Peripheral Endothelial Dysfunction Is Associated With Incident Major Depressive Disorder Authors: Jaskanwal D S Sara, Nazanin Rajai, Scott Breitinger, Betsy Medina‐Inojosa, Lilach O Lerman, Francisco Lopez‐Jimenez, Amir Lerman Published: 2024 ABSTRACT AbstractBackgroundA subset of individuals with major depressive disorder (MDD) have a high burden of cardiovascular risk factors and cerebral small‐vessel disease, implicating vascular disease in the development of depression. Cross‐sectional studies demonstrate a link between endothelial dysfunction and MDD, but the prospective association between peripheral endothelial dysfunction (PED) and an incident diagnosis of MDD is unknown.Methods and ResultsPatients undergoing a baseline assessment of cardiovascular risk were evaluated for PED using reactive hyperemia–peripheral arterial tonometry (≤1.8 consistent with PED). Patient medical records were reviewed to identify those who underwent a formal clinical evaluation of MDD after the index PED evaluation. The frequency of PED was compared in those with and without MDD. Logistic regression analyses were performed to assess the association between baseline PED and incident MDD. Between January 2006 and December 2020, 1614 patients underwent testing for PED. Four hundred eighty‐four (30.1%) patients underwent a formal evaluation for MDD after (0–15 years) the index procedure (mean±SD age, 52.8±13.8 years; 65.2% women). Of these, 157 (32.4%) had PED and 108 (31.0%) were diagnosed with MDD. Individuals with MDD had a higher frequency of PED (40.2% versus 30.2%; P=0.034) compared with those without MDD. In multivariable analyses, PED was significantly associated with MDD (odds ratio, 2.3 [95% CI, 1.4–3.8]; P<0.001).ConclusionsPED is significantly associated with incident MDD. Thus, PED may be a useful marker to identify individuals at increased risk of depression who may benefit from more frequent and earlier management strategies.Keywords: chest pain, depression, endothelial dysfunction, ischemiaSubject Categories: Biomarkers, Vascular Biology Nonstandard Abbreviations and Acronyms major depressive disorder peripheral endothelial dysfunction Patient Health Questionnaire‐9 reactive hyperemic index reactive hyperemia–peripheral arterial tonometry Clinical Perspective What Is New? Peripheral endothelial dysfunction, measured using reactive hyperemia peripheral arterial tonometry, is significantly associated with incident major depressive disorder at follow‐up in patients undergoing a baseline evaluation of cardiovascular risk. Peripheral endothelial dysfunction, measured using reactive hyperemia peripheral arterial tonometry, is significantly associated with incident major depressive disorder at follow‐up in patients undergoing a baseline evaluation of cardiovascular risk. What Are the Clinical Implications? Peripheral endothelial dysfunction may be a marker that could be used to identify individuals at increased risk of depression who may benefit from more frequent screening and earlier implementation of preventive and therapeutic measures. Peripheral endothelial dysfunction may be a marker that could be used to identify individuals at increased risk of depression who may benefit from more frequent screening and earlier implementation of preventive and therapeutic measures. Major depressive disorder (MDD) has a prevalence of 5% to 10% and is estimated to affect >300 million people worldwide. 1 MDD is characterized by low mood or anhedonia that is unremitting for a period of time >2 weeks combined with psychomotor, somatic, and cognitive changes that result in some degree of distress or functional impairment. 2 As such, MDD was estimated to be the third leading cause of years lived with disability between 2007 and 2017 3 and the second leading cause of disability‐adjusted life years by 2030. 4 The cause of MDD is complex and is hypothesized to be a diagnostic category composed of multiple causative subtypes with yet to be elucidated underlying pathologic mechanisms. 5 , 6 As such, individuals with this disorder often exhibit varying disease courses and unpredictable responses to therapy. 7 , 8 Thus, improving our understanding of the causes of MDD and developing effective preventive measures is an important public health need. MDD and cardiovascular disease (CVD) have a bidirectional relationship. 9 MDD is associated with an increased risk of coronary artery disease, myocardial infarction, 10 and cardiovascular mortality. 