Anxiety and depression one year after the first stroke among Lebanese survivors: proportions, changes, and predictors

Abstract

Background

About 1 in 2 stroke survivors suffer from anxiety and depression in the first year after stroke. This study aimed to calculate the proportions of 1-year post-stroke anxiety and depression (PSA and PSD), evaluate their changes over time, and identify the contributing factors among first-ever stroke survivors in Lebanon, where relevant research has been scarce.

Methods

A hospital-based multicenter study was conducted among 150 subjects aged ≥ 18 years involving scheduled home visits at 3, 6, and 12 months after stroke index. Several scales were employed, including Hospital Anxiety and Depression Scale (HADS), Mini-Mental State Examination (MMSE), modified Rankin Scale (mRS), Short Form Health Survey (SF12), National Institutes of Health Stroke Scale (NIHSS), Visual Analogue Scale (VAS), Douleur Neuropathique 4 (DN4), Modified Ashworth Scale (MAS), and Fatigue Severity Scale (FSS), to assess levels of anxiety and depression, cognitive function, disability degree, quality of life, stroke severity, general pain, central pain, spasticity, and fatigue after stroke, respectively. Descriptive analyses were performed to describe baseline and stroke characteristics and to calculate proportions of PSA and PSD, followed by univariate and multivariable analyses to identify the contributing factors.

Results

High HADS scores were reported in our cohort, with the most prevalent symptoms occurring within the first 3 months after stroke index (77.3% for PSD and 51.2% for PSA with HADS ≥ 8). Despite a slight decrease over the subsequent 6 and 12 months, proportions remained elevated, affecting at least 40% to 60% of survivors. PSA was a consistent independent predictor of PSD (Adjusted Odds Ratio ≅ 2). Other contributing factors to PSA and PSD were highlighted, including a history of atrial fibrillation, longer sedentary hours, high scores of NIHSS and mRS, lower scores of SF12 and MMSE, and the presence of immobility-related problems, of which high DN4 scores were independent predictors. Better 1-year psychological outcomes were noticed in those with higher educational levels and employment after stroke.

Conclusion

Routine psychological screening and support for stroke survivors are urgently needed. By identifying factors and emphasizing early detection, our research offers valuable insights that can inform clinical practice and improve the well-being of stroke survivors.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-025-06997-9.

Introduction

About 1 in 2 stroke survivors suffer from anxiety and depression in the first year after stroke, which hamper their recovery. Post-Stroke Anxiety and Depression (PSA and PSD) are associated with high mortality rate, more pronounced cognitive deficits, worsened functional outcome, lower Quality of Life (QoL), and high rates of suicidal ideation. Although these known impacts, stroke survivors at risk of psychological disorders are rarely routinely screened for anxiety or depression or formally assessed by neuropsychologists [1].

Multiple overlapping biological, physical, social, and psychological factors contribute to neuropsychiatric sequelae following stroke. Some risk factors are still controversial and need further investigations. The most common predictors of PSD and PSA have been found to be stroke severity [Adjusted Hazard Ratio (AHR) = 1.52], physical impairments, fatigue, history of depression, and post-stroke cognitive impairment [2–6]. Other factors were identified as predictors of PSD, including lack of social support after stroke [3, 4, 6], PSA [4, 6], and history of any mood disorder [6].

Two systematic reviews were conducted to assess PSA and PSD globally using different assessment tools from 2 days to 10 years post-stroke. They reported an overall percentage of anxiety “caseness” of 24.2% [95% Confidence Interval (CI), 21.5–26.9], a cumulative proportion of anxiety disorder of 21.4% (95% CI, 19.2–23.5) between 1 and 5 months, and 31.8% (95% CI, 17.8–45.7) between 6 and 12 months. However, the proportion of clinically significant anxiety symptoms has been ranged from 23.6% (95% CI, 18.9–28.2) between 1 and 5 months to 21.5% (95% CI, 15.3–27.8) between 6 and 12 months. The overall proportion of any depressive disorder was 33.5% (95%CI, 30.3- 36.8). However, in the community/outpatient setting (mean period of 10 months after stroke), the proportion of any depressive disorder was 12.8% (95%CI, 11.3–14.3) [7, 8].

In the Middle East and North Africa (MENA) region, Kaadan et al., in 2017 conducted a systematic review of 34 papers studying the prevalence of PSD at different time points using various assessment tools, and reported lower rates in Saudi Arabia (17%) and Iran (18%) and higher rates in Algeria (56.1%), Jordan (64%), and Morocco (73.2%) [9].

Nevertheless, to date, several stroke papers have been published in Lebanon [10–34] but there are no studies investigating the complications that would result from stroke, including neuropsychiatric problems that are believed to impede recovery and threaten stroke survivors’ lives. Hence, the present study aimed at calculating the rates of PSA and PSD through HADS scores, examining the changes in anxiety and depressive symptoms severity and profiles, and determining the various risk factors—baseline socio-demographics, pre-existing conditions, lifestyle, stroke-related factors, National Institutes of Health Stroke Scale (NIHSS) scores, Short Form Health Survey 12 (SF12) of QoL scores, modified Rankin Scale (mRS) scores, and complications, at 3-, 6-, and 12 months following FES among Lebanese survivors.

Methods

Study design

We conducted a hospital-based prospective, multicenter longitudinal study recruiting subjects from 10 medical centers in Beirut and Mount Lebanon. It spanned 15 months of follow-up, starting from February 2018.

Study population

The participants were first-ever ischemic or hemorrhagic stroke survivors who were admitted to the hospitals between February and May 2018. The inclusion criteria were: 1) age ≥ 18 years, 2) Lebanese nationality, 3) having had a first stroke, well identified by the following ICD-10 codes (I60-I64): stroke, stroke, ischemic stroke, hemorrhagic stroke, intracerebral hemorrhage or cerebrovascular/embolic thrombosis, and 4) diagnosis confirmed clinically and by brain imaging. The exclusion criteria were 1) admission for recurrent stroke or Transient Ischemic Attack (TIA), or 2) pre-existing neuro-psychiatric and cognitive disorders (dementia, Alzheimer disease, ataxia, Bell’s palsy, brain tumor, cerebral aneurysm, epilepsy, seizures, Parkinson disease, meningitis, hydrocephalus, encephalitis, aphasia, brain attack, attention disorders, anxiety, depression, and other mood disorders).

Study procedures

Following approval from ethics committees and hospital directors, we initiated contact with eligible subjects admitted to participating hospitals between February and May 2018. Subsequently, written informed consent was obtained from eligible subjects or their legal representatives to enroll in the study. Thereafter, clinical data were meticulously gathered by three extensively trained and proficient investigators. This was achieved through face-to-face interviews with subjects and their proxies during pre-scheduled visits, as well as subsequent follow-up home visits at 3-, 6-, and 12-month intervals post-stroke. The collection of clinical information was facilitated through a comprehensive questionnaire encompassing various domains, including socio-demographic characteristics such as age, gender, place of residence, marital status, number of children, age of the subject's custodian, level of education of the subject and custodian, employment status, number of household members, number of rooms, and type of health insurance. Lifestyle factors including dietary habits, smoking status, physical activity, alcohol and substance consumption, and social support were also recorded. Additionally, health indicators and medical history, incorporating anthropometric indices, family/medical/surgical history, comorbidities, and treatments received by subjects, were documented. Aspects of stroke and their severity, encompassing types/subtypes, location, symptoms, length of hospital stay, disease severity, degree of disability, assessment of Quality of Life (QoL), and potential post-stroke complications, such as neuropsychiatric disorders, cognitive impairments, hyperglycemia, fatigue, post-stroke pain, falls, pressure ulcers, pulmonary and urinary infections, deep vein thrombosis, pulmonary embolism, seizures, stroke recurrence, and mortality, were also comprehensively evaluated.

Definitions and outcome measures

The World Health Organization (WHO) defined stroke as “rapidly developed clinical signs of focal (or global) disturbance of cerebral function, lasting more than 24 h or leading to death, with no apparent cause other than of vascular origin” [35]. Ischemic stroke was classified using the Trial of Org 10,172 in the Acute Stroke Treatment (TOAST) into 5 subtypes: 1) Large-Artery Atherosclerosis (LAA), 2) Cardioembolism (CE), 3) Small-Vessel Occlusion (SVO), 4) Stroke of Other determined Etiology (SOE), and 5) Stroke of Undetermined Etiology (SUE) [36]. The same criteria of initial stroke were considered to define stroke recurrence. Both ischemic and hemorrhagic stroke recurrences were documented. Specifically, only recurrences emerging at least 21 days following the initial event were deemed eligible for inclusion [37]. Mortality was characterized as any death occurring within the 12-month period following the onset of the initial stroke. In instances where a patient passed away during the year-long follow-up, the cause of death was investigated by referencing hospital or primary care medical records.

