Association of treatment response with obesity and other metabolic risk factors in adults with depressive disorders: Results from a National Depression Cohort study in Korea (the CRESCEND study)

Changes in insulin resistance following antidepressant treatment mediate response in major depressive disorder

BACKGROUND

Insulin resistance (IR) is a potential predictor of antidepressant treatment response.

AIMS

We assess changes in IR after antidepressant treatment and whether these changes have any effect on treatment response. Also, to see whether changes in IR mediates relationship between C-reactive protein (CRP) and antidepressant efficacy.

METHODS

This is a secondary analysis of an 8-week, open-label clinical trial with 95 adults experiencing a major depressive episode. Response to vortioxetine was measured using the Montgomery-Åsberg Depression Rating Scale (MADRS). Generalized estimating equation models were utilized for this intent-to-treat analysis.

RESULTS

When adjusted for age, sex, and body mass index, there was a significant increase in IR following treatment in the overall sample (p = 0.035). This finding was detected in treatment non-responders (p = 0.019), whereas it was not observed in responders (p = 0.329). Mediation analysis revealed that change in IR during treatment was responsible for change in MADRS as well as the relationship between baseline CRP and treatment response.

CONCLUSIONS

Exacerbation of IR during antidepressant treatment mediated non-response. Conversely in treatment responders IR reduced. Like previous studies, baseline CRP moderated treatment response. This relationship was also mediated by changes in IR. These findings further elucidate the role of IR in terms of antidepressant response as well as potentially explain inflammation's relationship with the latter.

Overweight in mood disorders: Effects on morbidity and treatment response

OBJECTIVE

As it is not clear how body-mass index (BMI) may relate to diagnosis, symptom-severity, illness-course, and treatment-response among psychiatric patients, we related BMI to psychiatric diagnosis and to selected clinical and demographic factors in major affective disorder subjects.

METHODS

We analyzed mean BMI levels vs. diagnosis, and evaluated selected risk factors for association with overweight and obesity among subjects with DSM-5 major affective disorders.

RESULTS

In 1884 subjects, BMI ranged from 23.4 kg/m² with anxiety disorders to 27.6 with psychotic disorders, and averaged 24.1 among 1469 affective disorder subjects. Mood-disorder subjects with BMI ≥ 25 (overweight/obese) were more likely: men, older, married, with more children and siblings, less education, lower socioeconomic status, engaged less in physical exercise, smoked more, and lived in less densely populated areas. They also were more likely to have: BD than MDD, familial mood disorders, no co-occurring ADHD, higher serum triglyceride levels, more time depressed and less improvement in depression ratings with treatment.

CONCLUSIONS

Risk of being overweight or obese was greatest with psychoses, least with anxiety, personality, and minor depressive disorders, and intermediate with major mood disorders. Several plausible risk factors for high BMI were identified in mood disorder subjects, including male sex and with BD > MDD. Striking were selectively greater prospective morbidity and decreased treatment-response for depression vs. mania with BMI ≥ 25.

Pre-treatment allostatic load and metabolic dysregulation predict SSRI response in major depressive disorder: a preliminary report

BACKGROUND

Major depressive disorder (MDD) is associated with increased allostatic load (AL; a measure of physiological costs of repeated/chronic stress-responding) and metabolic dysregulation (MetD; a measure of metabolic health and precursor to many medical illnesses). Though AL and MetD are associated with poor somatic health outcomes, little is known regarding their relationship with antidepressant-treatment outcomes.

METHODS

We determined pre-treatment AL and MetD in 67 healthy controls and 34 unmedicated, medically healthy MDD subjects. Following this, MDD subjects completed 8-weeks of open-label selective serotonin reuptake inhibitor (SSRI) antidepressant treatment and were categorized as 'Responders' (⩾50% improvement in depression severity ratings) or 'Non-responders' (<50% improvement). Logistic and linear regressions were performed to determine if pre-treatment AL or MetD scores predicted SSRI-response. Secondary analyses examined cross-sectional differences between MDD and control groups.

RESULTS

Pre-treatment AL and MetD scores significantly predicted continuous antidepressant response (i.e. absolute decreases in depression severity ratings) (p = 0.012 and 0.014, respectively), as well as post-treatment status as a Responder or Non-responder (p = 0.022 and 0.040, respectively), such that higher pre-treatment AL and MetD were associated with poorer SSRI-treatment outcomes. Pre-treatment AL and MetD of Responders were similar to Controls, while those of Non-responders were significantly higher than both Responders (p = 0.025 and 0.033, respectively) and Controls (p = 0.039 and 0.001, respectively).