11 Conversely, the prevalence of MDD in individuals with established CVD is 3‐fold greater than in the general population. 12 Depression diagnosed later in life is believed to be pathophysiologically distinct from early‐onset depression and is associated with a greater burden of cardiovascular risk factors, 13 a stronger association with stroke, vascular dementia and white matter hyperintensities, 14 and higher resistance to conventional treatment strategies, such as cognitive‐behavioral therapy and antidepressants. 15 This observation has given rise to the notion that vascular disease may play a role in the cause of depression in a subset of patients along the lines of the “vascular depression” hypothesis. 16 Endothelial dysfunction, a marker of early atherosclerosis associated with a several‐fold increased risk of CVD 17 that can be measured noninvasively and peripherally with techniques such as reactive hyperemia–peripheral arterial tonometry (RH‐PAT), is believed to play a pathophysiological role in the link between CVD and MDD. 18 Cross‐sectional studies have demonstrated an association between direct measures of peripheral endothelial dysfunction (PED) 19 , 20 , 21 and MDD. However, prospective studies have focused on surrogate measures of PED, such as plasma endothelial biomarkers, or have relied on indirect markers of cerebrovascular endothelial function, such as white matter hyperintensities and cerebral microbleeds or microinfarcts that can only be identified using time‐consuming and expensive imaging protocols. 22 , 23 In the current study, we aimed to evaluate the association between a simple, noninvasive, and reliable method of directly evaluating peripheral microvascular endothelial function using RH‐PAT 24 and an incident diagnosis of MDD. We hypothesize that PED at baseline is associated with a diagnosis of incident MDD at follow‐up. METHODS This prospective cohort study was approved by the Mayo Clinic Institutional Review Board, and all study participants provided their informed consent. This study is reported in keeping with the strengthening the reporting of observational studies in epidemiology (STROBE) cohort study reporting guidelines. The data that support the findings of this study are available from the corresponding author upon reasonable request. Participants were referred to the Center for Coronary Physiology at our institution between 2006 and 2020 for an assessment of peripheral endothelial function using RH‐PAT testing. The test was performed as part of a baseline assessment of cardiovascular risk or in volunteers enrolling in clinical trials. Participants with inconclusive test results and those who were not evaluated for MDD at follow‐up were excluded from this study. Assessment of Peripheral Endothelial Function Peripheral microvascular endothelial function was assessed using reactive hyperemia by means of digital plethysmography (EndoPAT2000 device, Itamar Medical Inc, Caesarea, Israel), a validated method of measuring peripheral endothelial function. 24 A detailed study protocol was described elsewhere. 25 , 26 , 27 , 28 In brief, all patients were instructed to stop all vasoactive medications, including calcium channel blockers, β‐blockers, and long‐acting nitrates, for at least 24 hours before the RH‐PAT test. At the beginning of the test, a finger probe was placed on the middle fingers of both hands. The 15‐minute reactive hyperemia protocol then commenced as follows: first, an initial 5‐minute period consisted of equilibration and a baseline measurement of PAT; next, a blood pressure arm cuff applied to the test arm was inflated to 60 mm Hg above baseline systolic blood pressure, or at least 200 mm Hg for 5 minutes; the cuff was then deflated, and the postdeflation PAT tracing was recorded for an additional 5 minutes. The reactive hyperemia index (RHI) ratio was calculated by dividing the average pulse wave amplitude after deflation (postdeflation PAT) by the baseline pulse‐wave amplitude (baseline PAT) in the same arm. This was computed automatically by the EndoPAT device by normalizing this ratio to baseline and indexing it to the control arm. An RHI ≤1.8 was deemed consistent with PED, as demonstrated previously. 