Dispositional Optimism Predicts Antidepressant Treatment Response in Major Depressive Disorder: Potential Relevance for Positive Psychology Interventions

Abstract

Positive psychology interventions have demonstrated efficacy in treating Major Depressive Disorder (MDD), but the relationship between dispositional optimism and antidepressant treatment response remains unclear. In this study, optimism (Revised Life Orientation Test, LOT-R) and depression severity (Hamilton Depression Rating Scale, HDRS) were assessed in healthy unmedicated MDD participants (N=86) and healthy controls (N=65). Treatment response (≥50% reduction in HDRS) was measured in 46 of these MDD participants after 8 weeks of open-label selective serotonin reuptake inhibitor (SSRI) treatment. MDD participants reported significantly lower pre-treatment optimism than healthy controls. Higher pre-treatment optimism, as well as greater increase in optimism (pre-treatment to 8 weeks), were associated with greater likelihood of being an SSRI responder versus non-responder (β=0.21, p=0.013; β =0.32, p=0.01, respectively), even when ratings of optimism/pessimism were excluded from depression severity/response ratings. The findings are consistent with the importance of aspects of positive psychology in treatment response. Possible mechanisms are discussed.

INTRODUCTION

Major Depressive Disorder (MDD) is the second-leading cause of years lived with disability in the United States (The US Burden of Disease Collaborators et al., 2018) with an estimated lifetime prevalence of over 20% among adults (Hasin et al., 2018). Compounding the seriousness of MDD, pharmacological treatment is only effective for some patients, with response rates estimated between 49–65% (Rush et al., 2006; Sinyor et al., 2010). Investigating factors related to MDD treatment response is crucial for the development and tailoring of therapies that address patient needs.

Traditional disease models for MDD phenomenology and antidepressant treatment response have generally focused on the presence of pathology rather than on the absence of protective factors. However, more recently, constructs of positive psychology, such as dispositional optimism, have been researched in relation to physical and psychiatric health markers and outcomes, including inflammation, cardiovascular disease, stroke, recovery from surgery, and coping skills (Carver & Scheier, 2014; Nes & Segerstrom, 2006; Roy et al., 2010; Scheier & Carver, 2018). Optimism is correlated with several moderating factors, including genetics (Caprara et al., 2009), as well as environmental and life history variables, such as socioeconomic status, childhood family conditions, and employment status (Ek et al., 2004; Heinonen et al., 2006). Dispositional optimism, generally thought of as trait-like, given its relative stability over time (although studies have observed changes in optimism over time) (Scheier & Carver, 2018), refers to the tendency to have positive expectations for the future (Carver et al., 2010). Low optimism is often seen in MDD, and was found to be a risk factor for the development of MDD in a 10-year longitudinal study of U.S. nurses (Weitzer et al., 2022). Martínez et al. (2021) found low optimism to predict incidence of antidepressant medication among a cohort of nurses in Spain (Martos Martínez et al., 2021). In a cohort of Finnish public sector employees, Kronström et al. (2014) similarly found low optimism and high pessimism to positively predict antidepressant treatment initiation and length of pharmacologic treatment, further suggesting a link between depression, low optimism, and high pessimism (Kronström et al., 2014). However, optimism or pessimism are not diagnostic criteria of MDD per the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V) (American Psychiatric Association, 2013). Nonetheless, as a psychological variable, optimism may either affect or reflect the severity and course of MDD and relate to its treatment response.