A battery of tools and patient-reported scales were used to evaluate stroke severity, anxiety and depression as follows:

The NIHSS, developed in 1989 by Brott T. and colleagues [38], stands as the foremost reliable and valid instrument for assessing stroke severity [39]. Comprising 15 components, it evaluates consciousness, language, motor function, sensory loss, visual fields, extra-ocular movements, coordination, neglect, and speech. Scores range across five levels: 0 signifies no stroke, 1–4 denotes minor stroke, 5–14 indicates moderate stroke, 15–20 signifies moderate to severe stroke, and 21–42 represents severe stroke [40], with a Cronbach’s alpha coefficient of (r) = 0.942. We employed the validated Arabic version of the NIHSS, exhibiting intra-rater and inter-rater agreements of 0.94 and 0.95, respectively [41].

The HADS, designed by Zigmond and Snaith in 1983, serves to assess anxiety and depression levels [42]. Comprising 14 self-report items, it features two distinct scales: one for depression and another for anxiety. Scores are categorized as follows: 0–7 denotes normal, 8–10 signifies borderline, and 11–21 indicates abnormal. Widely recognized for its validity and reliability, the HADS is among the most frequently employed tools for screening anxiety and depression post-stroke [2, 6, 43–46]. Its use in stroke rehabilitation is endorsed by international guidelines [47].

Moreover, other instruments were applied to assess the degree of disability, QoL, cognitive performance, general pain, neuropathic pain, fatigue, and social support after stroke, were defined in the attached Additional file 1.

Statistical methods

Collected data were coded, introduced, and entered to the SPSS version 25 (SPSS™ Inc., Chicago, IL USA). Descriptive analyses were applied using means and standard deviations for continuous variables and percentages for categorical variables to describe the baseline characteristics and stroke characteristics, and to calculate the proportions of post-stroke complications, including psychiatric disorders, anxiety, and depression post-stroke. Univariate logistic regression analyses were performed. The explanatory variables were tested one by one against the dependent variable for the presence of a significant association using the multinomial logistic regression. The strength of association was interpreted using the Unadjusted Odds Ratio (UOR) with 95% CI. In addition, given the multiple comparisons conducted, we applied Bonferroni correction to adjust the significance threshold and reduce the risk of type I error. The adjusted significance level was set based on the number of comparisons, ensuring that the overall significance remained at p ≤ 0.05. Then, multivariable analysis using multinomial logistic regressions adjusted for age, gender, education, and stroke severity, was made to identify the independent predictors of borderline cases and confirmed cases of anxiety and depression post-stroke. The variables with univariate association at p ≤ 0.05 were entered into logistic regression. The goodness-of-fit statistic is reported to ensure the model provides a good fit for the data (p > 0.05). Strength of association was interpreted using the Adjusted Odds Ratio (AOR) with 95% CI. A p-value ≤ 0.05 was considered statistically significant.

Results

Baseline and stroke characteristics

Out of 183 subjects who met the predefined inclusion criteria, 150 subjects were enrolled in the study (82% of response rate). Of those, one was lost to follow-up at 12 months as well as a total of 32 mortalities were reported, 27 (18%) in the 3 months, three (2.4%) in the 3 to 6 months, and two (1.7%) in the 6 to 12 months following the first stroke. A total of 117 subjects survived up to 12 months (our last visit). The mean age of participants was 73.69 ± 12.11 years. Eighty-eight (58.7%) were males versus 62 (41.3%) were females. The majority were married (78%), 34.7% were provided care from caregivers, 87.3% were unemployed/without any profession/retired, and 21.2% were illiterate/less educated. One hundred sixteen (77.3%) and 78 (52%) had pre-existing HBP and dyslipidemia, respectively (Table 1). A collective of 143 individuals (comprising 95.3% of the sample) experienced ischemic stroke, delineated as follows: 48.8% were attributed to SVO, 46.3% to LAA, and 5% to CE etiology, while a smaller subset of 7 individuals (4.7%) suffered from intracerebral hemorrhagic stroke. In terms of localization, 46.7% exhibited lesions in the left hemisphere, whereas 40% displayed involvement in the right hemisphere. Merely 6.8% of the subjects underwent intravenous thrombolysis. A significant portion (70%) were unable to communicate verbally or in writing during the stroke episode and experienced unilateral limb weakness.

Table 1.

Baseline characteristics of the study population

Characteristics	Frequency (N)	Percentage (%)
Age (N = 150)	Mean (± SD) = 73.69 (± 12.11)
Gender (N = 150)
Male	88	58.7
Female	62	41.3
Marital status (N = 150)
Single/Widowed/Divorced	33	22.0
Married	117	78.0
Comorbidities (N = 150)
DM	60	40.0
DL	78	52.0
AF	47	31.3
MI	29	19.3
HBP	116	77.3
Other CVDa	20	13.3
Social security (n = 150)	125	83.3
Educational level (N = 150)
Illiterate/Primary or Complementary education	104	69.3
Secondary or University education	46	30.7
Professional status post stroke (N = 150)
Person without any profession/retired	101	67.3
Unemployed	30	20.0
Employed	19	12.7
Presence of a caregiver (N = 150)	52	34.7
Age of caregiver (N = 52)
Adult: 20–40 years	9	17.3
Middle age: 40–60 years	35	67.3
Elderly: > 60 years	8	15.4
Household members (N = 128)
Living alone	5	3.9
Living with family members	123	96.1
HCI (N = 128)
≤ 1	112	87.5
> 1	16	12.5
Region (N = 150)
Beirut	33	22.0
Mount Lebanon	107	71.3
Othersb	10	6.7

N Frequency, % Percentage, SD Standard Deviation, DM Diabetes Mellitus, DL Dyslipidemia, AF Atrial Fibrillation, MI Myocardial Infarction, HBP High Blood Pressure, CVD Cardiovascular Diseases

^aOther CVD: coronary artery disease, cardiomyopathy, arrhythmia, chronic heart failure, and thoracic aortic aneurysm

^bOther regions: North Lebanon, Bekaa, South Lebanon

Rates of anxiety and depression, and changes over 1-year post-stroke

High proportions of PSD and PSA were recorded in Lebanese stroke survivors in the 3 months following stroke. Out of 123 survivors, 95 experienced depressive symptoms (77.3%) [20 borderline cases (8 ≤ HADS_D ≤ 10) (16.3%) and 75 confirmed cases of depression (HADS_D ≥ 11) (61%)]. Sixty-three subjects (51.2%) experienced anxiety symptoms [22 borderline cases (8 ≤ HADS_A ≤ 10) (17.9%) and 41 confirmed cases of anxiety (HADS_A ≥ 11) (33.3%)]. These scores slightly decreased in the following 6 and 12 months but remained elevated accounting for at least 40% of the survivors for PSA and 60% for PSD. It is worth mentioning that the proportions of subjects who developed only anxiety were very low at 3-, 6-, and 12 months post-stroke (n = 1, 0.7% each) in addition to 8 borderline cases of anxiety at 3- and 6 months post-stroke (n = 4, 2.7% each) [Figs. 1, 2, and 1S (Additional file 1)].

Fig. 1.

Proportions of post-stroke anxiety at 3, 6, and 12 months after stroke among Lebanese first-ever stroke survivors. HADS_A, Hospital Anxiety and Depression Scale_ Anxiety score

Fig. 2.

Proportions of post-stroke depression at 3, 6, and 12 months after stroke among Lebanese first-ever stroke survivors. HADS_D, Hospital Anxiety and Depression Scale__Depression score

Sixty-three subjects remained depressed (confirmed cases) (52.5%) from 3 to 6 months, 48 from 6 to 12 months (41.7%), versus 45 from 3 to 12 months (39.2%). Thirty-five were improved (30.4%) from 3 to 12 months (changes from depression to borderline cases or normal cases as well as borderline cases to normal cases). On the other hand, 30 subjects remained anxious (confirmed cases) from 3 to 6 months (25%), 24 from 6 to 12 months (20.9%), versus 20 from 3 to 12 months (17.4%). Twenty-nine were improved (25.2%) from 3 to 12 months (changes from anxiety to borderline cases or normal cases as well as borderline cases to normal cases) (Fig. 2S, Additional file 1).

Risk factors of anxiety and depression over 1-year post-stroke

Tables 2, 3, and 4 represent the univariate analyses’ results of factors predicting PSA; however, Tables 1S, 2S, and 3S in the Additional file 1 represent the factors in numbers, percentages and means. One-year PSA (3, 6- and 12 months) was significantly higher in subjects who did not resume work post-stroke (32% to 39%) compared to those who did (11% to 13%) (UOR ≅ 5). Other factors were significantly positively associated with 1-year PSA: history of AF (42% to 54%) (UOR ≅ 3), increasing duration of hospital stay (UOR = 1.1), higher NIHSS (UOR = 1.2), mRS (UOR ≅ 5), FSS (UOR ≅ 4), VAS (UOR ≅ 2), DN4 scores (UOR ≅ 2–3), and MAS scores (UOR ≅ 3), falls at 6 months (UOR between 3 to 8), pressure ulcers at 3 months (UOR between 5 to 16), contractures (UOR between 4 and 7), and pneumonia at 3- and 6 months post-stroke (UOR between 6 to 9). However, the factors that were negatively associated with 1-year PSA were presence of health insurance (UOR ≅ 0.3), less sedentary hours (< 7 h/day) (UOR ≅ 0.1), increasing PCS and MCS component scores of the SF12 QoL (UOR = 0.8 each component), and higher MMSE scores (UOR ≅ 0.3). Higher educational level was negatively associated with PSA at 12 months only (UOR = 0.2). Moderate SSRS predicted a borderline case of anxiety at 6 months (UOR = 8) but lowered the risk of PSA at 12 months (UOR = 0.4). Nevertheless, recurrent stroke at 3 months increased almost 6 folds the risk of a borderline anxiety case at 3 months post-stroke.

