Depressive symptoms after stroke are associated with worse recovery

Abstract

Objective

More studies are needed on how depressive symptoms in stroke patients can impact outcomes. We evaluated the relationship between depression symptom severity and motor outcomes in a cohort of patients with motor impairment from ischemic stroke.

Method

We enrolled consecutive ischemic stroke patients without a clinical diagnosis of depression who presented to a single-center urban academic referral hospital. The Patient Health Questionnaire-9 (PHQ-9) scale was used to measure depression symptom severity at three months. Three assessments of motor function were collected at stroke onset and three months: Fugl-Meyer upper extremity (FM-UE), Motricity Index, and Action Research Arm Test (ARAT). We assessed the association between three-month severity on PHQ-9 scores with the outcome measures using univariable and multivariable linear regression models.

Results

Fifty-seven patients (mean age 67.8 ± 17.0 years; 50.9% male; 59.6% Caucasian) were included in the final analysis. Mean (standard deviation) outcome scores at three months were PHQ-9: 6.39 (5), Motricity Index: 86.93 (30.04), FM-UE: 52.67 (17.83), and ARAT: 43.77 (20.03). After adjusting for age, initial National Institute of Health Stroke Scale, and if patient discharged after hospitalization on a selective serotonin reuptake inhibitor, sex, and baseline motor outcome, we found that for every point increase in PHQ-9, the Motricity Index decreased by 0.82 points (p = 0.02) and the FM-UE decreased by 0.77 points (p = 0.049).

Conclusion

Depressive symptoms are common in the stroke population. Depressive symptoms after stroke are associated with multiple types of motor impairments. We need better understanding of the biologic and psychologic aspects of depression involved in stroke recovery.

Keywords

medical illness comorbidity biopsychosocial psychiatry depression stroke

Introduction

There are over 7 million survivors of stroke in the United States¹ and 80 million worldwide.² Globally, stroke remains the second leading cause of death with the highest incidence occurring in east Asia, followed by the eastern European region.² Stroke was also the second most common cause of disability-adjusted life-years with more patients now living with chronic stroke.² Common impairments occurring immediately after stroke include limb paralysis and aphasia, but there are other important complications that occur late after stroke impacting long-term function and quality of life.³ Depression after stroke is stroke significant complication that can be easily overlooked but when recognized and treated can improve overall outcome.⁴ Depression can be defined as feeling low in mood, with loss of interest, and lack of pleasure, persisting for at least two weeks after stroke.^5,6 Some studies report that depression can be divided into major and minor subtypes. Major depression is defined as major depressive disorder, while the minor depression is less well defined and usually diagnosed as a dysthymic disorder.⁷

Depression is common after stroke with a five-year prevalence ranging between 30% and 50%.^4,8 The highest incidence is between three to six months after stroke onset.⁹ Yet, there are studies that indicate depression can be present in up to 25% of patients at the time of index stroke hospitalization discharge.¹⁰ Importantly, these high rates of incidence and prevalence are important as depression is associated with increased rates of mortality.¹¹ Additionally, depression also has a negative impact on stroke functional outcomes.^12,13

While depression is associated with a worse overall functional outcome after stroke, few studies have reported the relationship between depression symptom severity and specific long-term motor and cognitive outcomes. This is especially important in identifying potential patients with depression after stroke. We followed a cohort of patients with ischemic stroke and motor deficits in order to determine if acquired depressive symptom severity at three months was also associated with motor and cognitive outcomes.