Although dispositional optimism is associated with better physical health outcomes in numerous areas, including cardiovascular disease, rehospitalization after surgery, and overall mortality (Scheier & Carver, 2018), its relationship to MDD treatment outcome is not well understood. Of relevance to the present study, dispositional optimism is associated with recovery from certain illnesses. A systematic review found that dispositional optimism is associated with positive health outcomes and greater engagement with health behaviors for patients with cardiovascular disease (Schiavon et al., 2017). A study of 309 post-coronary artery bypass graft (CABG) patients found that those identified as optimistic (assessed with the Revised Life Orientation Test, LOT-R) were less likely to be hospitalized post-surgery (Scheier et al., 1999). A study of 369 patients recovering from clinically diagnosed acute coronary syndrome found that optimism, also assessed with the LOT-R, predicted better overall physical health and lower risk of depressive symptoms 12 months after cardiac diagnosis (Ronaldson et al., 2015). In the same study population, depressive symptoms and optimism interacted in predicting future morbidity and mortality, in that depressive symptoms predicted poor outcomes only in those individuals with low optimism ratings.

The role of optimism in treatment response from psychiatric disease is understudied. Kronström et al. (2014) find that higher optimism and lower pessimism are associated with higher likelihood of stopping antidepressant medication before the end of the follow up, suggesting a possible link between optimism and treatment response; however, this study did not track depressive symptoms and it is unclear if medication discontinuation was due to greater likelihood of remission or dropout (Kronström et al., 2014). To our knowledge, only two studies to date have tracked depressive symptoms in relation to optimism. In the first study, which included 430 post-CABG surgery patients with mild-to-moderate depression (selected via Patient Health Questionnaire [PHQ]-9 screening), optimism was associated with lower baseline depression symptom ratings, more favorable response to selective serotonin reuptake inhibitor (SSRI) treatment (measured via Hamilton Depression Rating Scale, HDRS), and lower likelihood of rehospitalization within 8 months (Tindle et al., 2012). That study, however, did not select patients based on DSM MDD criteria, did not exclusively include unmedicated patients for the baseline assessments, and included only patients with severe medical comorbidity, which itself is associated with depression (Hare et al., 2014). The second study by Salazar-Fraile et al. (2018) measured the correlation of baseline optimism with antidepressant treatment response after 12 weeks of SSRI treatment in 89 patients diagnosed with MDD during screening in primary care offices in Spain. They excluded patients with dementia, bipolar disorder, psychosis, or substance abuse. They found that optimism positively predicted treatment response as measured by change in PHQ-9 score (Hazard ratio = 1.2). This study, however, did not exclude patients with other medical or psychiatric comorbidities or concurrent medication regimens, and relied on self-reporting for measuring depression severity and SSRI adherence (Salazar-Fraile et al., 2018). To our knowledge, the effects of optimism on antidepressant response in otherwise healthy and unmedicated participants with well-characterized MDD have yet to be examined. Understanding the effect of optimism on antidepressant response is important to determine whether optimism should be considered relevant to prognosis and treatment.

The present study 1) compared dispositional optimism in healthy MDD participants and healthy controls and 2) examined whether dispositional optimism at baseline, as well as changes in optimism over 8-weeks of guideline-based open-label SSRI treatment, predict antidepressant treatment response in physically healthy unmedicated individuals with rigorously diagnosed MDD who had antidepressant medication adherence specifically assessed. We hypothesized that (i) dispositional optimism would be lower in the unmedicated participants with MDD compared to healthy controls, (ii) higher dispositional optimism at baseline would predict more favorable response to SSRI treatment in MDD participants, and (iii) increases in optimism over the course of treatment would be associated with more favorable response to SSRI treatment in MDD participants, even when ratings of optimism/pessimism are excluded from the overall depression severity ratings.

