. Chakravarti A. To a future of genetic medicine. Nature. 2001;409(6822):822–823. doi:10.1038/35057281 [CrossRef] PMID:11236997
  14. Rosenblat JD, Lee Y, McIntyre RS. Does pharmacogenomic testing improve clinical outcomes for major depressive disorder? a systematic review of clinical trials and cost-effectiveness studies. J Clin Psychiatry. 2017;78(6):720–729. doi:10.4088/JCP.15r10583 [CrossRef] PMID:28068459
  15. Lubke GH, Hottenga JJ, Walters R, et al. Estimating the genetic variance of major depressive disorder due to all single nucleotide polymorphisms. Biol Psychiatry. 2012;72(8):707–709. doi:10.1016/j.biopsych.2012.03.011 [CrossRef] PMID:22520966
  16. McCaffery JM, Papandonatos GD, Faulconbridge LF, et al. Look AHEAD Research Group. Genetic predictors of depressive symptoms in the Look AHEAD trial. Psychosom Med. 2015;77(9):982–992. doi:10.1097/PSY.0000000000000242 [CrossRef] PMID:26489030
  17. Toups MS, Trivedi MH. Role of metabolic dysfunction in treatment resistance of major depressive disorder. Neuropsychiatry. 2011;1(5):441–455. Accessed May 14, 2020.
  18. Dinoff A, Herrmann N, Lanctôt KL. Ceramides and depression: a systematic review. J Affect Disord. 2017;213:35–43. doi:10.1016/j.jad.2017.02.008 [CrossRef] PMID:28189963
  19. Czysz AH, South C, Gadad BS, et al. Can targeted metabolomics predict depression recovery? Results from the CO-MED trial. Transl Psychiatry. 2019;9(1):11. doi:10.1038/s41398-018-0349-6 [CrossRef] PMID:30664617
  20. Disner SG, Beevers CG, Haigh EA, Beck AT. Neural mechanisms of the cognitive model of depression. Nat Rev Neurosci. 2011;12(8):467–477. doi:10.1038/nrn3027 [CrossRef] PMID:21731066
  21. Kamijo K, Pontifex MB, Khan NA, et al. The negative association of childhood obesity to cognitive control of action monitoring. Cereb Cortex. 2014;24(3):654–662. doi:10.1093/cercor/bhs349 [CrossRef] PMID:23146965
  22. Huang T, Tarp J, Domazet SL, et al. Associations of adiposity and aerobic fitness with executive function and math performance in Danish adolescents. J Pediatr. 2015;167(4):810–815. doi:10.1016/j.jpeds.2015.07.009 [CrossRef] PMID:26256018
  23. Nemeroff CB, Musselman DL. Are platelets the link between depression and ischemic heart disease?Am Heart J. 2000;140(4)(suppl):57–62. doi:10.1067/mhj.2000.109978 [CrossRef] PMID:11011349
  24. Laghrissi-Thode F, Wagner WR, Pollock BG, Johnson PC, Finkel MS. Elevated platelet factor 4 and beta-thromboglobulin plasma levels in depressed patients with ischemic heart disease. Biol Psychiatry. 1997;42(4):290–295. doi:10.1016/S0006-3223(96)00345-9 [CrossRef] PMID:9270907
  25. Nakatani D, Sato H, Sakata Y, et al. Osaka Acute Coronary Insufficiency Study Group. Influence of serotonin transporter gene polymorphism on depressive symptoms and new cardiac events after acute myocardial infarction. Am Heart J. 2005;150(4):652–658. doi:10.1016/j.ahj.2005.03.062 [CrossRef] PMID:16209960
  26. Mayberg HS, Robinson RG, Wong DF, et al. PET imaging of cortical S2 serotonin receptors after stroke: lateralized changes and relationship to depression. Am J Psychiatry. 1988;145(8):937–943. doi:10.1176/ajp.145.8.937 [CrossRef] PMID:3394877
  27. Lener MS, Niciu MJ, Ballard ED, et al. Glutamate and gamma-aminobutyric acid systems in the pathophysiology of major depression and antidepressant response to ketamine. Biol Psychiatry. 2017;81(10):886–897. doi:10.1016/j.biopsych.2016.05.005 [CrossRef]
  28. Cheng S-Y, Zhao Y-D, Li J, Chen X-Y, Wang R-D, Zeng J-W. Plasma levels of glutamate during stroke is associated with development of post-stroke depression. Psychoneuroendocrinology. 2014;47:126–135. doi:10.1016/j.psyneuen.2014.05.006 [CrossRef] PMID:25001962
  29. Spalletta G, Bossù P, Ciaramella A, Bria P, Caltagirone C, Robinson RG. The etiology of poststroke depression: a review of the literature and a new hypothesis involving inflammatory cytokines. Mol Psychiatry. 2006;11(11):984–991. doi:10.1038/ [CrossRef] PMID:16894392
  30. Colasanti A, Guo Q, Giannetti P, et al. Hippocampal neuroinflammation, functional connectivity, and depressive symptoms in multiple sclerosis. Biol Psychiatry. 2016;80(1):62–72. doi:10.1016/j.biopsych.2015.11.022 [CrossRef] PMID:26809249
  31. Mason BL. Mapping immune system dysfunction to provide clinically actionable biomarkers and to understand psychiatric pathology. Psychiatr Ann. 2020;50(6):244–249. doi:10.3928/00485713-20200506-01 [CrossRef]
  32. Polonsky WH, Fisher L, Earles J, et al. Assessing psychosocial distress in diabetes: development of the diabetes distress scale. Diabetes Care. 2005;28(3):626–631. doi:10.2337/diacare.28.3.626 [CrossRef] PMID:15735199
  33. Trivedi MH, Chin Fatt CR, Jha MK, et al. Comprehensive phenotyping of depression disease trajectory and risk: Rationale and design of Texas Resilience Against Depression study (T-RAD). J Psychiatr Res. 2020;122:22–32. doi:10.1016/j.jpsychires.2019.12.004 [CrossRef] PMID:31918350

Andrew Czysz, MD, PhD, is an Assistant Professor of Psychiatry, Center for Depression Research and Clinical Care, Department of Psychiatry, University of Texas Southwestern Medical Center.

Address correspondence to Andrew Czysz, MD, PhD, Department of Psychiatry, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390-9119; email:

Disclosure: Andrew Czysz discloses contracted research support from Janssen Research and Development, LLC.


Sign up to receive

Journal E-contents