29 Patient Information Data were collected on conventional cardiovascular risk factors, including hypertension, diabetes, hyperlipidemia, smoking status, and body mass index, and biochemical parameters, such as total and high‐density lipoprotein cholesterol. Smoking was categorized as a history of current smoking, former smoking, or never smoking; hyperlipidemia was defined by a documented history of hyperlipidemia, treatment with lipid‐lowering therapy, a low‐density lipoprotein cholesterol value above target (<130 mg/dL for low‐risk patients, <100 mg/dL for moderate‐ to high‐risk patients, <70 mg/dL for very high‐risk patients, and <55 mg/dL for extremely high‐risk patients based on 10‐year atherosclerotic CVD risk), high‐density lipoprotein cholesterol <40 mg/dL in men or <50 mg/dL in women, or triglycerides >150 mg/dL. Type 2 diabetes was defined as a documented history of or treatment for type 2 diabetes, or a hemoglobin A1c of >6.5, if available. Hypertension was defined as a documented history of or treatment of the condition, a systolic blood pressure measurement >130 mm Hg, or diastolic blood pressure measurement >90 mm Hg. All blood test results included in this study are based on blood samples obtained on the morning of the index procedure. A history of myocardial infarction was diagnosed as the presence of at least 2 of the following: (1) typical chest pain for at least 20 minutes; (2) increased creatinine kinase (or the MB fraction) or troponin level; or (3) new ST‐segment elevation, Q‐waves, or left bundle‐branch block on ECG. Information was also collected on medical history, including other vascular diseases (defined as a documented history of peripheral vascular disease, stroke, or transient ischemic attack), and documented history of migraine headache. 30 , 31 , 32 , 33 , 34 Patients with a baseline diagnosis of MDD were excluded from this study. To assess their temporal relationships, we retrospectively evaluated the medical records of all remaining patients to identify those who were referred for and underwent a clinically indicated formal evaluation for a diagnosis of MDD at a time after the index date of RH‐PAT testing. In cases with >1 formal evaluation for MDD, the first evaluation was used. The diagnosis of MDD was determined by means of a structured clinical interview by a trained clinician. We did not base the diagnosis of MDD on the results of Patient Health Questionnaire‐9 (PHQ‐9) when this was available as this is used to screen and to evaluate symptoms of depression rather than to diagnose MDD per se. 35 Statistical Analysis Patients were categorized as having PED versus normal peripheral endothelial function at baseline, using the cutoff value of ≤1.8 to define PED. Patients were then prospectively identified as developing an incident diagnosis of MDD. The proportion of individuals with PED who developed MDD was compared with those who did not develop MDD in all patients and across both sexes. Furthermore, RH‐PAT, as a continuous variable and index of peripheral microvascular endothelial function, was compared between patients who developed MDD versus those who did not in all patients and after stratifying by sex. Logistic regression analyses were used to estimate the odds ratio (OR) of developing incident MDD using PED as the categorical independent variable in both univariable and multivariable analyses. Continuous variables are presented as a mean (SD), where data are normally distributed and as a median (quartile 1–quartile 3) for skewed data. Categorical variables are presented as frequencies (percentages). Differences between groups were analyzed using Student t test and Wilcoxon rank‐sum test for continuous variables and Pearson χ2 test for proportions. P<0.05 was considered significant, and all statistical analyses were performed using JMP 9 software (SAS Institute, Inc, Cary, NC). RESULTS Sample Overview Between January 2006 and December 2020, 1614 patients underwent testing using RH‐PAT to evaluate peripheral endothelial function. Four hundred eighty‐four (30.1%) were referred for and underwent a formal evaluation for MDD at a date after the index test procedure (range, 0–15 years) and had complete measurements available for RH‐PAT and baseline CVD variables (mean±SD age, 52.8±13.8 years; 65.2% women; mean±SD body mass index, 29±6.7 kg/m2), of whom 157 (32.4%) had PED. Among female participants, 83 (26.3%) had PED, whereas in men, 74 (43.8%) had PED at baseline. Table 1 summarizes the baseline characteristics of all patients categorized according to those with versus without PED. Individuals with PED comprised a lower proportion of women (53.2% versus 70.9%; P<0.001), had a higher frequency of diabetes (16.5% versus 10.3%; P=0.03), had lower total cholesterol values (174.2±38.5 versus 184.1±42.6 