MATERIALS AND METHODS

Recruitment procedures and study participants

MDD outpatients (N = 86) and healthy controls (N = 65) were recruited by flyers, Craigslist postings, newspaper ads and clinical referrals. All study procedures were approved by the UCSF Institutional Review Board, and all participants signed informed consent forms and were compensated for their participation. SSRI treatment was provided at no cost. All diagnoses, including MDD, were made according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) guidelines, which were in use at the start of this study. Diagnoses were established using the Structured Clinical Interview for DSM-IV-Text Revision (Kübler, 2013) and verified in a separate unstructured diagnostic evaluation by a Board-certified psychiatrist. Depression symptom severity was assessed in MDD participants using the observer-rated 17-item Hamilton Depression Rating Scale (HDRS) (Hamilton, 1960). All MDD participants had a minimum 17-item HDRS score of 17. MDD participants were excluded if they met DSM-IV criteria for any of the following: (1) bipolar disorder, (2) alcohol or substance abuse within the preceding 6 months, (3) Post-traumatic stress disorder (PTSD) or an eating disorder within 1 month of entering the study, and (4) any history of psychosis outside of a major depressive episode, or the presence of any psychotic symptoms during the current major depressive episode. To remain in the study, MDD participants could not begin new courses of psychotherapy during the study period, but those already in ongoing psychotherapy, with no planned changes during the course of the study, were allowed to continue.

Potential healthy controls, also screened with the Structured Clinical Interview for DSM-IV TR, were excluded for any history of DSM-IV Axis-I diagnoses. All study participants (MDD and controls) were free of chronic illnesses or acute illnesses or infections, inflammatory disorders, neurological disorders, or any other medical conditions considered to be potentially confounding, as assessed by history, physical examinations, and complete routine blood screening.

All participants were free of psychotropic medications (including antidepressants), hormone supplements, steroid-containing birth control or other potentially interfering medications, including vitamin supplements above the U.S. recommended daily allowances (e.g. >90 mg/day for Vitamin C) for at least 6 weeks prior to enrollment. They had not had any vaccinations for at least 6 weeks prior to enrollment. For MDD participants, short-acting sedative-hypnotics were allowed as needed for sleep up to a maximum of three times per week, but none within one week prior to any evaluations. All participants had to have a negative urine toxicology screen for drugs of abuse and a urine test for pregnancy in women of child-bearing age on the study visit days.

Procedure

At the baseline interview, MDD participants (N=86) completed the observer-rated 17-item HDRS, and all participants (including healthy controls, N=65) completed the self-rated LOT-R, described below. Following this, 46 MDD participants completed 8-weeks of standard, protocol-based open-label outpatient SSRI treatment. The remaining 40 MDD participants were studied before SSRI treatment was added to our study protocol. The choice of SSRI was based on physician and patient discussion of prior history, family history and expected side effects. Outpatient medication compliance, as well as clinical evaluations and assessments of drug tolerability, were assessed by telephone after 1-week and in-person after 4- and 8-weeks of treatment, when pill counts were performed, and blood was drawn for ascertainment of plasma SSRI concentrations. In all cases, plasma antidepressant concentrations were in the expected clinical range, except in one participant (a ‘Responder’) whose final sertraline dose was low [25 mg] due to intolerance of side effects at higher doses, suggesting excellent outpatient compliance. At the end of 8 weeks of SSRI-treatment, depressive symptom severity of MDD participants was again observer-rated with the HDRS and participants completed the LOT-R.

Measures

Optimism

The Revised Life Orientation Test (LOT-R) (Herzberg et al., 2006) is the most widely utilized self-rated measure of optimism. It is a 10-item questionnaire pertaining to one’s expectations for the future. The overall LOT-R score consists of an optimism and a pessimism subscale (positively and negatively worded items, respectively). The optimism and pessimism subscales have been found in at least one study to behave statistically as independent constructs, suggesting the possibility that they represent distinct psychological phenomena (Herzberg et al., 2006). However, previous work on optimism has utilized the overall LOT-R score (Salazar-Fraile et al., 2018). Therefore, we report our findings using the overall score, as well as each of the optimism and pessimism subscales. A higher optimism subscale score represents higher optimism, a higher pessimism score represents lower pessimism, and the sum of the optimism and pessimism scores equals the overall LOT-R score. Throughout the manuscript, “optimism” refers to the total LOT-R score, unless otherwise specified that the optimism or pessimism subscale is being discussed. Participants completed the LOT-R at baseline and at the end of 8 weeks of SSRI-treatment. A change (Δ) LOT-R was computed by subtracting the baseline LOT-R score from the week 8 LOT-R score.