CONCLUSIONS

These preliminary findings suggest that indices of metabolic and hypothalamic-pituitary-adrenal-axis dysregulation are associated with poorer SSRI-treatment response. To our knowledge, this is the first study to demonstrate that these markers of medical disease risk also predict poorer antidepressant outcomes.

Insulin resistance is associated with deficits in hedonic, self-reported cognitive, and psychosocial functional response to antidepressant treatment in individuals with major depressive disorder

BACKGROUND

To assess the effect of insulin resistance (IR) on treatment response to the antidepressant, vortioxetine, in patients with Major Depressive Disorder (MDD).

METHODS

This is a secondary analysis of an 8-week, open-label clinical trial. Ninety-five adults in a primary care setting experiencing a major depressive episode were included. Response to vortioxetine was measured using the THINC-integrated tool, Montgomery Åsberg Depression Rating Scale (MADRS), the Snaith-Hamilton Pleasure Scale (SHAPS), the Perceived Deficits Questionnaire (PDQ-5), and the Sheehan Disability Scale (SDS). Generalized estimating equation models were utilized for data analysis.

RESULTS

When adjusted for age, gender, dose, and BMI, there was a significant baseline IR by time interaction for SHAPS (p = 0.022), PDQ-5 (p = 0.037), and SDS (p = 0.013). Higher baseline IR predicted decreased early improvements in anhedonia. It also predicted poorer subjective assessments of cognition and increased functional impairment at the endpoint of treatment. For functional capacity (i.e. SDS) other covariates including severity of symptoms, illness course, other metabolic factors (e.g. cholesterol), and physical activity were included with no changes to the moderating effect of baseline IR.

LIMITATIONS

This was a post-hoc analysis of a primarily non-diabetic sample. Also, only one agent was assessed.

CONCLUSIONS

IR was a predictor of response to vortioxetine. This persisted after controlling for other factors including, but not limited to, BMI. These findings strengthen the link between depression and IR and may point to another novel metabolic predictor of response. These findings should be replicated using other antidepressants.

The impact of insulin on the serotonergic system and consequences on diabetes-associated mood disorders

The idea that insulin could influence emotional behaviours has long been suggested. However, the underlying mechanisms have yet to be solved and there is no direct and clear-cut evidence demonstrating that such action involves brain serotonergic neurones. Indeed, initial arguments in favour of the association between insulin, serotonin and mood arise from clinical or animal studies showing that impaired insulin action in type 1 or type 2 diabetes causes anxiety- and depressive symptoms along with blunted plasma and brain serotonin levels. The present review synthesises the main mechanistic hypotheses that might explain the comorbidity between diabetes and depression. It also provides a state of knowledge of the direct and indirect experimental evidence that insulin modulates brain serotonergic neurones. Finally, it highlights the literature suggesting that antidiabetic drugs present antidepressant-like effects and, conversely, that serotonergic antidepressants impact glucose homeostasis. Overall, this review provides mechanistic insights into how insulin signalling alters serotonergic neurotransmission and related behaviours bringing new targets for therapeutic options.

Physical health in affective disorders: a narrative review of the literature

This article reviews the most common non-psychiatric comorbidities associated with affective disorders, examining the implications of their possible bidirectional link. A narrative review was conducted on the association among the three most common non-psychiatric diseases in major depressive disorder and bipolar disorder (obesity, metabolic syndrome, and cardiovascular diseases) in articles published from January 1994 to April 2020. The evidence suggests that obesity, metabolic syndrome, and cardiovascular diseases are highly prevalent in patients diagnosed with affective disorders. The presence of non-psychiatric comorbidities significantly worsens the therapeutic management and prognosis of affective disorders and vice versa. In many cases, these comorbidities may precede the onset of affective disorders, although in most cases they appear after it. The presence of these concurrent non-psychiatric diseases in an individual diagnosed with an affective disorder is associated with a more complex disease presentation and management. For professionals, the evidence unequivocally supports routine surveillance of comorbidities from a multidisciplinary approach.