Methods

Study population

The local Institutional Review Board approved the study. We enrolled consecutive patients from January 2017 to December 2017 presenting to a single-center urban academic referral hospital with imaging-verified ischemic stroke. Written informed consent was obtained from patients. Information about index stroke hospitalization was prospectively recorded by two investigators (CL and AB) blinded to outcome data. We screened for depression by chart abstraction, medication review, and patient interview. Ischemic stroke was defined as sudden onset of neurologic deficits and confirmed with corresponding hyperintense and hypointense lesions on diffusion-weighted imaging and acquired diffusion coefficient sequences, respectively. We recorded prospectively: demographics, past medical history including depression, hospital course, use of selective serotonin reuptake inhibitor (SSRI) since stroke onset.^3,14 Clinical and radiographic data were prospectively reviewed to determine Trial of Org 10172 in Acute Stroke Treatment (TOAST) subtype for each confirmed case by consensus.¹⁵

Depressive symptoms assessment

Depressive symptom severity was measured using the Patient Health Questionnaire nine-item depression scale (PHQ-9) after three months from stroke. PHQ-9 assessments were collected by trained research assistant. Research assistants were blind to stroke outcome data. The PHQ-9 is a nine-item self-administered depression screening and diagnostic tool that is commonly used to evaluate patients with stroke for depression.^16–18 Our group decided to use this measure because in patients with stroke, the measure has excellent test–retest reliability (Intraclass Correlation Coefficient = 0.98), excellent interrater reliability (Intraclass Correlation Coefficient = 0.98), and excellent internal consistency (Cronbach’s alpha = 0.79).

Outcome measures

The stroke outcome measures were collected by Neurologist and Physical Medicine and Rehabilitation physicians and blind to depressive symptom assessment. Motor measures were collected by physical assessment within a week of stroke and after three months from stroke using the following validated measures for stroke patients: Motorcity index evaluating the hemiparetic side¹⁹ with a range of 0–100 points; Fugl-Meyer upper extremity (FM-UE) measure to assess motor functioning in the paretic arm²⁰ ranging from 0 to 66 points; and the Action Research Arm Test (ARAT) evaluation to assess specific arm function and mechanics²¹ ranging from 0 to 57 points. For these three measures, higher scores equate to better performance. In addition, we also collected the following: functional outcome using the modified Rankin Scale (mRS),²² stroke severity with the National Institute of Health Stroke Scale (NIHSS), and cognitive screen with the Montreal Cognitive Assessment (MOCA). Motor, mRS, NIHSS, and MOCA were all collected at the initial stroke hospitalization within a week of stroke onset and repeated at three months. Score changes were calculated by subtracting the initial score from the three-month score.

Statistical analysis

Descriptive statistics were calculated for all variables in the study. Frequency and percentage were reported for categorical variables, whereas mean (standard deviation) or median (interquartile range) were reported for continuous variables. Associations between PHQ-9 score and motor outcomes, both measured at three months, were assessed both visually and by the use of unadjusted linear regression models. For significant associations, adjusted models were then utilized to control for factors such as sex, age at presentation, stroke severity (NIHSS), whether the patient was discharged on SSRI, and baseline motor outcome value. Normality and linearity assumptions, and the overall model fit were assessed using graphical diagnostic techniques and tests for outliers. Reported p-values are unadjusted for multiple comparisons. All analyses were performed using R v3.4.3 statistical software. Underlying research materials related to the manuscript can be requested from the corresponding author at their discretion.

Results

Patient baseline characteristics are summarized in Table 1. Our sample included 57 patients (mean age 67.8 ± 17 years; 50.9% male; 59.6% Caucasian). Three patients (6.0%) had a prior history of depression. Table 2 summarizes outcome measures at baseline, three-month follow-up (F/U), and changes from baseline to three-month F/U, as well as PHQ-9 score at three-month F/U. All outcome measures improved from initial evaluation. In particular, negative changes in NIHSS and mRS scores demonstrated improved stroke severity and function at three months.

Table 1.

Overall baseline characteristics (n = 57).