Table 2.

Baseline factors associated with borderline case of anxiety (8 ≤ HADS_A ≤ 10) and post-stroke anxiety (PSA) (HADS_A ≥ 11) at 3, 6, and 12 months using multinomial logistic regression analysis

Baseline characteristics	3 months post-stroke				6 months post-stroke				12 months post-stroke
	Borderline case		PSA		Borderline case		PSA		Borderline case		PSA
	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]
Age, mean (years)	0.655	1.010 [0.967—1.055]	0.722	1.006 [0.972—1.042]	0.476	1.017 [0.971—1.064]	0.555	0.989 [0.955—1.025]	0.85	0.996 [0.950—1.043]	0.784	1.006 [0.967—1.046]
Gender, Female	0.657	1.250 [0.467—3.349]	0.551	0.778 [0.340—1.778]	0.186	0.486 [0.166—1.417]	0.328	0.658 [0.284—1.524]	0.643	0.769 [0.253—2.339]	0.247	0.569 [0.220—1.476]
Educational level, Higher education	0.499	0.700 [0.249—1.971]	0.106	0.484 [0.201—1.167]	0.339	0.592 [0.202—1.733]	0.183	0.548 [0.226—1.329]	0.411	0.616 [0.194—1.954]	0.018	0.246 [0.077—0.787]
Professional status post stroke
Person without any profession/retired	0.490	0.625 [0.165—2.374]	0.941	1.045 [0.319—3.423]	0.358	0.532 [0.139—2.043]	0.728	0.809 [0.244—2.680]	0.49	0.625 [0.165—2.374]	0.941	1.045 [0.319—3.423]
Unemployed	0.187	2.917 [0.594—14.327]	0.047	4.333 [1.022—18.382]	0.132	3.5 [0.685—17.889]	0.030	5.2 [1.173—23.043]	0.187	2.917 [0.594—14.327]	0.047	4.333 [1.022—18.382]
Employed	Ref		Ref		Ref		Ref		Ref		Ref
Health insurance	0.642	0.704 [0.160—3.095]	0.032	0.303 [0.102—0.902]	0.004	0.138 [0.035—0.537]	0.017	0.215 [0.061—0.757]	0.004	0.149 [0.040—0.554]	0.136	0.376 [0.104—1.358]
SSRS, Moderate	0.832	1.133 [0.357—3.600]	0.320	0.643 [0.269—1.536]	0.048	8.182 [1.020—65.615]	0.527	0.755 [0.317—1.802]	0.78	1.216 [0.308—4.798]	0.049	0.383 [0.147—0.995]
Sedentary duration
1 to 6 h/day	0.013	0.210 [0.061—0.720]	< 0.001	0.035 [0.007—0.177]	0.216	0.446 [0.124—1.603]	< 0.001	0.078 [0.020—0.312]	0.013	0.121 [0.023—0.645]	0.001	0.069 [0.014—0.346]
7 to 11 h/day	0.032	0.239 [0.065—0.882]	0.076	0.407 [0.151—1.098]	0.696	0.782 [0.227—2.690]	0.142	0.479 [0.179—1.279]	0.172	0.420 [0.121—1.458]	0.085	0.400 [0.141—1.134]
≥ 12 h/day	Ref		Ref		Ref		Ref		Ref		Ref
History of AF	0.01	4 [1.402—11.409]	0.021	2.833 [1.167—6.877]	0.169	2.110 [0.729—6.109]	0.017	2.914 [1.210—7.018]	0.299	1.833 [0.584—5.753]	0.008	3.565 [1.397—9.098]

Bold numbers represent significant association with p-value ≤ 0.05

HADS_A Hospital Anxiety and Depression Scale_ Anxiety score, PSA Post-stroke Anxiety, UOR Unadjusted Odds Ratio, CI Confidence Interval, SSRS Social Support Rating Scale, AF Atrial Fibrillation 

Table 3.

Stroke-related factors associated with borderline case of anxiety (8 ≤ HADS_A ≤ 10) and post-stroke anxiety (PSA) (HADS_A ≥ 11) at 3, 6, and 12 months using multinomial logistic regression analysis

Stroke-related factors	3 months post-stroke				6 months post-stroke				12 months post-stroke
	Borderline case		PSA		Borderline case		PSA		Borderline case		PSA
	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]
Duration of hospital stay	0.338	1.042 [0.958—1.134]	0.007	1.099 [1.026—1.176]	0.152	1.057 [0.980—1.140]	0.010	1.089 [1.021—1.162]	0.177	1.052 [0.977—1.131]	0.025	1.072 [1.009—1.140]
NIHSS at 3 months	< 0.001	1.170 [1.082—1.265]	< 0.001	1.207 [1.122—1.298]	< 0.001	1.145 [1.063—1.234]	< 0.001	1.193 [1.113—1.279]	0.004	1.113 [1.035—1.196]	< 0.001	1.169 [1.094—1.249]
NIHSS at 6 months					0.003	1.136 [1.044—1.235]	< 0.001	1.198 [1.109—1.294]	0.003	1.134 [1.043—1.233]	< 0.001	1.198 [1.110—1.294]
NIHSS at 12 months									0.009	1.141 [1.034—1.259]	< 0.001	1.2 [1.099—1.312]
mRS at 3 months	0.030	1.520 [1.041—2.218]	< 0.001	2.541 [1.693—3.549]	0.138	1.330 [0.912—1.939]	< 0.001	2.413 [1.650—3.527]	0.007	1.847 [1.179—2.892]	< 0.001	2.689 [1.725—4.192]
mRS at 6 months					0.161	1.307 [0.899—1.899]	< 0.001	2.036 [1.458—2.844]	0.003	1.976 [1.268—3.079]	< 0.001	2.539 [1.681—3.835]
mRS at 12 months									0.001	2.196 [1.406—3.430]	< 0.001	2.817 [1.851—4.286]
SF12 of QoL
PCS at 3 months	0.454	0.975 [0.913—1.042]	0.005	0.908 [0.850—0.971]	0.502	0.977 [0.914—1.045]	0.002	0.893 [0.830—0.961]	0.049	0.909 [0.827—0.999]	0.006	0.888 [0.816—0.966]
MCS at 3 months	0.978	1.001 [0.948—1.056]	0.005	0.935 [0.892—0.979]	0.188	1.039 [0.981—1.101]	0.014	0.942 [0.897—0.988]	0.144	0.955 [0.898—1.016]	0.010	0.932 [0.884—0.983]
PCS at 6 months					0.813	0.994 [0.941—1.048]	< 0.001	0.899 [0.848—0.953]	0.028	0.923 [0.860—0.991]	< 0.001	0.881 [0.822—0.945]
MCS at 6 months					0.164	1.037 [0.985—1.091]	0.001	0.924 [0.882—0.968]	0.033	0.939 [0.886—0.995]	0.002	0.922 [0.877—0.970]
PCS at 12 months									0.001	0.895 [0.840—0.953]	< 0.001	0.885 [0.835—0.937]
MCS at 12 months									< 0.001	0.905 [0.858—0.955]	< 0.001	0.892 [0.848—0.937]

Bold numbers represent significant association with p-value ≤ 0.05

HADS_A Hospital Anxiety and Depression Scale_ Anxiety score, PSA Post-stroke Anxiety, UOR Unadjusted Odds Ratio, CI Confidence Interval, NIHSS National Institutes of Health Stroke Scale, mRS modified Rankin Scale, SF12 of QoL Short Form Health Survey of Quality of Life, PCS physical component summary, MCS mental component summary

Table 4.