Depressive symptoms

The 17-item HDRS is among the most widely used observer-rated scales for assessing MDD severity and changes with treatment (Hamilton, 1960; Timmerby et al., 2017). The HDRS was obtained at baseline and at 8 weeks. MDD participants with ⩾50% improvement in HDRS ratings from baseline to week 8 were classified as “SSRI responders,” while those with <50% improvement were classified as “SSRI non-responders.” Since the 17-item HDRS contains one item (“Depressed Mood”) that includes ratings similar to pessimism or lack of optimism (“sadness, hopeless, helpless, worthless”), we additionally calculated a 16-item version that excluded this item for a sensitivity analysis, in order to avoid overlap with the major independent variable in this study, optimism. The rest of the items in the 17-item HDRS do not overlap with the LOT-R (they included Feelings of Guilt, Suicide, Insomnia, Work and Activities, Retardation, Agitation, Anxiety Psychic, Anxiety Somatic, Somatic Symptoms, Genital Symptoms, Hypochondriasis, Loss of Weight, Insight).

Statistical Analysis

An independent samples t-test was used to compare baseline optimism between the MDD group and the healthy control group; baseline optimism was checked for normality with a normal-appearing Q-Q plot and skewness statistic between −0.5 and 0.5. A series of logistic regressions was used to examine associations between LOT-R, Δ LOT-R, and their subscales with SSRI responder status as the dichotomous outcome variable, while controlling for baseline HDRS score. All analyses were completed using SPSS software. Data are available upon request to the authors.

RESULTS

Baseline Differences in Sample Characteristics

We compared baseline sample characteristics between control and MDD groups, and SSRI responder vs. non-responder groups (Table 1). T-tests performed for age, years of education, and current MDD episode duration (SSRI responder vs. non-responder), revealed no significant group differences (all ps > 0.05), except that the total MDD group had significantly fewer years of education compared to the controls. Chi-square tests performed for race, ethnicity, gender, and prescribed SSRI revealed no significant group differences (all ps > 0.05).

Table 1:

Demographics and Descriptive Statistics of Study Variables

	Control (N=65)	MDD (N=86)	MDD w/follow up
			SSRI Responder (N=14)	SSRI Non-responder (N=32)
Race
Asian	17 (26%)	18 (21%)	0 (0%)	6 (19%)
Black or African American	2 (3%)	7 (8%)	1 (7%)	2 (6%)
Caucasian	37 (57%)	45 (52%)	8 (57%)	18 (56%)
More than one	5 (8%)	8 (9%)	3 (21%)	3 (9%)
Other	4 (6%)	8 (9%)	2 (14%)	3 (9%)
Ethnicity
Hispanic or Latino/a	8 (12%)	14 (16%)	4 (29%)	4 (13%)
Not Hispanic or Latino/a	57 (88%)	72 (84%)	10 (71%)	28 (88%)
Gender
Male	25 (38%)	38 (44%)	4 (29%)	15 (47%)
Female	40 (62%)	48 (56%)	10 (71%)	17 (53%)
Age [mean (SD)]	36.8 (13.6)	36.9 (13.1)	35.4 (9.9)	37.7 (12.5)
Years of education [mean (SD)]	16.9 (1.7)	15.9 (2.3)*	16.1 (2.4)	15.6 (2.5)
Current MDD episode duration (days)	N/A	2347 (3661)	2204 (3458)	3419 (3967)
Prescribed Selective Serotonin Reuptake Inhibitor (SSRI)
Citalopram	N/A	N/A	1 (7%)	5 (16%)
Escitalopram	N/A	N/A	6 (43%)	8 (25%)
Fluoxetine	N/A	N/A	3 (21%)	5 (16%)
Sertraline	N/A	N/A	4 (29%)	14 (44%)
HDRS (baseline) [mean (SD)]	N/A	19.6 (3.2)	18.8 (3.6)	19.7 (3.3)
HDRS (week 8) [mean (SD)]	N/A	11.2 (5.8)	5.8 (3.3)	15.3 (3.9)*
Optimism (baseline) [mean (SD)]
LOT-R Total score	18.6 (3.1)	9.3 (4.6)*	12.1 (4.2)	8.3 (4.1)*
LOT-R Optimism subscale	9.3 (1.7)	4.3 (2.3)*	5.9 (2.2)	3.6 (2.1)*
LOT-R Pessimism subscale	9.3 (1.8)	5.0 (2.6)*	6.2 (2.3)	4.6 (2.5)*
Optimism (week 8)
LOT-R Total score	N/A	N/A	17.6 (3.5)	10.5 (3.7)*
LOT-R Optimism subscale	N/A	N/A	8.8 (2.2)	4.9 (2.1)*
LOT-R Pessimism subscale	N/A	N/A	8.9 (2.4)	5.6 (2.0)*