mg/dL; P=0.016), and had lower high‐density lipoprotein cholesterol values (53.5±17.4 versus 60.3±17.8 mg/dL; P<0.001) compared with individuals without PED. Otherwise, there were no other significant differences between groups with respect to demographic or clinical variables. Table 1Baseline Characteristics of Participants With Versus Without PEDCharacteristicTotalPEDNo PED P value(N=484)(N=157 [32.4%])(N=327 [67.5%])Female sex, %65.253.270.9<0.001Age, y52.8±13.852.09±14.853.14±13.30.45BMI, kg/m2 29±6.729.4±5.728.9±7.050.6Systolic blood pressure, mm Hg125.1±16.4125±17.1124±14.80.3Diastolic blood pressure, mm Hg76±12.376±13.575.9±90.47Smoking, %27.730.426.40.2Hypertension, %44.74345.50.3Diabetes, %12.316.510.30.03Hyperlipidemia, %27.328.526.70.37Myocardial infarction, %9.28.99.40.5Cerebrovascular disease, %6.65.17.30.31Heart failure, %6.67.66.10.32Migraine headache, %35.333.836.10.34Peripheral artery disease, %8.58.38.60.53Total cholesterol, mg/dL181.03±41.5174.2±38.5184.1±42.60.016HDL cholesterol, mg/dL58.1±17.953.5±17.460.3±17.8<0.001Data are given as mean±SD unless otherwise indicated. BMI indicates body mass index; HDL, high‐density lipoprotein; and PED, peripheral endothelial dysfunction. Baseline Characteristics of Participants With Versus Without PED Data are given as mean±SD unless otherwise indicated. BMI indicates body mass index; HDL, high‐density lipoprotein; and PED, peripheral endothelial dysfunction. Frequency of MDD According to PED Table 2 shows a comparison of clinical variables, including PED and RHI, between patients with versus without MDD. One hundred eight (31.0%) of all individuals were diagnosed with MDD. Individuals with MDD were slightly younger (50.3±12.5 versus 53.55±14.0 years; P=0.03) and comprised a higher proportion of women (74.1% versus 62.6%; P=0.02) compared with those without MDD. Individuals with MDD also had a higher frequency of smoking (42.6% versus 23.9%; P<0.001), hypertension (60.2% versus 40.6%; P<0.001), and diabetes (22.2% versus 9.7%; P<0.001) compared with those without MDD. Furthermore, individuals with MDD had a higher frequency of PED defined as an RHI ≤1.8 (40.2% versus 30.2%; P=0.034) compared with those without MDD, and had on average higher PHQ‐9 scores (mean [range]) at the time of evaluation for MDD (7 [3–12] versus 4 [1–10]; P<0.019). In men, the frequency of PED in individuals with MDD was not significantly different compared with that in individuals without MDD (53.6% versus 41.8%; P=0.17). The PHQ‐9 score between men with and without MDD was not significantly different at the time of evaluation for MDD (6 [2–11] versus 3 [1–8]; P=0.13). In women, those with MDD had a higher frequency of PED compared with those without MDD (35.4% versus 23.3%; P=0.02). The average PHQ‐9 score in women with MDD was greater than that in those without MDD (6 [2–11] versus 3 [1–8]; P=0.12). When comparing RHI as a continuous variable between individuals with MDD versus those without MDD, there were no significant differences between groups in all patients (2.02 [1.68–2.6] versus 2.22 [1.67–2.71]; P=0.2), in women only (2.11 [1.74–2.66] versus 2.34 [1.85–2.79]; P=0.06), or in men only (1.79 [1.59–2.34] versus 1.91 [1.59–2.58]; P=0.8). Table 2Comparison of Clinical Variables Between Patients With Versus Without MDDVariableMDDNo MDD P value(N=108 [22.3%])(N=376 [77.7%])Demographics108 (of 486)378Age, y50.3±12.553.55±140.03Sex (female), %74.162.60.02Medical comorbidities, %Smoking42.623.85<0.001Hypertension60.240.6<0.001Diabetes22.29.7<0.001Hyperlipidemia30.626.40.23Myocardial infarction7.49.70.3Heart failure7.46.30.4Cerebrovascular diseases6.56.50.6Peripheral endothelial functionRHI2.02 (1.68–2.6)2.22 (1.67–2.71)0.2Depression scorePHQ‐9 score7 (3–12)4 (1–10)<0.019Data are given as mean±SD or median (range) unless otherwise indicated. MDD indicates major depressive disorder; PHQ‐9, Patient Health Questionnaire‐9; and RHI, reactive hyperemia index. Comparison of Clinical Variables Between Patients With Versus Without MDD Data are given as mean±SD or median (range) unless otherwise indicated. MDD indicates major depressive disorder; PHQ‐9, Patient Health Questionnaire‐9; and RHI, reactive hyperemia index. PED and MDD: Univariable and Multivariable Analyses Table 3 demonstrates the associations between PED type (as a categorical variable) and MDD in univariable analyses and multivariable analyses adjusted for age, sex, cerebrovascular disease, heart failure, diabetes, hypertension, hyperlipidemia, history