Post-stroke complications associated with borderline case of anxiety (8≤HADS_A≤10) and post-stroke anxiety (PSA) (HADS_A≥11) at 3, 6, and 12 months using multinomial logistic regression analysis

Post-stroke complications	3 months post-stroke				6 months post-stroke				12 months post-stroke
	Borderline case		PSA		Borderline case		PSA		Borderline case		PSA
	p-value	UOR [95%CI]	p-value	UOR [95%CI]	p-value	UOR [95%CI]	p-value	UOR [95%CI]	p-value	UOR [95%CI]	p-value	UOR [95%CI]
Recurrent stroke at 3 months	0.028	5.588 [1.210 - 25.818]	0.526	0.475 [0.048 - 4.733]	0.276	2.417 [0.494 - 11.822]	0.833	0.829 [0.144 - 4.761]	0.590	1.595 [0.291 - 8.739]	0.466	0.447 [0.051 - 3.897]
MMSE at 3 months (≤ 23/>23)	0.689	0.811 [0.291 - 2.260]	0.014	0.233 [0.073 - 0.749]	0.440	1.504 [0.533 - 4.245]	0.084	0.382 [0.128 - 1.137]	0.202	0.417 [0.109 - 1.599]	0.012	0.072 [0.009 - 0.565]
MMSE at 6 months (≤ 23/>23)					0.019	0.293 [0.105 - 0.820]	0.004	0.290 [0.124 - 0.679]	0.002	0.135 [0.039 - 0.465]	<0.001	0.149 [0.055 - 0.406]
MMSE at 12 months (≤ 23/>23)									0.002	0.168 [0.053 - 0.534]	<0.001	0.175 [0.067 - 0.461]
FSS at 3 months	0.027	1.644 [1.058 - 2.552]	<0.001	2.362 [1.526 - 3.656]	0.651	1.088 [0.754 - 1.571]	0.001	2.166 [1.391 - 3.375]	0.011	2.192 [1.199 - 4.008]	0.001	2.347 [1.398 - 3.938]
FSS at 6 months					0.094	1.397 [0.945 - 2.064]	<0.001	3.469 [2.132 - 5.645]	0.001	2.613 [1.492 - 4.576]	<0.001	3.887 [2.185 - 6.912]
FSS at 12 months									<0.001	2.383 [1.568 - 3.622]	<0.001	2.655 [1.794 - 3.929]
VAS at 3 months	0.456	1.062 [0.907 - 1.244]	0.001	1.296 [1.119 - 1.502]	0.655	1.037 [0.885 - 1.215]	0.003	1.250 [1.080 - 1.445]	0.062	1.196 [0.991 - 1.444]	0.003	1.286 [1.088 - 1.520]
VAS at 6 months					0.200	1.138 [0.934 - 1.388]	<0.001	1.476 [1.229 - 1.772]	0.010	1.341 [1.071 - 1.678]	<0.001	1.509 [1.229 - 1.854]
VAS at 12 months									0.001	1.512 [1.194 - 1.914]	<0.001	1.664 [1.339 - 2.067]
DN4 at 3 months (<4/≥4)	0.016	4.245 [1.313 - 13.726]	0.005	4.286 [1.556 - 11.804]	0.113	2.650 [0.795 - 8.831]	0.003	4.517 [1.676 - 12.177]	0.104	2.818 [0.807 - 9.841]	0.001	5.314 [1.916 - 14.739]
DN4 at 6 months (<4/≥4)					0.071	3.5 [0.900 - 13.615]	0.034	3.6 [1.102 - 11.758]	0.741	1.327 [0.249 - 7.077]	0.015	4.127 [1.319 - 12.915]
DN4 at 12 months (<4/≥4)									0.011	1.698 [1.132 - 2.549]	0.001	1.872 [1.292 - 2.712]
MAS at 3 months	<0.001	2.816 [1.720 - 4.612]	<0.001	2.996 [1.885 - 4.763]	<0.001	2.423 [1.505 - 3.899]	<0.001	3.147 [2.000 - 4.953]	0.006	1.719 [1.166 - 2.536]	<0.001	2.275 [1.610 - 3.214]
MAS at 6 months					<0.001	2.991 [1.664 - 5.376]	<0.001	3.984 [2.263 - 7.015]	0.004	1.928 [1.226 - 3.031]	<0.001	2.472 [1.651 - 3.702]
MAS at 12 months									0.001	2.979 [1.536 −5.777]	<0.001	3.773 [2.026 - 7.025]
Shoulder subluxation at 3 months	0.480	2.762 [0.165 - 46.169]	0.023	11.941 [1.408 - 101.253]	0.445	3 [0.179 - 50.211]	0.016	14 [1.646 - 119.068]	0.006	23.667 [2.433 −230.249]	0.025	12.909 [1.370 - 121.610]
Contractures at 3 months	0.003	5.127 [1.767 - 14.880]	<0.001	4.947 [2.014 - 12.153]	0.461	1.500 [0.510 - 4.413]	0.002	3.938 [1.642 - 9.439]	0.806	1.159 [0.357 - 3.760]	0.004	4.080 [1.587 - 10.489]
Contractures at 6 months					0.522	1.446 [0.468 - 4.470]	0.003	3.816 [1.567 - 9.297]	0.255	2.028 [0.601 - 6.843]	<0.001	7.138 [2.645 - 19.268]
Contractures at 12 months									0.409	1.846 [0.431 - 7.908]	0.010	4.235 [1.421 - 12.625]
Falls at 3 months	0.014	3.6 [1.295 - 10.010]	0.028	2.591 [1.110 - 6.047]	0.440	1.504 [0.533 - 4.245]	0.028	2.580 [1.107 - 6.015]	0.805	1.152 [0.375 - 3.537]	0.283	1.646 [0.662 - 4.089]
Falls at 6 months					0.410	1.9 [0.413 - 8.742]	0.001	6.939 [2.241 - 21.483]	0.499	1.641 [0.390 - 6.900]	0.015	3.765 [1.295 - 10.949]
Falls at 12 months									0.119	5.071 [0.658 - 39.082]	0.014	8.452 [1.528 - 46.760]
Pressure ulcers at 3 months	<0.001	13.2 [3.812 - 45.702]	<0.001	15.529 [5.136 - 46.953]	0.025	3.624 [1.172 - 11.209]	<0.001	7.729 [2.958 - 20.194]	0.006	5.083 [1.595 - 16.201]	<0.001	5.931 [2.206 - 15.940]
Confirmed pneumonia at 3 months	0.019	5.250 [1.318 - 20.905]	0.002	6.714 [2.008 - 22.630]	0.106	3.412 [0.771 - 15.102]	<0.001	9.227 [2.739 - 31.091]	0.01	6.182 [1.534 - 24.910]	0.003	6.4 [1.857 - 22.056]
Confirmed pneumonia at 6 months					0.266	3.158 [0.416 - 23.966]	0.017	7.241 [1.415 - 37.062]	0.493	2.367 [0.201 - 27.819]	0.034	6.762 [1.157 - 39.533]

Bold numbers represent significant association with p-value £0.05

HADS_A Hospital Anxiety and Depression Scale_ Anxiety score, PSA Post-stroke Anxiety, UOR Unadjusted Odds Ratio, CI Confidence Interval, MMSE Mini-Mental State Examination, FSS Fatigue Severity Scale, VAS Visual Analogue Scale, DN4 Douleur Neuropathique 4, MAS Modified Ashworth Scale

The multivariable analysis findings were displayed in Table 5. Only pressure ulcers emerged as significantly independent predictor of PSA at 3 months post-stroke by stepwise multivariable analysis (AOR = 4.7, 95% CI 1.02–21.2). However, PSA at 6 months was predicted by lower PCS scores at 3 months (AOR = 0.9 by increasing 1 point, 95% CI 0.8–0.98), higher mRS scores at 3 months (AOR = 2.2, 95% CI 1.2–4.04), and higher FSS scores at 6 months (AOR = 6.7, 95% CI 2.7–16.8). 12-month PSA independent predictors were AF (AOR = 5.3, 95% CI 1.001–28.1), higher FSS scores at 12 months (AOR = 2.7, 95% CI 1.3–5.8), and higher DN4 scores at 3 months (AOR = 6.9, 95% CI 1.1–44.8).

Table 5.

Independent predictors of borderline cases of anxiety (8 ≤ HADS_A ≤ 10) and post-stroke anxiety (PSA) (HADS_A ≥ 11) at 3, 6, and 12 months using multivariable analysis

Independent predictors	Borderline cases		PSA
	p-value	AOR [95%CI]	p-value	AOR [95% CI]
At 3 months post-stroke
History of AF	0.015	6.074 [1.422—25.938]	0.151	2.317 [0.735—7.301]
Recurrent stroke	0.031	10.768 [1.242—93.320]	0.659	0.551 [0.039—7.784]
Falls	0.014	8.809 [1.546—50.183]	0.496	1.490 [0.473—4.697]
Pressure ulcers	0.002	26.590 [3.240—218.216]	0.046	4.669 [1.026—21.248]
At 6 months post-stroke
PCS of SF12 at 3 months	0.026	0.893 [0.809—0.987]	0.025	0.877 [0.781—0.984]
mRS at 3 months	0.035	1.966 [1.050—3.680]	0.011	2.204 [1.202—4.040]
FSS at 6 months	0.015	3.025 [1.244—7.353]	< 0.001	6.728 [2.693—16.807]
At 12 months post-stroke
History of AF	0.258	2.519 [0.508—12.484]	0.050	5.3 [1.001—28.051]
FSS at 12 months	0.021	2.224 [1.129—4.382]	0.011	2.694 [1.254—5.791]
DN4 at 3 months	0.552	1.693 [0.299—9.585]	0.044	6.870 [1.053—44.824]