^*p < 0.05, comparing MDD to Control, and SSRI Non-responder to SSRI Responder.

HDRS = 17-item Hamilton Depression Rating Scale; LOT-R = Life Orientation Test-Revised.

Greater HDRS scores indicate greater depression severity. Greater Total LOT-R scores and Optimism subscale scores indicate greater optimism. Lower pessimism subscale scores indicate greater pessimism.

Baseline Differences in Optimism between MDD and healthy controls

A t-test showed that baseline optimism (LOT-R total score) was higher among the control group compared to the MDD group, t(149) = 14.8, p < 0.001; for mean and SD see Table 1. As expected, the MDD group was less optimistic on the optimism subscale, t(149) = 15.0; p < 0.001; Table 1) and more pessimistic on the pessimism subscale, t(149) = 11.9; p < 0.001; Table 1).

Baseline Optimism and SSRI Treatment Response

A logistic regression model showed an association between baseline optimism (LOT-R total score) and odds of antidepressant response (for statistics, see Table 2, Test 1; also see Figure 1a). For each unit increase in baseline optimism, the odds of SSRI response increased by 1.24-fold, controlling for baseline depression severity (total HDRS score). Analyses with the LOT-R subscales yielded similar results. Logistic regression showed an association between baseline optimism subscale and odds of SSRI response (Table 2, Test 2). For each unit increase on the optimism subscale, odds of SSRI response increased by 1.62-fold, controlling for baseline depression severity. There was also a trend (p = 0.064) of an association between baseline pessimism subscale and odds of SSRI response (Table 2, Test 3). For each unit increase on the pessimism subscale (higher score represents less pessimism), the odds of SSRI response increased 1.30-fold.

Table 2:

Logistic Regression Analyses Examining the Relationship Between Optimism and SSRI Treatment Response

Test		β (Standard Error)	Exp(β)	p-value
1	Baseline optimism (LOT-R total score)	0.21 (0.086)	1.24	0.013
2	Baseline optimism subscale	0.48 (0.18)	1.62	0.006
3	Baseline pessimism subscale	0.26 (0.14)	1.30	0.064
4	Δ Optimism (Δ LOT-R total score)	0.32 (0.12)	1.37	0.011
5	Δ Optimism subscale	0.40 (0.19)	1.49	0.038
6	Δ Pessimism subscale	0.63 (0.26)	1.87	0.014

Outcome: Odds of SSRI response.

LOT-R = Life Orientation Test-Revised

Δ = change in LOT-R (from baseline to 8 weeks)

Note: All models control for baseline depression severity (17-item HDRS).

Figure 1: Baseline Optimism and Δ Optimism among SSRI Responders and Non-Responders.

1a. Baseline optimism (LOT-R total score) among SSRI responders and non-responders.

1b. Δ Optimism (from baseline to 8 weeks) among SSRI responders and non-responders.