Sex (%)
Female	28 (49.1)
Male	29 (50.9)
TOAST subtype (%)
Cardioembolic	14 (24.6)
Cryptogenic	7 (12.3)
Large artery atherosclerosis	13 (22.8)
Other determined	7 (12.3)
Small artery disease	16 (28.1)
Acute stroke treatment (IV-alteplase or thrombectomy) (%)
No	37 (68.5)
Yes	17 (31.5)
Age (mean (SD))	67.78 (16.95)
Race (%)
African-American/Black	11 (21.6)
Asian	4 (7.8)
Caucasian/White	34 (66.7)
Native American/Alaska Native	1 (2.0)
Other	1 (2.0)
Private insurance (%)
No	38 (66.7)
Yes	19 (33.3)
Premorbid functioning (%)
Independent	52 (91.2)
Any neuropsychologic impairment	5 (8.8)
Living situation (%)
Assisted living	1 (2.0)
Home alone	15 (30.6)
Home with family	32 (65.3)
Other	1 (2.0)
Discharged from hospital with SSRI (%)
No	43 (75.4)
Yes	14 (24.6)
Medical history prior to stroke
Depression (%)
No	47 (94.0)
Yes	3 (6.0)
Hypertension (%)
No	25 (43.9)
Yes	32 (56.1)
Diabetes (%)
No	45 (78.9)
Yes	12 (21.1)
Dyslipidemia (%)
No	34 (59.6)
Yes	23 (40.4)
Obesity (%)
No	54 (94.7)
Yes	3 (5.3)
Atrial fibrillation (%)
No	50 (87.7)
Yes	7 (12.3)
Cardiac disease (%)
No	45 (78.9)
Yes	12 (21.1)
Peripheral Artery disease (%)
No	56 (98.2)
Yes	1 (1.8)
Current smoking (%)
No	53 (93.0)
Yes	4 (7.0)

SD: standard deviation; TOAST: Trial of Org 10172 in Acute Stroke Treatment; SSRI: selective serotonin reuptake inhibitor.

Table 2.

PHQ-9 score and outcome measures (n = 57).

Mean (SD)
	Baseline scores	Three-month scores	Change in score from initial presentation to three months
NIHSS	7.79 (7.27)	2.32 (3.13)	−4.00 (−6.00, −2.00)^a
mRS	3.44 (1.60)	1.84 (1.29)	−1.60 (1.58)^a
Motricity arm	66.93 (35.81)	81.68 (27.31)	7.00 (0.00, 27.00)^a
Motricity leg	78.56 (28.66)	90.02 (14.15)	8.00 (0.00, 24.00)^a
Motricity side	72.92 (30.04)	86.93 (18.07)	8.00 (0.00, 28.00)^a
FM-UE	42.70 (22.65)	52.67 (17.83)	4.00 (1.00, 20.00)^a
MOCA	22.16 (7.16)	23.70 (7.14)	2.66 (3.46)^a
ARAT	32.58 (23.32)	43.77 (20.03)	3.00 (0.00, 14.25)^a
PHQ-9	–	6.39 (5.00)	–

SD: standard deviation; PHQ-9: Patient Health Questionnaire; mRS: modified Rankin Scale; NIHSS: National Institute of Health Stroke Scale; FM-UE: Fugl-Meyer upper extremity; ARAT: Action Research Arm Test; MOCA: Montreal Cognitive Assessment.

^aMedian (interquartile range) reported due to highly non-normal distribution.

Unadjusted, univariable linear regression models (Table 3) report associations between three-month F/U outcomes and three-month F/U PHQ-9 score. The PHQ-9 scores at three months was statistically significantly associated with Motricity arm (p = 0.006), Motricity side (p = 0.032), and FM-UE (p = 0.041). There was a borderline association between three-month PHQ-9 and ARAT score (p = 0.05). No association was found with NIHSS, mRS, Motricity leg, and MOCA score (Supporting Figure 1). An increase in PHQ-9 score (an increased severity of depression symptoms at three months relative to baseline) was associated with worse motor outcomes (Figure 1). In particular, a 1-point increase in PHQ-9 score was associated with a decrease of 2.46 points in the Motricity arm index (p = 0.006), 1.26 points in the Motricity side index (p = 0.032), 1.14 points in the FM-UE (p = 0.041), and 1.09 points in the ARAT (p = 0.055). When adjusted for potential confounders and clinically relevant factors, these associations were minimally changed (Table 3). After adjusting for sex, age at presentation, stroke severity (NIHSS), prescribed SSRI at acute hospital discharge, and baseline motor outcome, three-month F/U PHQ-9 score remained significantly associated with three-month F/U outcomes: Motricity arm (β = −1.99, p = 0.012), motricity side (β = −0.82, p= 0.022), and FM-UE (β = −0.77, p = 0.049). After adjusting, no significant association was found between three-month F/U PHQ-9 and three-month F/U ARAT (β = −0.73, p = 0.198).