Bold numbers represent significant association with p-value ≤ 0.05

HADS_A Hospital Anxiety and Depression Scale_ Anxiety score, PSA Post-stroke Anxiety, AOR Adjusted Odds Ratio, CI Confidence Interval, AF Atrial Fibrillation, PCS of SF12 physical component summary of Short Form Health Survey 12, mRS modified Rankin Sale, FSS Fatigue Severity Scale, DN4 Douleur Neuropathique 4

The risk factors that were significantly positively associated with PSD within 1-year post-stroke were subjects without any profession/retired and unemployed (UOR ranging between 4 and 7, UOR between 6 and 23, respectively), higher NIHSS (UOR ≅ 2), mRS (UOR ≅ 5), HADS_A scores (UOR ≅ 2), FSS scores (UOR between 3 to 6), VAS scores (UOR ≅ 2), DN4 scores (UOR ≅ 3), and MAS scores (UOR between 3 to 8), contractures (UOR between 3 and 13), and pressure ulcers at 3- and 6 months (UOR between 7 and 22). Whereas less sedentary hours (< 7 h/day) (UOR between 0.01 and 0.2), higher PCS and MCS component scores of SF12 QoL (UOR = 0.8 each component), and higher MMSE scores (UOR = 0.1) were negatively associated with 1-year PSD. Similarly, higher educational level was a protective factor against PSD at 6- and 12 months post-stroke (UOR = 0.4). On the other hand, prolonged hospital stay increased the risk of PSD at 3- and 6 months post-stroke (UOR ≅ 2). Moderate SSRS and former smoking had inversely impacted 3-month PSD (UOR = 0.3 and UOR = 6, respectively). Similar opposite relation was demonstrated between presence of health insurance and presence of a caregiver with the risk of borderline depression case and PSD at 6 months, respectively (health insurance: UOR = 0.08 and caregiver: UOR = 2.7) (Tables 6, 7, and 8) (Tables 4S, 5S, and 6S in the Additional file 1 represent the factors in numbers, percentages and means).

Table 6.

Baseline factors associated with borderline case of depression (8≤HADS_D≤10) and post-stroke depression (PSD) (HADS_D≥11) at 3, 6, and 12 months using multinomial logistic regression analysis

Baseline characteristics	3 months post-stroke				6 months post-stroke				12 months post-stroke
	Borderline case		PSD		Borderline case		PSD		Borderline case		PSD
	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]
Age, mean (years)	0.920	1.003 [0.953 - 1.055]	0.988	1 [0.962 - 1.039]	0.208	1.033 [0.982 - 1.088]	0.614	1.010 [0.973 - 1.048]	0.868	0.996 [0.949 - 1.045]	0.337	1.017 [0.982 - 1.053]
Gender, Female	0.517	1.473 [0.456- 4.755]	0.784	1.135 [0.460 - 2.797]	0.964	0.974 [0.312 - 3.044]	0.925	1.043 [0.436 - 2.491]	0.791	1.167 [0.374 - 3.642]	0.901	0.950 [0.424 - 2.128]
Educational level, Higher education	0.843	0.889 [0.277 - 2.854]	0.154	0.519 [0.210 - 1.278]	0.103	0.375 [0.115 - 1.220]	0.024	0.364 [0.151 - 0.878]	0.575	1.381 [0.446 - 4.273]	0.040	0.4 [0.167 - 0.959]
Professional status post stroke
Person without any profession/retired	0.743	1.250 [0.330 - 4.740]	0.001	8.125 [2.241 - 29.452]	0.411	1.779 [0.451 - 7.023]	0.002	7.765 [2.178 - 27.678]	0.808	1.2 [0.276 - 5.209]	0.029	4.533 [1.167 - 17.611]
Unemployed	0.108	5 [0.704 - 35.495]	0.001	23.750 [3.689 - 152.885]	0.063	5.5 [0.912 - 33.184]	0.002	14.667 [2.727 - 78.877]	0.436	2 [0.350 - 11.439]	0.023	6 [1.274 - 28.254]
Employed	Ref		Ref		Ref		Ref		Ref		Ref
Health insurance					0.027	0.083 [0.009 - 0.757]	0.065	0.141 [0.018 - 1.131]	0.140	0.333 [0.077 - 1.437]	0.356	0.569 [0.172 - 1.881]
Presence of a caregiver	0.685	0.750 [0.187 - 3.011]	0.083	2.357 [0.894 - 6.214]	0.760	0.807 [0.204 - 3.196]	0.044	2.707 [1.028 - 7.130]	0.096	2.758 [0.836 - 9.092]	0.072	2.245 [0.929 - 5.424]
SSRS, Moderate	0.660	1.5 [0.247 - 9.111]	0.031	0.280 [0.088 - 0.890]	0.785	0.817 [0.192 - 3.484]	0.076	0.376 [0.128 - 1.109]	0.913	1.083 [0.258 - 4.545]	0.104	0.471 [0.190 - 1.168]
Sedentary duration
1 to 6 hours/day	0.053	0.097 [0.009 - 1.028]	<0.001	0.013 [0.002 - 0.112]	0.019	0.065 [0.007 - 0.640]	<0.001	0.015 [0.002 - 0.129]	0.018	0.162 [0.036 - 0.730]	<0.001	0.057 [0.015 - 0.213]
7 to 11 hours/day	0.427	0.375 [0.033 - 4.228]	0.050	0.114 [0.013 - 1.004]	0.132	0.167 [0.016 - 1.718]	0.031	0.095 [0.011 - 0.807]	0.022	0.159 [0.033 - 0.769]	0.010	0.220 [0.069 - 0.701]
≥ 12 hours/day	Ref		Ref		Ref		Ref		Ref		Ref
Smoking status
Never smoker	0.591	1.5 [0.341 - 6.592]	0.580	1.304 [0.509 - 3.343]	0.906	1.083 [0.288 - 4.081]	0.641	1.254 [0.484 - 3.244]	0.515	1.652 [0.364 - 7.503]	0.685	0.826 [0.329 - 2.076]
Ex-smoker	0.004	16.250 [2.462 - 107.242]	0.029	5.935 [1.196 - 29.452]	0.062	4.333 [0.928 - 20.244]	0.071	3.250 [0.903 - 11.696]	0.058	4.750 [0.946 - 23.845]	0.090	2.523 [0.864 - 7.369]
Current smoker	Ref		Ref		Ref		Ref		Ref		Ref
History of AF	1.000	1 [0.266 - 3.763]	0.243	1.787 [0.674 - 4.738]	0.393	1.714 [0.497 - 5.911]	0.166	1.993 [0.751 - 5.288]	0.803	1.182 [0.317 - 4.400]	0.010	3.136 [1.310 - 7.510]

Bold numbers represent significant association with p-value £0.05

HADS_D Hospital Anxiety and Depression Scale_ Depression score, PSD Post-stroke Depression, UOR Unadjusted Odds Ratio, CI Confidence Interval, AF Atrial Fibrillation

Table 7.

Stroke-related factors associated with borderline case of depression (8≤HADS_D≤10) and post-stroke depression (PSD) (HADS_D≥11) at 3, 6, and 12 months using multinomial logistic regression analysis

Stroke-related factors	3 months post-stroke				6 months post-stroke				12 months post-stroke
	Borderline case		PSD		Borderline case		PSD		Borderline case		PSD
	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]
Duration of hospital stay	0.019	1.191 [1.029 - 1.378]	0.009	1.204 [1.048 - 1.383]	0.007	1.209 [1.054 - 1.387]	0.008	1.196 [1.048 - 1.364]	0.855	1.009 [0.920 - 1.106]	0.119	1.048 [0.988 - 1.112]
NIHSS at 3 months	0.003	1.232 [1.074 - 1.412]	<0.001	1.272 [1.117 - 1.448]	0.012	1.145 [1.031 - 1.272]	<0.001	1.196 [1.087 - 1.317]	0.018	1.107 [1.018 - 1.204]	<0.001	1.163 [1.086 - 1.245]
NIHSS at 6 months					0.008	1.256 [1.063 - 1.484]	<0.001	1.331 [1.135 - 1.561]	0.003	1.188 [1.062 - 1.328]	<0.001	1.249 [1.132 - 1.378]
NIHSS at 12 months									0.005	1.360 [1.099 - 1.683]	<0.001	1.468 [1.199 - 1.797]
mRS at 3 months	0.002	2.620 [1.420 - 4.835]	<0.001	5.264 [2.875 - 9.635]	0.001	2.783 [1.544 - 5.017]	<0.001	5.051 [2.822 - 9.039]	0.009	1.909 [1.176 - 3.099]	<0.001	3.199 [2.076 - 4.930]
mRS at 6 months					0.005	2.819 [1.365 - 5.819]	<0.001	5.934 [2.930 - 12.020]	0.002	2.280 [1.343 - 3.879]	<0.001	3.739 [2.335 - 5.987]
mRS at 12 months									<0.001	3.433 [1.791 - 6.582]	<0.001	5.386 [2.914 - 9.954]
SF12 of QoL
PCS at 3 months	0.063	0.928 [0.858 - 1.004]	<0.001	0.836 [0.777 - 0.900]	0.018	0.908 [0.839 - 0.984]	<0.001	0.863 [0.807 - 0.924]	0.057	0.920 [0.844 - 1.003]	<0.001	0.864 [0.804 - 0.928]
MCS at 3 months	0.572	0.977 [0.902 - 1.059]	<0.001	0.831 [0.773 - 0.895]	0.286	0.960 [0.891 - 1.034]	<0.001	0.844 [0.788 - 0.904]	0.116	0.946 [0.884 - 1.014]	<0.001	0.872 [0.824 - 0.923]
PCS at 6 months					0.075	0.939 [0.876 - 1.006]	<0.001	0.813 [0.755 - 0.876]	0.025	0.922 [0.859 - 0.990]	<0.001	0.853 [0.800 - 0.909]
MCS at 6 months					0.052	0.934 [0.872 - 1.001]	<0.001	0.840 [0.784 - 0.900]	0.016	0.922 [0.863 - 0.985]	<0.001	0.871 [0.823 - 0.921]
PCS at 12 months									<0.001	0.861 [0.800 - 0.926]	<0.001	0.807 [0.752 - 0.867]
MCS at 12 months									<0.001	0.888 [0.832 - 0.947]	<0.001	0.844 [0.795 - 0.897]