of myocardial infarction, and smoking. In univariable analyses, PED was associated with MDD with borderline significance (OR, 1.55 [95% CI, 0.99–2.4]; P=0.051). Female sex, age, diabetes, hypertension, and history of smoking were all significantly associated with MDD in univariable analyses. In multivariable analyses, PED was significantly associated with MDD (OR, 2.3 [95% CI, 1.4–3.8]; P<0.001), as were female sex, age, diabetes, hypertension, and history of smoking. Table 3Univariable and Multivariable Logistic Regression Analyses Evaluating the Association Between Risk Factors and MDDVariableUnivariableMultivariableOR (95%CI) P valueOR (95% CI) P valueSex (reference: male)1.67 (1.04–2.7)0.032.1 (1.3–3.6)0.004Age0.98 (96–0.99)0.030.96 (0.94–0.98)<0.001Cerebrovascular disease0.99 (0.41–2.37)0.990.99 (0.39–2.5)0.99Heart failure1.2 (0.5–2.7)0.660.89 (0.35–2.2)0.8Diabetes2.5 (1.4–4.5)0.0012.2 (1.2–4.2)0.012Hypertension2.2 (1.4–3.4)<0.0012.5 (1.5–4.2)<0.001Hyperlipidemia1.18 (0.7–1.89)0.481.3 (0.7–2.3)0.39Myocardial infarction0.75 (0.34–1.66)0.480.84 (0.33–2.1)0.72Smoking2.34 (1.5–3.68)<0.0011.7 (1.04–2.7)0.034PED1.55 (0.99–2.4)0.0512.3 (1.4–3.8)<0.001MDD indicates major depressive disorder; OR, odds ratio; and PED, peripheral endothelial dysfunction. Univariable and Multivariable Logistic Regression Analyses Evaluating the Association Between Risk Factors and MDD MDD indicates major depressive disorder; OR, odds ratio; and PED, peripheral endothelial dysfunction. DISCUSSION In the current study, we demonstrated the following: (1) of consecutive patients who presented to our institution for baseline assessment of CVD risk, almost one‐third of patients were subsequently referred for and underwent an evaluation of MDD; (2) of those referred for an evaluation of MDD, almost one‐third were diagnosed with MDD; (3) PED was prevalent in those with MDD, affecting >40% of all individuals; (4) the frequency of PED was significantly higher in women with MDD compared with women without MDD, although this difference was not observed in men; and (5) PED was associated with incident MDD in univariable (albeit with borderline significance) and multivariable analyses. In keeping with previous data, our findings highlight the fact that MDD is a prevalent disorder. 1 Together with its high burden of disability, 3 , 4 MDD represents an important public health problem for which a better understanding of the underlying pathology would be of great value. The current study supports a potential role of endothelial dysfunction in the development of depression and suggests that PED could be a marker of risk for the development of a subset of MDD even after adjusting for other cardiovascular risk factors. Given that effective management of MDD is often hampered by patients having unpredictable disease courses and variable responses to treatment, 7 , 8 harnessing a unique therapeutic target, such as endothelial dysfunction, offers the potential for novel ways in which to manage and improve clinical outcomes in certain depression subtypes. MDD is more prevalent in patients with acute coronary syndromes and chronic CVD compared with the general population. 12 , 36 At the same time, individuals with MDD are at an increased risk of coronary artery disease and myocardial infarction, 10 as well as a premature mortality from CVD. 11 These observations suggest that MDD and CVD may contribute to each other's pathophysiology in a reciprocal manner. 9 Endothelial dysfunction, a marker of preclinical atherosclerosis characterized by impaired endothelium‐dependent vasodilatation 18 that is associated with an increased risk of CVD, 17 has been shown to be associated with MDD in several meta‐analyses. 19 , 20 , 21 The studies included in these reports evaluated endothelial dysfunction in a variety of ways. Some relied on peripherally circulating markers of inflammation and endothelial function 8 that provide surrogate measures of vascular health rather than a functional assessment of the endothelium that specifically relates to the effects of nitric oxide. This molecule has vasodilator properties and inhibits several proatherogenic processes, such as the oxidation of low‐density lipoproteins and the proliferation of smooth muscle cells, 37 making it a key marker of endothelial function. Other studies made use of techniques such as flow‐mediated dilatation, 19 which noninvasively assesses vasodilator response in response to increased shear stress 18 using ultrasound to