Error bars represent 95% confidence intervals.

Change in Optimism (Δ Optimism) and SSRI Treatment Response

A logistic regression analysis showed an association between Δ optimism over the study period (week 8 optimism score minus baseline optimism score) and odds of SSRI response (for statistics, see Table 2, Test 4; also see Figure 1b). For each unit increase in Δ optimism, odds of SSRI response increased 1.37-fold, controlling for baseline depression severity. Analyses with the LOT-R subscales paralleled this finding. Logistic regression showed an association between Δ optimism subscale over the study period and odds of SSRI response (Table 2, Test 5). For each increase in Δ optimism subscale over the study period, odds of SSRI response increased 1.49-fold, controlling for baseline depression severity. Logistic regression showed an association between Δ pessimism subscale over the study period and odds of SSRI response (Table 2, Test 6). For each unit increase on the Δ pessimism subscale over the study period (representing a decrease in pessimism), the odds of SSRI response increased 1.87-fold.

Sensitivity Analyses

Since the “Depressed Mood” item of the HDRS (item #1 of the scale) partly assesses pessimism, this could have accounted for the relationships we observed. Therefore, we also assessed the HDRS score minus that particular item. When we removed the depressed mood item from the 17-item HDRS score, baseline optimism and Δ optimism remained associated with increased likelihood for SSRI treatment response in logistic regression analyses, controlling for baseline depression severity (baseline optimism: β = 0.16, SE = 0.08, p = 0.046; Δ optimism: β = 0.28, SE = 0.11, p=0.015).

DISCUSSION

As expected, dispositional optimism, as assessed by the LOT-R total score, was lower in unmedicated MDD participants compared to healthy controls. Our novel findings include an association of higher baseline optimism with an increased likelihood of SSRI antidepressant treatment response and an association of greater increase in optimism over the study period with increased likelihood of SSRI treatment response.

This study builds on results of previous studies that have found an association between baseline optimism and SSRI treatment response in patients recovering from surgery (Tindle et al., 2012) and those presenting to a primary care physician with depressive symptoms (Salazar-Fraile et al., 2018) by reproducing these findings in depressed participants who 1) were physically healthy and rigorously screened for MDD diagnosis and for the absence of psychiatric, somatic, or substance abuse comorbidities, 2) were medication-free for a minimum of 6 weeks before testing, (3) were treated prospectively with standard outpatient SSRI treatment, and 4) had their plasma SSRI levels monitored to ensure medication compliance and adequate treatment. This study also extends prior ones by showing that SSRI responders not only had higher baseline optimism but had higher increases in optimism over the study period, even when a restricted HDRS scale (eliminating ratings on a pessimism item) was used to assess treatment response.

While the mechanisms linking optimism with SSRI response are incompletely understood, biological mechanisms may play a role. Several studies have linked optimism with decreased threat-related expression of genes involved with inflammation (Uchida et al., 2018) and with decreased circulating inflammatory cytokines (Ikeda et al., 2011; Roy et al., 2010). Treatment-associated decreases in inflammation are associated with better antidepressant response (Lindqvist et al., 2017). Further, optimism has been related to anatomic and functional brain or neural properties that affect emotion-regulating parts of the brain (Aue et al., 2021; Dolcos et al., 2016; Kuzmanovic et al., 2016; Sharot et al., 2007; Yang et al., 2013), which may interact with SSRI effects (Godlewska et al., 2012; Harmer et al., 2009; Harmer & Cowen, 2013). Future studies should examine inflammation and neural mechanisms as mediators of the optimism-treatment response relationship.

Our finding that participants who responded to SSRI treatment tended to demonstrate a stronger increase in optimism during the study period (even after excluding optimism/pessimism ratings from determinations of overall response), may also be consistent with recent neuropsychological theories of SSRI action suggesting that the ability of SSRIs to produce “positive biases in processing emotional information” may contribute to their antidepressant efficacy (Harmer & Cowen, 2013).