Bold numbers represent significant association with p-value £0.05

HADS_D Hospital Anxiety and Depression Scale_ Depression score, PSD Post-stroke Depression, UOR Unadjusted Odds Ratio, CI Confidence Interval, NIHSS National Institutes of Health Stroke Scale, mRS modified Rankin Scale, SF12 of QoL Short Form Health Survey of Quality of Life, PCS physical component summary, MCS mental component summary

Table 8.

Post-stroke complications associated with borderline case of depression (8≤HADS_D≤10) and post-stroke depression (PSD) (HADS_D≥11) at 3, 6, and 12 months using multinomial logistic regression analysis

Post-stroke complications	3 months post-stroke				6 months post-stroke				12 months post-stroke
	Borderline case		PSD		Borderline case		PSD		Borderline case		PSD
	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]	p-value	UOR [95% CI]
HADS_A at 3 months	<0.001	1.490 [1.197 - 1.856]	<0.001	1.661 [1.359 - 2.032]	0.023	1.234 [1.030 - 1.479]	<0.001	1.499 [1.271 - 1.768]	0.097	1.144 [0.976 - 1.342]	<0.001	1.336 [1.175 - 1.519]
HADS_A at 6 months					0.002	1.360 [1.119 - 1.653]	<0.001	1.594 [1.327 - 1.915]	0.016	1.210 [1.036 - 1.414]	<0.001	1.396 [1.226 - 1.590]
HADS_A at 12 months									<0.001	1.765 [1.325 - 2.351]	<0.001	2.258 [1.679 - 3.037]
MMSE at 3 months (≤ 23/>23)	1.000	1 [0.317 - 3.151]	<0.001	0.103 [0.035 - 0.301]	0.695	0.800 [0.262 - 2.440]	<0.001	0.122 [0.043 - 0.351]	0.859	0.900 [0.282 - 2.870]	0.001	0.133 [0.041 - 0.429]
MMSE at 6 months					0.086	0.296 [0.074 - 1.187]	<0.001	0.076 [0.024 - 0.243]	0.455	0.621 [0.178 - 2.168]	<0.001	0.102 [0.041 - 0.256]
MMSE at 12 months									0.216	0.409 [0.099 - 1.687]	<0.001	0.072 [0.026 - 0.204]
FSS at 3 months	0.076	1.511 [0.957 - 2.387]	<0.001	4.045 [2.400 - 6.818]	0.016	1.778 [1.114 - 2.837]	<0.001	3.177 [1.995 - 5.060]	0.096	1.494 [0.931 - 2.396]	<0.001	2.913 [1.827 - 4.643]
FSS at 6 months					0.001	2.596 [1.444 - 4.666]	<0.001	5.773 [3.094 - 10.775]	0.001	2.591 [1.477 - 4.546]	<0.001	4.190 [2.497 - 7.031]
FSS at 12 months									<0.001	3.056 [1.803 - 5.182]	<0.001	5.492 [3.081 - 9.790]
VAS at 3 months	0.025	1.269 [1.031 - 1.563]	<0.001	1.537 [1.284 - 1.840]	0.188	1.136 [0.940 - 1.373]	<0.001	1.424 [1.212 - 1.673]	0.115	1.167 [0.963 - 1.415]	<0.001	1.483 [1.260 - 1.744]
VAS at 6 months					0.044	1.317 [1.008 - 1.721]	<0.001	1.792 [1.410 - 2.278]	0.141	1.207 [0.940 - 1.550]	<0.001	1.795 [1.444 - 2.231]
VAS at 12 months									0.004	1.727 [1.188 - 2.509]	<0.001	2.534 [1.796 - 3.575]
DN4 at 3 months	<0.001	2.478 [1.542 - 3.981]	<0.001	2.227 [1.435 - 3.456]	0.003	1.658 [1.187 - 2.316]	0.001	1.628 [1.214 - 2.183]	0.213	1.196 [0.902 - 1.587]	0.005	1.343 [1.092 - 1.651]
DN4 at 6 months					0.010	1.797 [1.149 - 2.811]	0.002	1.878 [1.259 - 2.804]	0.204	1.274 [0.877 - 1.851]	0.002	1.553 [1.177 - 2.048]
DN4 at 12 months									0.042	2.006 [1.027 - 3.921]	<0.001	3.015 [1.681 - 5.407]
MAS at 3 months	0.005	3.282 [1.419 - 7.590]	0.001	3.710 [1.664 - 8.271]	0.012	2.198 [1.192 - 4.053]	0.001	2.546 [1.447 - 4.479]	0.009	1.839 [1.165 - 2.901]	<0.001	2.102 [1.443 - 3.062]
MAS at 6 months					0.008	2.618 [1.289 - 5.314]	0.002	2.877 [1.489 - 5.558]	0.007	2.123 [1.232 - 3.658]	<0.001	2.412 [1.521 - 3.823]
MAS at 12 months									0.018	3.995 [1.274 - 12.529]	<0.001	8.410 [2.964 - 23.867]
Contractures at 3 months	0.025	7 [1.270 - 38.574]	0.001	12.649 [2.795 - 57.236]	0.246	2.229 [0.576 - 8.623]	0.004	4.754 [1.630 - 13.867]	0.834	1.152 [0.309 - 4.292]	0.005	3.442 [1.455 - 8.145]
Contractures at 6 months					0.162	3.625 [0.597 - 22.013]	0.001	13.257 [2.926 - 30.069]	0.178	2.667 [2.635 - 20.409]	<0.001	7.333 [2.635 - 20.409]
Contractures at 12 months									0.142	3.615 [0.651 - 20.072]	0.006	6.451 [1.719 - 24.215]
Falls at 3 months	1.000	1 [0.266 - 3.763]	0.065	2.488 [0.944 - 6.554]	0.393	1.714 [0.497 - 5.911]	0.033	2.873 [1.091 - 7.561]	0.243	2 [0.624 - 6.410]	0.083	2.095 [0.909 - 4.831]
Falls at 6 months									0.024	8 [1.307 - 48.953]	0.003	10.588 [2.271 - 49.362]
Falls at 12 months									0.043	11.308 [1.085 - 117.896]	0.124	5.568 [0.626 - 49.519]
Pressure ulcers at 3 months		0.030	11.571 [1.265 - 105.823]	0.003					0.008	10.909 [1.866 - 63.781]	<0.001	18 [3.923 - 82.584]
Pressure ulcers at 6 months									0.016	16.333 [1.670 - 159.754]	0.006	18.2 [2.274 - 145.636]

Bold numbers represent significant association with p-value £0.05

HADS_D Hospital Anxiety and Depression Scale_ Depression score, PSD Post-stroke Depression, UOR Unadjusted Odds Ratio, CI Confidence Interval, HADS_A Hospital Anxiety and Depression Scale_ Anxiety Score, MMSE Mini-Mental State Examination, FSS Fatigue Severity Scale, VAS Visual Analogue Scale, DN4 Douleur Neuropathique 4, MAS Modified Ashworth Scale

In the multivariable analysis (Table 9), our findings showed that increasing HADS_A scores was the most consistent independent predictor of PSD within 1-year post-stroke (AOR = 1.7, 95%CI 1.2–2.5 at 3 months; AOR = 1.5, 95%CI 1.1–1.9 at 6 months; and AOR = 1.6, 95%CI 1.1–2.3).

Table 9.