compare brachial artery diameter before and after reactive hyperemia. Results derived using this technique are operator dependent, in contrast to RH‐PAT used in the current study, which assesses endothelial dysfunction in a computerized operator‐independent manner that normalizes endothelial function to the contralateral arm. 24 Furthermore, RH‐PAT assesses endothelial function in the microcirculation, as opposed to large conductance vessels, which may be more relevant to the intracranial vasculature and neuropsychiatric diseases, such as MDD. 22 Most studies that have assessed the relationship between endothelial function and MDD have also had a cross‐sectional design, 19 , 20 , 21 making it difficult to infer a causal relationship between endothelial dysfunction and MDD in the direction of that association. Indeed, these reports have often postulated that MDD results in a higher incidence of CVD in a manner that is mediated by endothelial dysfunction, 38 which itself arises in the context of inflammation, maladaptive lifestyle habits, and hypothalamic‐pituitary‐adrenal and autonomic dysregulation. 38 However, several prospective studies have demonstrated that endothelial dysfunction is in fact associated with a future diagnosis of MDD. 22 These studies imply that cerebrovascular disease, which may even be subclinical, plays a role in the cause of depression in accordance with the “vascular depression” hypothesis. 16 Indeed, depression diagnosed later in life is associated with stroke, vascular dementia, and white matter hyperintensities. 14 Furthermore, late‐onset depression is thought to be pathophysiologically distinct to early‐onset depression, with the latter exhibiting psychobiological vulnerability, such as neuroticism and parental history, whereas the former has a greater preponderance of vascular risk factors 13 and is more resistant to cognitive‐behavioral therapy and antidepressants. 15 Prospective studies identified white matter hyperintensities 22 , 39 or subcortical microbleeds or microinfarcts on brain imaging 40 as proxies of cerebrovascular endothelial dysfunction. Another study assessed cerebrovascular endothelial function more directly using transcranial Doppler ultrasound to measure blood flow velocities. 41 These techniques rely on expensive and complicated imaging protocols that can be cumbersome for patients. In the current study, and to our knowledge for the first time, we demonstrate that individuals with PED at baseline have a greater odds of incident MDD at follow‐up. Our study used RHI to directly measure peripheral endothelial function in a nonintrusive expedient manner that could be performed in an office‐based setting. Thus, our findings underscore the notion that endothelial dysfunction is a systemic process that can affect and be measured in different vascular beds and still be associated with disease in distant organ systems. In this way, endothelial dysfunction measured peripherally may be a marker of risk for future depression. Further larger prospective studies will be required to confirm these findings and to determine the clinical implications of identifying individuals at risk of MDD using PED. Measures of endothelial dysfunction provide an integrated index of the cumulative effects of risk (and protective) factors on the vascular wall across time. 42 In fact, population studies have demonstrated that endothelial function, measured using noninvasive technique, such as RHI, may be a superior measure of cardiovascular risk compared with the sum of the estimated effect of traditional risk factors, which do not always accurately predict CVD. 43 , 44 Thus, endothelial dysfunction in particular, and vascular health more generally, may also be a predictor of depression, or a hypothetical causative subtype of depression, a disease not traditionally included under the umbrella of vascular disease. Our multivariable analyses demonstrated a significant association between PED and MDD even after adjusting for traditional cardiovascular risk factors as well as cerebrovascular disease, suggesting that the presence of these variables cannot explain this association. This suggests that endothelial dysfunction may play an independent role in the mechanism of at least a subset of depression. For example, cerebral endothelial dysfunction may result in chronic ischemia and cerebral perfusion deficits to regions involved in mood regulation. 