Even apart from a biological explanation, it is possible that those with a more optimistic bias have different health behaviors (Yue et al., 2022), which influence antidepressant response. Patients with greater optimism regarding medical treatments tend to be more medication-compliant (Millstein et al., 2016), but in our study, all participants showed evidence of excellent medication adherence.

A psychosocial lens may also be applied to understanding possible connections between optimism and treatment response. Optimism is correlated with resilience, or one’s ability to cope with or recover from challenges. There is an association between dispositional optimism and self-identifying as having an “approach”-style response to stress, aimed towards reducing or managing to stressors, instead of an avoidant response (Nes & Segerstrom, 2006). A study of men aged 70–89 found that dispositional optimism was associated with greater resilience to loneliness, with optimism predicting lower self-rated loneliness over time independent of social network and depression (Rius‐Ottenheim et al., 2012), and loneliness itself is increasingly tied to depression (Hussain et al., 2023; Lee et al., 2019; Straus et al., 2022). Optimists report greater social network size (Andersson, 2012) and perceived social support (Srivastava et al., 2006), which is in turn correlated with resilience. Since our study found optimism to be linked to increased likelihood of antidepressant response, future research may also focus on resilience and social support as optimism-related mediators or therapeutic targets for patients receiving antidepressant treatment.

It is also possible that an “expectancy bias” or “placebo effect,” contributes to the association between optimism and antidepressant response, such that those who expect to become better tend to improve more, independent of the direct effect of the treatment itself. The literature shows a possible association between optimism and the placebo response (Kern et al., 2020). While investigators try to minimize or account for expectancy or placebo responses in designing treatment trials (Evans et al., 2021), clinicians typically value and enlist “placebo” or “expectancy” responses as valuable adjuncts to other treatment modalities (Rosenfield & Bernstein, 2021). It is increasingly difficult to ascribe expectancy effects or placebo responses to purely psychological processes, since a growing literature indicates biological mechanisms underlying their therapeutic effects. A substantial body of research, largely in Parkinson’s disease patients, indicates that “placebo” treatments, combined with the expectation of improvement, activate dopamine reward pathways in the ventral striatum and nucleus accumbens (Scott et al., 2007), which, in turn, activates endogenous opioid systems, at least in studies of placebo analgesia (Benedetti, 2014). This may partially imitate SSRI-mediated brain activation of the limbic system and prefrontal cortex (Mayberg et al., 2002). In the case of social anxiety disorder, expectancy (overt administration of escitalopram compared to “covert” administration of escitalopram in which participants were incorrectly told they were receiving active placebo but actually received escitalopram) resulted in substantially greater anti-anxiety effects but did not change the degree of serotonin transporter (SERT) binding on PET scan. However, dopamine transporter (DAT) binding significantly decreased in the right putamen and pallidum and the left thalamus in the “expectant” group, which was associated with more favorable responses, but DAT binding increased in those areas in the non-expectant group (Hjorth et al., 2021), again linking expectancy with activation of dopaminergic reward pathways. In a separate analysis by the same group, overt and covert escitalopram administration also resulted in differential fMRI responses to treatment between groups in the posterior cingulate, superior temporal, and inferior frontal gyri (Faria et al., 2017). The authors concluded that their results point to a pronounced placebo component in SSRI-treatment of social anxiety disorder that favors a biopsychosocial over a purely biomedical explanatory model for SSRI efficacy. Finally, certain early metabolomic changes are parallel between placebo and active SSRI treatment (Kaddurah-Daouk et al., 2013). In any event, our results suggest that optimism is a useful predictor of, and perhaps a contributor to, treatment response, and further, that optimism, although conceptualized as a trait variable, may be modifiable with antidepressant treatment (Harmer et al., 2009; Harmer & Cowen, 2013).