Table 3.

Linear regression models with three-month F/U PHQ-9 as a predictor.

	Unadjusted models			Adjusted models
Outcome (at three-month F/U)	Estimate	Standard error	p-Value	Estimate	Standard error	p-Value
Motricity arm	−2.46	0.85	0.00599	−1.99	0.75	0.01210
Motricity side	−1.26	0.57	0.03231	−0.82	0.34	0.02166
FM-UE	−1.14	0.54	0.04054	−0.77	0.38	0.04947
ARAT	−1.09	0.55	0.05458	−0.73	0.55	0.19780
NIHSS	0.14	0.09	0.12659	–	–	–
mRS	0.06	0.04	0.14752	–	–	–
Motricity leg	−0.57	0.45	0.20821	–	–	–
MOCA	0.24	0.24	0.32422	–	–	–

PHQ-9: Patient Health Questionnaire; mRS: modified Rankin Scale; NIHSS: National Institute of Health Stroke Scale; FM-UE: Fugl-Meyer Upper Extremity; ARAT: Action Research Arm Test; MOCA: Montreal Cognitive Assessment.

Figure 1.

Plot of PHQ-9 scores with motor outcomes measures. PHQ-9: Patient Health Questionnaire; F/U: follow-up; FM-UE: Fugl-Meyer upper extremity; ARAT: Action Research Arm Test; MOCA: Montreal Cognitive Assessment.

Discussion

Our results demonstrate that in a cohort of patients with stroke presenting with motor weakness, an increase in severity of depressed symptoms as measured by PHQ-9 at three months was associated with several motor outcomes. We did not find a similar association with cognitive measures. At three months, our cohort of patients could be classified as having mild depressive symptoms on the PHQ-9. Our longitudinal outcomes also demonstrate expected improvements after stroke in lower NIHSS and mRS along with increased motor outcome measures. These findings indicate that even though our cohort had the clinically expected motor improvements after stroke, that they still exhibited mild depressive symptoms is concerning. In clinical practice after stroke, we often assume that improved motor deficits confirm that the patient is overall improving. Clinicians also rarely formally screen for depressive symptoms at the time of clinical F/U. Our results highlight the importance of screening for depressive symptoms as it might impact motor recovery.

SSRIs, the most common medication class used to treat depression, may have a positive impact on motor recovery. Only a quarter of our patients were discharged from acute hospitalization on SSRIs, as a new medication, though they all had documented motor weakness. A randomized controlled trial comparing a SSRI to placebo within 5 to 10 days after stroke showed significantly improved 90-day FM-UE scores.^22,23 This trial excluded patients with depression but did track depression severity symptoms. They found that the group on the SSRI had improved depression scores along with improved motor outcome scores. Our trial was not powered to distinguish between outcomes of patients discharged with SSRIs compared with those that were not. However, our results are consistent with this trial: in patients with increased severity in depressive symptoms, their motor outcomes were likely to be worse including on the FM-UE. Our findings are unique because we have also included the Motricity scores in addition to the FM-UE and ARAT measures. This highlights the important relationship of screening for depressive symptoms, motor outcomes, and the use of SSRIs.