Impact of Psychotropic Medication Effects on Obesity and the Metabolic Syndrome in People With Serious Mental Illness

People with serious mental illness (SMI), including schizophrenia, bipolar disorder, and major depressive disorder, have a higher mortality rate and shortened life expectancy. This is mainly attributable to physical diseases, particularly cardiovascular diseases (CVDs). Important risk factors for CVDs are obesity and other metabolic abnormalities, which are especially prevalent in people with SMI. Several factors contribute to this increased risk, including unhealthy lifestyles. Psychotropic medication independently further increases this risk. In this review we want to examine the relationship between obesity and other components of the metabolic syndrome and psychotropic medication in people with SMI.

Excess body weight as a predictor of response to treatment with antidepressants in patients with depressive disorder

BACKGROUND

Depression and obesity are debilitating conditions representing an enormous health and economic burden worldwide. Depression is common among patients with excess weight, but more importantly, these patients may be at risk for poor response when treated with antidepressant medications (AD).

METHODS

We conducted a comprehensive scoping review to summarize the evidence regarding the difference in response to treatment of depression with AD among patients with excess weight as compared to normal weight patients and to identify knowledge gaps.

RESULTS

The search of the Medline and PsycINFO (2004-2019) identified twelve relevant studies. Tabulation and frequency analysis of the charted data along with a narrative synthesis were performed. Nine studies (75%) reported clinically relevant negative association between patients' high BMI or obesity and treatment response to either nortriptyline, fluoxetine, or various AD; one study (8.3%) reported no difference in response to various AD combinations between BMI groups. One study showed benefits of bupropion and escitalopram combination in patients with morbid obesity (BMI > 35 kg/m²) as compared with escitalopram monotherapy. Another study reported benefits when using venlafaxine-XR in patients with morbid obesity. We also acknowledge the possible role of sex and genetic factors predicting AD treatment response.

LIMITATIONS

The search was restricted to two most relevant sources, publications in four languages and adult population.

CONCLUSION

The synthesized data may be useful to physicians in their decision regarding the choice of AD in patients with excess weight. Researchers need to address causality of association between obesity and treatment response to individual AD types.

SOD2 genetic polymorphism (rs4880) has no impact on 6-month response to antidepressant treatment and inflammatory biomarkers in depressed patients

Two thirds of patients suffering from a major depressive episode (MDE) do not reach a complete response with antidepressant drugs. This lack of response is due to several factors, including genetic determinants. Since major depressive disorder is associated with inflammatory and oxidative stress abnormalities, the metabolism of superoxide anions might be involved in non-response to antidepressant drugs. Superoxide anions are metabolized by manganese-dependent superoxide dismutase (SOD2) in the mitochondria. A functional genetic polymorphism (SOD2, rs4880), responsible of a 40% reduction in enzyme activity, is associated with anti-inflammatory response of rosuvastatin. We investigated the association of ala-allele of SOD2 rs4880 and both antidepressant efficacy and inflammatory parameters in patients treated for a MDE with antidepressant drugs. The Hamilton Depression Rating Scale (HDRS) score and levels of plasma CRP and inflammatory cytokines were assessed at baseline, one month (M1), 3 months (M3) and 6 months (M6) after antidepressant treatment. They were compared according to SOD2 genetic polymorphism. Of the 484 patients studied, 361 (74.6%) carried the ala-allele (Ala group), 123 (25.4%) of them had Val/Val genotype (Val/Val group). No significant difference was observed between the Ala and Val/Val groups neither for baseline clinical characteristics, nor for HDRS scores, response/remission rates, plasma CRP and cytokine levels throughout the study. The rs4880 SOD2 genetic polymorphism was not associated with the clinical response and cytokines levels after antidepressant treatment. These data suggest that SOD2 is not a major genetic determinant of antidepressant response. Other genes of the oxidative stress pathways should be explored in further studies.

Regulators of glucocorticoid receptor function in an animal model of depression and obesity