Independent predictors of borderline cases of depression (8≤HADS_D≤10) and post-stroke depression (PSD) (HADS_D≥11) at 3, 6, and 12 months using multivariable analysis

Independent predictors	Borderline case		PSD
	p-value	AOR [95%CI]	p-value	AOR [95% CI]
At 3 months post-stroke
mRS at 3 months	0.007	10.364 [1.901 - 56.503]	0.028	4.784 [1.186 - 19.296]
PCS of SF12 at 3 months	0.040	0.809 [0.662 - 0.990]	0.029	0.803 [0.659 - 0.978]
DN4 at 3 months	0.010	3.416 [1.341 - 8.704]	0.009	3.246 [1.334 - 7.896]
Sedentary behavior	0.599	1.696 [0.237 - 12.140]	0.028	6.920 [1.234 - 38.823]
HADS_A at 3 months	0.097	1.380 [0.943 - 2.020]	0.003	1.744 [1.211 - 2.510]
At 6 months post-stroke
Resume of work post-stroke	0.239	0.528 [0.182 - 1.528]	0.011	0.199 [0.057 - 0.691]
mRS at 6 months	0.513	1.448 [0.478 - 4.385]	0.041	3.084 [1.050 - 9.061]
HADS_A at 6 months	0.189	1.215 [0.909 - 1.625]	0.014	1.461 [1.080 - 1.977]
FSS at 6 months	0.123	1.984 [0.831 - 4.737]	0.033	2.632 [1.079 - 6.420]
At 12 months post-stroke
History of AF	0.853	0.812 [0.089 - 7.398]	0.034	15.217 [1.232 - 187.976]
Resume of work post-stroke	0.562	1.502 [0.379 - 5.957]	0.047	0.107 [0.012 - 0.973]
HADS_A at 12 months	0.155	1.275 [0.912 - 1.781]	0.012	1.583 [1.106 - 2.266]

Bold numbers represent significant association with p-value £0.05

HADS_D Hospital Anxiety and Depression Scale_ Depression score, PSD Post-stroke Depression, AOR Adjusted Odds Ratio, CI Confidence Interval, mRS modified Rankin Sale, PCS of SF12 physical component summary of Short Form Health Survey 12, DN4 Douleur Neuropathique 4, HADS_A Hospital Anxiety and Depression Scale_ Anxiety Score, FSS Fatigue Severity Scale, AF Atrial Fibrillation 

Other factors have been found to be independently predictive of higher PSD risk at 3 months post-stroke as follows: prolonged sedentary hours (≥ 12 h/day) (AOR = 6.9, 95%CI 1.2–38.8), higher mRS scores (AOR = 4.8, 95%CI 1.2–19.3), higher DN4 scores (AOR = 3.2, 95%CI 1.3–7.9), and lower PCS scores (AOR = 0.8 with increasing 1 point, 95%CI 0.7–0.9). Furthermore, higher mRS scores and FSS scores at 6 months were independently determinants of 6-month PSD (AOR = 3.1, 95%CI 1.1–9.1 and AOR = 2.6, 95%CI 1.1–6.4, respectively). On the other hand, employment post-stroke was a protective factor against PSD at 6- and 12 months after stroke (AOR = 0.2, 95%CI 0.06–0.7 and AOR = 0.1, 95%CI 0.01–0.97, respectively).

Discussion

Anxiety and depression are common and major consequences after stroke. The present study is the first in Lebanon estimating a higher risk of PSA and PSD in stroke survivors, considering depressive and anxiety symptoms mainly occurring in the early stage after the acute event, in the first 3 months (51.2% for PSA and 77.3% for PSD). These proportions slighlty decreased in the following 6 to 12 months but remained elevated accounting for at least 40% of the survivors for PSA and 60% for PSD. This was consistent with the rates obtained in Jordan and Morocco ranging between 55 and 75% [9, 48] but higher than those reported in France, Taiwan, USA, and Brazil varying between 11 to 25% for PSA and 20% to 40% for PSD [49–52]. The first possible explanation could be the different used scales for assessment of anxiety and depression. Second, study designs, sample sizes, stroke types, and geographic variation could affect these rates. Third, our study population were patients living in their homes and care was provided by family members who are usually unprepared to face significant challenges residual from stroke, worsened by the lack of financial resources in Lebanon (poverty, food insecurity, low educational level, lack of access to care, lack of transportation and medication affordability) [5, 9, 53–55].

Characterizing the control of risk factors is very essential for which treatment is of proven benefit [54]. After thorough consideration based on preceding research, we could investigate the factors contributing to these PSA and PSD high rates.

Our findings showed that subjects with functional and immobility-related problems were more vulnerable to develop anxiety and depression at any time-point within the first year following stroke. Subjects with PSA and PSD within the first year of stroke were more likely to be dependent in ADL. Higher mRS scores (mRS > 2), especially in the acute phase, were recorded in 60% to 90% of both, PSA and PSD patients. Lower PCS component scores of QoL within 3 months post-stroke were significant independent predictors of PSD developed at 3 months and PSA at 6 months. Similar to the findings of the present study, previous literature found that PSD and PSA were associated with poorer physical functioning at early phase and greater decline within 1-year post-stroke [2, 4–6, 56]. Unmet needs of those patients for physical recuperation, Activities of Daily Living (ADL), mobility, pain control, and communication remain predominant and depend on the temporal phase of their illness, the cause and severity of their stroke. Therefore, post-stroke patient care is an iterative process of assessment and management that begins in the hospital and continues in the community, where recovery, reintegration, and health maintenance take place over the following years according to the patient’s changing needs [54].

Moreover, Post-Stroke Fatigue (PSF) at 6 months post-stroke increased 7 folds the risk of PSA and 3 folds the risk of PSD at 6 months. In addition, PSF at 12 months increased 3 times the risk of 12-month PSA. Survivors with PSA and PSD were experiencing pain, notably Central Post-Stroke Pain (CPSP). Neuropathic pain in the early phase was an independent predictive factor to the risk of PSD at 3 months and PSA at 6 months. Patients with persistent fatigue and pain, specifically CPSP, may negatively impact ADL and may be more prone to PSA and PSD [2, 57–61]. Fatigue is an overwhelming sense of tiredness or exhaustion which may start in the first few days post-stroke and last for years. It has been labeled as an “invisible” deficit and a time-dependent factor having a significant impact on functioning. It may predispose individuals to other complications that are hard to handle, like falls and fractures, affecting balance and gait control, as well as mobility, ADL, and further anxiety and depression [59]. Likewise, pain, including CPSP, is a frequent and neglected complication that can happen immediately, weeks, or months afer a stroke event, resulting in fatigue, sleep disorders, memory impairment, anxiety, depression, low self-perceived health status, increased cognitive impairment, functional dependence, poorer recovery post-stroke, and suicidal ideation [57, 58, 62, 63, 63, 64].

Anxiety was the most consistent independent predictor of depression at 3-, 6-, and 12 months after stroke. This was consistent with previous studies [4–6, 65, 66]. Researchers suggested that shared genetic and biological etiologies are the reason behind this relationship [66, 67]. Camargos et al. in 2020 reported that secretion of pro-inflammatory mediators, cytotoxic substances, cytokines and neurotrophic factors and changes in their levels have been related to the pathophysiology mechanisms of both depression and anxiety after stroke [67].

AF was another independent predictor of long-term PSA and PSD (12 months post-stroke). Although its relation in our study did not demonstrate great precision, it was shown correlated with higher risk of anxiety and depression post-stroke [68]. Prolonged hospital stay and pressures ulcers were also independently risk factors for PSD and PSA at 3 months, respectively. Stroke patients with initial severe status at stroke onset have longer length of stay and they are less likely to be discharged to an independent or assisted living situation, leading to poorer functional and mental outcomes in the rehabilitation phase [69]. The application of bundled nursing management effectively promotes the cure rate of pressure ulcers and improve all negative emotions, including anxiety, QoL, sleep quality, and patient satisfaction [70].

Other important associated determinants of PSA and PSD of which effect was lost in the multivariable analysis are worth to mention. Anxious and depressed survivors had elevated NIHSS scores, VAS scores, and MAS scores, and lower MMSE scores. They were more exposed to contractures, falls, and longer sedentary sitting hours (≥ 12 h). Previous literature has supported these associations [2–6]. Furthermore, spasticity, contractures, and falls lead to motor impairment, gait abnormalities, pain, difficulties in performing ADL, such as feeding, dressing, and bathing, causing significant loss of QoL and self-perceived health status, increased cognitive impairment, and poorer recovery post-stroke. Stroke survivors become less likely to regain independent mobility leading to lower QoL and psychosocial issues [61–64, 71]. Stroke survivors tend to spend more time sedentary compared to stroke-free people. The brain damage affects their mental and functional status, they become less energetic, less active, less concentrated and more prone to social isolation, anxiety and depression [72–75]. The majority of stroke survivors in the present study were illiterate/low educational level (70%) and unemployed/without any profession/retired (87%) post-stroke predicting higher rates of PSA and PSD. Employment post-stroke was negatively correlated with PSA and PSD in our study population as well as in other studies [76–78]. Regarding educational level, three recent reviews have found no clear association between educational level and development of depression after stroke, and so far, no studies have proven an association with anxiety after stroke [4–6, 79]. However, higher educational level in our study was a protective factor against PSA at 12 months and PSD at 6- and 12 months post-stroke. Despite the unclear association mentioned by these three reviews, some papers have supported our findings revealing the impact of educational level on PSD development [80–83]. The relation between educational level and PSA could be related to the effect of PSD since the majority of our subjects with PSA had also PSD. Hence, these associations need further studies with larger cohorts that might identify the predictors of PSA and PSD separately.