One of the documented outcomes of positive psychology in depression has been in positive-psychology-oriented CBT, showing superior efficacy and greater patient preference compared to traditional problem-focused CBT (Geschwind et al., 2019, 2020). However, the importance of the focus on optimism has not been specifically assessed, and the evidence so far has been limited by small sample sizes (White et al., 2019). Our findings that baseline optimism predicted better antidepressant responses, and that optimism improved more over the course of treatment for treatment responders, are congruent with the literature suggesting that enhancing optimism through strategies such as cognitive behavioral therapy may improve treatment response (Webb et al., 2013). Indeed, research suggests that specifically teaching learned optimism techniques can significantly reduce depression (Doering et al., 2018; Meevissen et al., 2011; Miranda et al., 2017) and that positive psychology-based interventions (Pan et al., 2022) and promoting positive response expectations (Kirsch, 2017) may hold promise for treating depression. Our data cannot directly address that, however, since we did not experimentally target optimism biases.

Limitations.

Limitations of our study include modest sample sizes, as well as the limited nature of the LOT-R scale, in that it does not probe one’s reasons for having positive or negative expectations for their future, nor does it explicitly assess expectancy of benefit from the upcoming drug trial. It is additionally important to note that the 17-item HDRS includes “depressed mood” (including hopelessness/helplessness and pessimism) as one of its 17 items. However, sensitivity analyses showed that baseline optimism and Δ optimism remained associated with increased likelihood for treatment response even when the “depressed mood” HDRS item containing hopelessness/helplessness and pessimism was removed from the scale. Also, it is important to note that our SSRI treatment protocol was community-standard open- label, and our results should not be uncritically extrapolated to double-blind treatment studies, in which expectancy and “placebo” effects are lessened.

Strengths.

Our study has several key strengths: (1) We employed a rigorous methodology for diagnosing MDD and for evaluating patient adherence to the SSRI regimen. This is in contrast to the approach taken by Salazar-Fraile and colleagues (2018), who did not objectively measure adherence and thus speculated that optimism’s effect on antidepressant response may be mediated through better medication adherence —a theory our data does not support. (2) We instituted a minimum of six weeks for antidepressant washout before starting our study. (3) We carefully selected participants by excluding patients with other medical or psychiatric comorbidities or concurrent medications. (4) We measured both the overall LOT-R score as well as the optimism and pessimism subscales. (5) Finally, we assessed the effect of optimism independent of changes in depressed mood.

Conclusions.

Our results suggest that baseline optimism and increases in optimism during treatment are associated with increased likelihood of SSRI treatment response in MDD.

Indeed, expectancies may be responsible for part of the improvement observed with antidepressant treatment. If so, managing patients’ expectations may be an important part of clinical care (Rutherford et al., 2010). While causality cannot be inferred from our data, it is possible that psychological therapies aimed at increasing optimism (e.g., positive CBT) may augment antidepressant responses and may yield superior outcomes compared to traditional CBT (Geschwind et al., 2019, 2020). Prospective randomized trials assessing this important hypothesis are warranted. Future studies may examine more deeply the factors behind patients’ feelings of optimism and pessimism in order to better characterize the role of optimism in depression and response to antidepressants, and to identify how optimism can be enhanced and therapeutic approaches be tailored accordingly. Qualitative interviews with patients may reveal whether those with higher dispositional optimism tend to be more optimistic about the prospect of their antidepressant treatment, and whether this relates to their more favorable response to treatment. Further, our finding that optimism increases over the course of treatment in responders but not in non-responders raises the possibility that increases in optimism or positive bias may predict future response, although we did not specifically assess that. Our findings also raise the possible importance of biological correlates of optimism as contributing to antidepressant efficacy; the biology of “optimism” is a still-understudied attribute. In summary, our results may have both therapeutic implications as well as implications for understanding personality and biological factors relevant to antidepressant treatment response.

DATA AVAILABILITY STATEMENT:

The authors may be contacted for access to data. Authors agree to make data and materials supporting the results or analyses presented in their paper available upon reasonable request.