In another randomized controlled pragmatic trial, patients on SSRI versus placebo were compared based on overall functional status using the mRS.²⁴ The study did not find any significant difference in functional outcome with the mRS between the SSRI and placebo groups. This result also parallels our result where increased depressive symptoms did not improve overall functional outcome as measured by the mRS. Functional measures such as the mRS are routinely used as primary outcome in randomized clinical trials of acute stroke treatment. The mRS is not sensitive enough to give clinicians adequate information about recovery patterns in longitudinal rehabilitation studies.²⁵ Patients with depressive symptoms often experience concomitant symptoms such as anxiety, sleep issues, and fatigue that can also impact recovery. Global functional outcomes do not generally account for these issues. Additionally, the mRS is more weighted toward independence in walking. Motor outcomes assessing for upper extremity like the FM-UE gives more information on specific modalities than what the mRS was designed to measure. Therefore, in patients with stroke, particularly with depressive symptoms and motor deficits, it is important to more carefully assess these patients beyond what global functional outcome measures like the mRS can capture.

The mechanism of why depression and SSRIs impact motor outcomes need to be further examined. In animal models, SSRIs have neuroprotective effects and can improve cognitive deficits post-stroke.²⁶ This finding led to a hypothesis that the brain’s serotoninergic system is one of the main facilitators of motor output.²⁷ In human patients with stroke, functional MRI studies have shown that SSRIs, compared to placebo, activate specific areas of the motor cortex and that the same motor regions are found hyperexcitable using transcranial magnetic stimulation.²⁸ Like other psychiatric disorders, there are also strong psychosocial factors that influence biologic factors associated with mood.⁴ To understand the full mechanism of post-stroke depression on stroke recovery, a thorough model that incorporates the biologic, psychologic, and social/economic factors must be considered.

Study limitations

A major strength of our study is the natural history design with rigorous F/U and objective measures of cognition, and motor outcomes. There are, however, several limitations. First, this study was conducted in a single urban academic institution. The differences between institutions and patient characteristics should not vary substantially, but our results may not be generalizable to all settings. Second, our cohort included only ischemic stroke patients, most of whom had mild to moderate impairment. Third, our cohort was underpowered to detect change based on SSRI use. We did not record psychotropic or serotonin and norepinephrine reuptake inhibitor use. Fourth, we did not obtain substantial social economic status of patients, which may have helped to develop a more complete model. Fifth, we have a limited F/U period of up to three months. Ideally, future studies would also collect outcomes at 6 months where a high incidence and 12 months where a high prevalence of depression after stroke occur. Finally, we did not assess pre-stroke depression severity. We only noted, based on medical record review and patient interview, whether there was prior a diagnosis or treatment of depression. Therefore, it is impossible to know how pre-stroke depression might influence post-stroke measures of depressed mood. Pre-stroke measures of depression are cumbersome to perform and not validated in hospitalized patients. Rather, we included the baseline functional status where >90% of our cohort were independent without any reported issues prior to their stroke.

Conclusion

Unfortunately, depressive symptoms are common in patients who have suffered stroke. Our results demonstrate that an increase in severity of depressed symptoms was associated with several motor outcomes after stroke. We believe that future studies should focus on larger cohorts of patients beginning at initial hospitalization in order to determine the natural history of post-stroke depression including change in symptoms and interventions taken to improve symptom severity. This study reaffirms the need to have a better understanding in the overlap of biologic, psychologic, and social aspects of depression involved in the patient’s stroke recovery.

Supplemental Material

IJP905459 Supplemental Material1 - Supplemental material for Depressive symptoms after stroke are associated with worse recovery

Supplemental material, IJP905459 Supplemental Material1 for Depressive symptoms after stroke are associated with worse recovery by Chen Lin, Ahmed Babiker, Nina Srdanovic, Masha Kocherginsky and Richard L Harvey in The International Journal of Psychiatry in Medicine

Footnotes

Acknowledgments

The authors would like to acknowledge Drs. Shyam Prabhakaran, Yurany Andrea Arevalo, and Muhammad A Mansour in their assistance in designing the study and data acquisition.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Chen Lin

Supplemental Material

Supplemental material for this article is available online.

References

Mozaffarian

, et al. Executive summary: heart disease and stroke statistics–2016 update: a report from the American Heart Association. Circulation 2016; 133: 447–454.

Global, regional, and national burden of stroke 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol 2019; 18: 439–458.

Lin

, et al. Predicting domain-specific health-related quality of life using acute infarct volume. Stroke 2017; 48: 1925–1931.