Obesity is a disease that often co-occurs with depression, and some evidence indicates that chronic stress in the perinatal period, in association with overactive glucocorticoids, can cause permanent changes that increase the risk of the development of both depression and obesity later in life. However, the mechanism responsible for the overly potent action of glucocorticoids in both depression and obesity is not known. The aim of the present study was to determine the expression of glucocorticoid receptors (GRs) and mineralocorticoid receptors (MRs) and the factors that affect GR function (FKBP51, Bag-1 and HSP70) in a prenatal stress animal model of depression, a model of obesity and a model of both depression and obesity. Prenatal stress but not high-fat diet (HFD) was found to decrease the GR concentration in the frontal cortex. The level of the Bag-1M (46 kDa) isoform was also decreased in this structure but only in prenatal-stressed animals that did not show depression-like behaviour in the Porsolt test and were fed the standard diet (STD). In the model of depression employed here, decreases in MR expression and GR co-chaperone (FKBP51) levels in the hippocampus were also observed, and HFD intensified the prenatal stress-induced changes in MR expression. The obtained results indicated that prenatal stress affected the expression of GRs, MRs and their co-chaperones in the brain, but its effects were different in the frontal cortex and hippocampus. The decrease in MR density in the hippocampus and increased plasma insulin level seemed to be the most significant changes observed in the model of the co-occurrence of depression and obesity, which could limit the neuroprotective effects associated with the activation of MR and be a marker of peripheral insulin resistance, respectively. This article is protected by copyright. All rights reserved.

Brain structural networks and connectomes: the brain-obesity interface and its impact on mental health

PURPOSE

Obesity is a complex and multifactorial disease identified as a global epidemic. Convergent evidence indicates that obesity differentially influences patients with neuropsychiatric disorders providing a basis for hypothesizing that obesity alters brain structure and function associated with the brain's propensity toward disturbances in mood and cognition. Herein, we characterize alterations in brain structures and networks among obese subjects (ie, body mass index [BMI] ≥30 kg/m²) when compared with non-obese controls.

PATIENTS AND METHODS

We obtained noninvasive diffusion tensor imaging and generalized q-sampling imaging scans of 20 obese subjects (BMI=37.9±5.2 SD) and 30 non-obese controls (BMI=22.6±3.4 SD). Graph theoretical analysis and network-based statistical analysis were performed to assess structural and functional differences between groups. We additionally assessed for correlations between diffusion indices, BMI, and anxiety and depressive symptom severity (ie, Hospital Anxiety and Depression Scale total score).

RESULTS

The diffusion indices of the posterior limb of the internal capsule, corona radiata, and superior longitudinal fasciculus were significantly lower among obese subjects when compared with controls. Moreover, obese subjects were more likely to report anxiety and depressive symptoms. There were fewer structural network connections observed in obese subjects compared with non-obese controls. Topological measures of clustering coefficient (C), local efficiency (E_local), global efficiency (E_global), and transitivity were significantly lower among obese subjects. Similarly, three sub-networks were identified to have decreased structural connectivity among frontal-temporal regions in obese subjects compared with non-obese controls.

CONCLUSION

We extend knowledge further by delineating structural interconnectivity alterations within and across brain regions that are adversely affected in individuals who are obese.

Effect of Metabolic Syndrome on Late-Life Depression: Associations with Disease Severity and Treatment Resistance

BACKGROUND/OBJECTIVES

Metabolic syndrome (MetS) is the co-occurrence of obesity and metabolic derangements. Prior research implicates MetS in prolongation of the course of depression in older adults, but its effect on antidepressant response is unknown in this population. The objective was to determine whether MetS and related metabolic dyscrasias are associated with decreased rate of remission from depression in older adults treated pharmacologically for depression.

DESIGN

Secondary analysis of a randomized controlled trial.

SETTING

Three academic medical centers in North America.

PARTICIPANTS

Adults aged 60 and older (mean age 69.1) with major depressive disorder (MDD) (N = 435).

INTERVENTION

Open-label, protocolized treatment with extended-release venlafaxine for 12 or more weeks.

MEASUREMENTS

Time to remission from depression, with remission defined as a Montgomery-Åsberg Depression Rating Scale (MADRS) score of 10 or less at last two visits.

RESULTS

Two hundred twenty-two participants (51%) met criteria for MetS at baseline; MetS was associated with greater severity (MADRS score) and chronicity of depression at baseline. Remission was achieved in 182 participants (42%). In the unadjusted analysis, MetS was associated with prolonged time to remission (hazard ratio for remission = 0.71, 95% confidence interval = 0.52-0.95), but this relationship was not significant in the adjusted model; greater number of MetS components and lower high-density lipoprotein cholesterol had similar effects. Only diastolic blood pressure (DBP) was a significant predictor of time to remission before and after adjustment, with higher DBP predicting longer time to remission. Insulin sensitivity did not predict time to remission.

CONCLUSION

The presence of MetS in older adults with depression was associated with greater symptom severity and chronicity of depression, which appears to have accounted for the poorer antidepressant response observed in those with MetS. Additionally, our preliminary finding of an association between higher DBP and poorer antidepressant response bears further examination and replication.