Our study was the first hospital-based study in Lebanon investigating the rates of post-first-ever stroke complications and their changes over time as well as identifying their risk factors. The present study was a multicenter longitudinal prospective study which may have decreased recall and selection bias. The major strength of our study lay in the use of repeated international standardized, validated, and reliable measuring instruments, effectively mitigating numerous common biases inherent in trials and hospital-based studies. We took measures to circumvent information bias by employing the available Arabic versions of these instruments. Administered by highly qualified and extensively trained investigators through face-to-face interviews with participants, these instruments markedly reduced the potential for bias typically associated with self-administered questionnaires, stemming from potential misunderstandings of the questions.

Furthermore, our study's robustness was bolstered by the exclusion of patients with a history of previous stroke or TIA, those admitted with a diagnosis of TIA or subsequent stroke, or those with pre-existing neuro-psychiatric and cognitive disorders (such as dementia, Alzheimer's disease, ataxia, Bell's palsy, brain tumor, cerebral aneurysm, epilepsy, seizures, Parkinson's disease, meningitis, hydrocephalus, encephalitis, aphasia, cerebrovascular accident, attention disorders, anxiety, depression, and other mood disorders). Inclusion of individuals with these conditions could have introduced significant confounding factors that might have influenced our study's outcomes. It is worth mentioning that unlike to some previous literature, we did not exclude patients with post-stroke language or cognitive disorders in the analysis of physical and neuro-psychological complications as we believe that many other previous papers, cited in this manuscript, highlighted their main influence. We believe that their exclusion might be a selection bias and underestimation of other studied variables and estimates may not be generalizable to all stroke survivors [4, 6, 84]. We overcame communication and cognitive barriers using by-proxy assessment methods and observed behaviors when necessary to evaluate patients.

Nevertheless, several limitations require consideration. First, the main limitation was the small sample size recruited following the unique study considering a low prevalence of stroke in Lebanon of 3.9% according to other countries [22]. Second, as Lebanon is a small, highly interconnected country, hospitals serve poorly defined catchment areas. This is especially true for patients with complicated illnesses like Stroke where it is not uncommon for patients to travel the relatively short distances (2–3 h by car) to receive care in a different governorate. Therefore, we cannot claim that each hospital serves a specific defined population. According to published numbers from 2018–2019 by the Central Administration of Statistics in Lebanon [85], Mount Lebanon is the largest governorate in the country, hosting 2,032,600/4,842,500 (42.0%) of residents. Although the capital of Lebanon, Beirut, claims only 7.1% of residents (341,700/4,842,500), however it has the largest catchment area, around 2,000,000 due to the presence of major tertiary care centers. Despite this, our study could not be representative of the overall population of Lebanon. Third, our study uniquely encompasses a comprehensive range of post-stroke sequelae, thus introducing a novel dimension to the research landscape. However, the longitudinal nature of our study, spanning 12 months, led to the repetition of certain measures, raising questions regarding potential learning effects [86]. Moreover, it's essential to acknowledge the confounding influence of spontaneous recovery observed in individuals who have experienced neurological events [87]. Fourth and lastly, despite achieving statistically significant results, we encountered a loss to follow-up due to mortality within three months post-stroke. This circumstance prevented the assessment of cognitive, neuropsychological, and motor functions in these individuals, which were strongly suspected to be impaired. Consequently, we posit that our study might underestimate the extent of impairment in these functions. In addition, our studied outcomes, such as seizures, cognitive impairment, psychological disorders, and physical deficits (general pain, shoulder pain and subluxation, headaches, limb pain, spasticity, contractures, fatigue) were based on patient-reported standard valid measures and pre-specified definitions conducted by the study’s investigators and were not then clinically confirmed by a documented official medical report, the fact that may affect the estimated rates. However, due to differences in cultural factors, health systems and resource availability in each country, we believe that the most accurate measures that should be followed for both unified clinical decision-making and prognosis are the patient’s reported outcome measures [88].

Therefore, this study could function as a preliminary study among Lebanese survivors. Moreover, conclusions should be confirmed in a larger cohort. Our findings may be useful to draw hypotheses on stroke burden and its heavy physical, cognitive, and neuropsychological implications for further analyses. Future studies should consider the heterogeneity between studies and all the weak points in our study as well as a larger sample size across Lebanon must be considered to confirm our findings and cover the unaddressed factors.

Conclusion

In consequence, anxiety and depression are multifactorial complications post-stroke. Although their serious impact on stroke survivors, they are often neglected and ignored. Anxiety and depression are known to be associated with worsened functional outcomes, lower QoL, and more pronounced cognitive deficits, at any time following stroke, especially within the first year. Despite these recognized impacts, stroke survivors at risk of psychological disorders are rarely routinely screened for anxiety or depression or formally assessed by neuropsychologists [1]. While international guidelines vary on who, when, and how screening for PSA and PSD should take place, stroke patients may benefit from extensive regular psychological screening and support from clinicians and service providers after stroke, pondering the various contributing factors [89]. The acute phase is the mainstay for a successful secondary prevention that requires close monitoring, early recognition of potential challenges, and appropriate timely-dependent evidence-based interventions [90, 91]. Although the benefit of non-invasive psychosocial therapies has not yet been proven in homogenous randomized clinical trials [4], a new study, published in January 2023 [89], endorsed the application of self-management programs (having the confidence to deal with medical management, role management and emotional management of their conditions) which have shown promising results in improving patients’ Health-Related QoL and survival [92, 93].

Abbreviations

PSA

Post-Stroke Anxiety

PSD

Post-Stroke Depression

QoL

Quality of Life

AHR

Adjusted Hazard Ratio

CI

Confidence Interval

ICD

International Classification of Diseases

DSM

Diagnostic and Statistical Manual of Mental Disorders

MENA

Middle East and North Africa

FES

First-Ever Stroke

HAM-D

Hamilton Depression Rating Scale

DePreS

Depression Prediction Scale

BDI

Beck Depression Inventory

MADRS

Montgomery Åsberg Depression Rating Scale

HADS_D

Hospital Anxiety and Depression Scale_Depression

NIHSS

National Institutes of Health Stroke Scale

SF12

Short Form Health Survey 12

mRS

Modified Rankin Scale

TIA

Transient Ischemic Attack

WHO

World Health Organization

TOAST

Trial of Org 10,172 in the Acute Stroke Treatment

LAA

Large-Artery Atherosclerosis

CE

Cardioembolism

SVO

Small-Vessel Occlusion

SOE

Stroke of Other determined Etiology

SUE

Stroke of Undetermined Etiology

QVSFS

Questionnaire for Verifying Stroke-Free Status

HCI

Household Crowding Index

HBP

High Blood Pressure

AF

Atrial Fibrillation

DM

Diabetes Mellitus

BMI

Body Mass Index

NHANES-FFQ

National Health and Nutrition Examination Survey Food Frequency Questionnaire

PCS

Physical Component Summary

MCS

Mental Component Summary

SPSS

Statistical Package for Social Sciences

MMSE

Mini-Mental State Examination

HAS

“HauteAutorité de Santé”

VAS

Visual Analogue Scale

DN4

“DouleurNeuopathique”4

MAS

Modified Ashworth Scale

FSS

Fatigue Severity Scale

SSRS

Social Support Rating Scale

UOR

Unadjusted Odds Ratio

AOR

Adjusted Odds Ratio

ADL

Activities of daily living

PSF

Post-Stroke Fatigue

CPSP

Central Post-Stroke Pain

Authors’ contributions

HH, PS, CB contributed to the conceptualization and design of the study. CB, WK, MT, NS contributed to the data collection. CB analyzed and interpreted the data and wrote the first draft of the manuscript. All authors read and approved the final manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate

Our study protocol was reviewed and approved by the ethics committees of two participating hospitals (NEUR-2018–001, HDF-1152), and approvals from other hospitals were granted verbally by their directors. Ethical clearance letter was provided in compliance with the World Medical Association Declaration of Helsinki in 2013 [94]. After obtaining approvals, we contacted eligible participants, we explained all study components (importance, objectives, procedures, risks, benefits, voluntary nature, confidentiality, privacy, and right to withdraw) for a verbal consent after which a written informed consent (or their legal representatives) was necessary to participate in the study. Lastly, to maintain confidentiality, all data were coded in the questionnaire, and the questionnaire papers will be discarded once the legal retention period expired.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.