Robinson

Jorge

RE.

Post-stroke depression: a review. Am J Psychiatry 2016; 173: 221–231.

Hackett

, et al. Frequency of depression after stroke: a systematic review of observational studies. Stroke 2005; 36: 1330–1340.

Douven

, et al. Imaging markers of post-stroke depression and apathy: a systematic review and meta-analysis. Neuropsychol Rev 2017; 27: 202–219.

Morris

, et al. Are there two depressive syndromes after stroke? J Nerv Ment Dis 1994; 182: 230–234.

Ayerbe

, et al. Natural history, predictors and outcomes of depression after stroke: systematic review and meta-analysis. Br J Psychiatry 2018; 202: 14–21.

Angelelli

, et al. Development of neuropsychiatric symptoms in poststroke patients: a cross-sectional study. Acta Psychiatr Scand 2004; 110: 55–63.

10.

Astrom

Adolfsson

Asplund

Major depression in stroke patients. A 3-year longitudinal study. Stroke 1993; 24: 976–982.

11.

Williams

Ghose

Swindle

RW.

Depression and other mental health diagnoses increase mortality risk after ischemic stroke. Am J Psychiatry 2004; 161: 1090–1095.

12.

Bilge

, et al. Depression and functional outcome after stroke: the effect of antidepressant therapy on functional recovery. Eur J Phys Rehabil Med 2008; 44: 13–18.

13.

Paolucci

, et al. One-year follow-up in stroke patients discharged from rehabilitation hospital. Cerebrovasc Dis 2000; 10: 25–32.

14.

Sangha

, et al. Quality of life in patients with TIA and minor ischemic stroke. Neurology 2015; 85: 1957–1963.

15.

Goldstein

, et al. Improving the reliability of stroke subgroup classification using the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria. Stroke 2001; 32: 1091–1098.

16.

Lincoln

, et al. The validity of questionnaire measures for assessing depression after stroke. Clin Rehabil 2003; 17: 840–846.

17.

Parikh

, et al. The impact of poststroke depression on recovery in activities of daily living over a 2-year follow-up. Arch Neurol 1990; 47: 785–789.

18.

Prisnie

, et al. Validating screening tools for depression in stroke and transient ischemic attack patients. Int J Psychiatry Med 2016; 51: 262–277.

19.

Demeurisse

Demol

Robaye

Motor evaluation in vascular hemiplegia. Eur Neurol 1980; 19: 382–389.

20.

Prabhakaran

, et al. Inter-individual variability in the capacity for motor recovery after ischemic stroke. Neurorehabil Neural Repair 2008; 22: 64–71.

21.

Yozbatiran

Der-Yeghiaian

Cramer

SC.

A standardized approach to performing the action research arm test. Neurorehabil Neural Repair 2008; 22: 78–90.

22.

Mills

JA.

Systemic lupus erythematosus. N Engl J Med 1994; 330: 1871–1879.

23.

Chollet

, et al. Fluoxetine for motor recovery after acute ischaemic stroke (FLAME): a randomised placebo-controlled trial. Lancet Neurol 2011; 10: 123–130.

24.

Dennis

, et al. Effects of fluoxetine on functional outcomes after acute stroke (FOCUS): a pragmatic, double-blind, randomised, controlled trial. Lancet 2019; 393: 265–274.

25.

Cramer

, et al. Stroke recovery and rehabilitation research: issues, opportunities, and the National Institutes of Health StrokeNet. Stroke 2017; 48: 813–819.

26.

Lim

, et al. Fluoxetine affords robust neuroprotection in the postischemic brain via its anti-inflammatory effect. J Neurosci Res 2009; 87: 1037–1045.

27.

Jacobs

Fornal

CA.

Serotonin and motor activity. Curr Opin Neurobiol 1997; 7: 820–825.

28.

Dam

, et al. Effects of fluoxetine and maprotiline on functional recovery in poststroke hemiplegic patients undergoing rehabilitation therapy. Stroke 1996; 27: 1211–1214.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.22 MB