45 Endothelial dysfunction may also play a role in the cause of depression through concurrent low‐grade inflammation and oxidative stress that have also been associated with depression, 46 , 47 or by disrupting the blood‐brain barrier, 22 thereby depriving the brain of nutrients and exposing it to toxins. Further studies are required to clarify these pathophysiological mechanisms. In the current study, we used the formal diagnosis of MDD rather than depressive symptoms as the outcome of interest. Indeed, symptoms may fluctuate throughout the course of disease even when an individual fulfills the diagnostic criteria for MDD. 35 Interestingly, in the current study, MDD, but not PED, was associated with higher self‐reported PHQ‐9 scores. The link between depressive symptoms and endothelial dysfunction has been less consistent than that with a diagnosis of MDD per se. 48 Although symptoms of depression are common and distressing in and of themselves, the prevention and treatment of MDD represents a more pressing public health need. 3 , 4 The findings of the current study suggest that endothelial dysfunction may form a useful therapeutic target when managing MDD if, for instance, treatment strategies designed to effect endothelial function to reduce cardiovascular risk may instead be deployed to reduce the risk of or potentially even treat MDD. Such approaches shown to improve endothelial dysfunction include exercise programs, 49 dietary modification, 50 and even statin therapy. 51 Interestingly, selective serotonin receptor inhibitors used to treat moderate to severe MDD favorably modulate biochemical markers of endothelial damage, such as circulating endothelial cells and endothelial progenitor cells as well as plasma levels of soluble von Willebrand factor in patients with MDD. 52 Thus, it is conceivable that these medications could exert some of their therapeutic antidepressant effects by improving endothelial function, potentially paving the way for earlier and broader indications for this pharmacotherapy. Further studies will be required to further evaluate the implications of this hypothesis, and to determine the effects of both pharmacologic and nonpharmacologic interventions on endothelial dysfunction in reducing MDD risk. The current study has some limitations. First, our sample is made up of patients who were referred to a tertiary referral center and thus constitute a select population. Second, individuals who were referred for and underwent an evaluation of MDD after the index RH‐PAT test also constituted a select group as their referral was clinically indicated. Furthermore, although each of these patients underwent an independent structured clinical interview by a clinician, the results of that process were ascertained for this study by medical record review rather than being adjudicated by the investigators of our study. Also, we cannot account for the latent time that may be present between the onset of diagnostic features and the time of an actual clinical diagnosis. Although included study participants did not carry a formal diagnosis of MDD at the time of their baseline evaluation for PED, we cannot exclude the possibility that a subset of patients did not already have the features of depression at that time that were either subclinical, but still detrimental to the vascular health of the affected individual, or simply not diagnosed at that time. Under such circumstances, PED would be a characteristic, as opposed to a predictor, of MDD. Thus, prospective studies that also formally assess for MDD at baseline in parallel with an assessment of baseline endothelial function will be required. Last, we did not evaluate psychosocial and other interpersonal variables at baseline that may have also played a role in the cause of MDD. CONCLUSIONS PED is significantly associated with incident MDD in patients undergoing a baseline evaluation of cardiovascular risk. Thus, PED may be a marker that could be used to identify individuals at increased risk of depression who may benefit from more frequent screening and earlier implementation of preventive and therapeutic measures. Furthermore, endothelial dysfunction may represent a small but significant contributary mechanism for MDD, potentially implicating the role of vascular disease in the cause of this condition. Sources of Funding None. Disclosures None. This manuscript was sent to William W. Aitken, MD, Assistant Editor, for review by expert referees, editorial decision, and final disposition. For Sources of Funding and Disclosures, see page 7. References