Association of ischemic stroke onset time with presenting severity, acute progression, and long-term outcome: A cohort study.

BACKGROUND: Preclinical data suggest circadian variation in ischemic stroke progression, with more active cell death and infarct growth in rodent models with inactive phase (daytime) than active phase (nighttime) stroke onset. We aimed to examine the association of stroke onset time with presenting severity, early neurological deterioration (END), and long-term functional outcome in human ischemic stroke. METHODS AND
FINDINGS: In a Korean nationwide multicenter observational cohort study from May 2011 to July 2020, we assessed circadian effects on initial stroke severity (National Institutes of Health Stroke Scale [NIHSS] score at admission), END, and favorable functional outcome (3-month modified Rankin Scale [mRS] score 0 to 2 versus 3 to 6). We included 17,461 consecutive patients with witnessed ischemic stroke within 6 hours of onset. Stroke onset time was divided into 2 groups (day-onset [06:00 to 18:00] versus night-onset [18:00 to 06:00]) and into 6 groups by 4-hour intervals. We used mixed-effects ordered or logistic regression models while accounting for clustering by hospitals. Mean age was 66.9 (SD 13.4) years, and 6,900 (39.5%) were women. END occurred in 2,219 (12.7%) patients. After adjusting for covariates including age, sex, previous stroke, prestroke mRS score, admission NIHSS score, hypertension, diabetes, hyperlipidemia, smoking, atrial fibrillation, prestroke antiplatelet use, prestroke statin use, revascularization, season of stroke onset, and time from onset to hospital arrival, night-onset stroke was more prone to END (adjusted incidence 14.4% versus 12.8%, p = 0.006) and had a lower likelihood of favorable outcome (adjusted odds ratio, 0.88 [95% CI, 0.79 to 0.98]; p = 0.03) compared with day-onset stroke. When stroke onset times were grouped by 4-hour intervals, a monotonic gradient in presenting NIHSS score was noted, rising from a nadir in 06:00 to 10:00 to a peak in 02:00 to 06:00. The 18:00 to 22:00 and 22:00 to 02:00 onset stroke patients were more likely to experience END than the 06:00 to 10:00 onset stroke patients. At 3 months, there was a monotonic gradient in the rate of favorable functional outcome, falling from a peak at 06:00 to 10:00 to a nadir at 22:00 to 02:00. Study limitations include the lack of information on sleep disorders and patient work/activity schedules.
CONCLUSIONS: Night-onset strokes, compared with day-onset strokes, are associated with higher presenting neurologic severity, more frequent END, and worse 3-month functional outcome. These findings suggest that circadian time of onset is an important additional variable for inclusion in epidemiologic natural history studies and in treatment trials of neuroprotective and reperfusion agents for acute ischemic stroke.

Introduction

Despite a well-known morning increase in adverse cardiovascular events such as stroke [1,2] and myocardial infarction [3], there are only a few relatively small studies on the relationship between stroke onset time and short-term/long-term outcomes, with results being inconsistent [4-6]. In addition, these studies often did not account for subtypes of ischemic stroke, a heterogenous disease caused by different pathophysiological mechanisms, and often did not focus solely upon witnessed strokes with confirmed onset timing [7].

A recent preclinical study suggested that circadian rhythms may modulate the extent of brain ischemia and effects of treatment in acute stroke. Neuroprotective treatments reduced infarct growth in day-onset (inactive phase) rodent models of stroke (which corresponds to night-onset stroke in humans), but not in night-onset (active phase) rodent models of stroke (which corresponds to day-onset stroke in humans) [8]. Compared with active phase stroke models, inactive phase stroke models had more active cell death and infarct growth, which is the leading cause of early neurological deterioration (END) in patients with acute ischemic stroke [9,10]. To our knowledge, circadian effect on poststroke END has never been investigated yet.

In this nationwide multicenter study of 17,461 consecutive patients with witnessed ischemic stroke or transient ischemic attack (TIA), we investigated whether stroke onset time correlates with END as well as with initial neurological severity on admission and 3-month functional outcomes. Further, we explored whether these potential associations varied according to different stroke subtypes.

Methods

Study population

This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist). This study was conducted using a prospective multicenter stroke registry: Clinical Research Collaboration for Stroke-Korea (CRCS-K) [11-16]. Using a standardized protocol [17], data were collected from all patients with acute ischemic stroke or TIA who were admitted to 11 academic hospitals within 7 days of symptom onset between May 2011 and July 2020. Inclusion criteria for this study were as follows: (1) witnessed stroke onset; (2) hospital arrival within 6 hours of onset; and (3) agreement to be monitored for the CRCS-K registry-related poststroke outcomes. Among these (n = 60,634), 42,903 were excluded sequentially according to the following reasons: (1) 3,375 patients refused to give research consent to being monitored for stroke outcomes; (2) 16,047 did not have the onset time information (9,922 wake-up stroke, 6,123 unwitnessed stroke, and 2 data missing); and (3) 23,481 arrived to the medical facility >6 hours of onset. The remaining 17,461 patients with witnessed ischemic stroke (n = 14,890) or TIA (n = 2,481) constituted the study cohort. A total of 1,893 (12.8%) patients were lost to follow-up: did not complete hospital visit nor reply to phone call for outcome capture. The institutional review boards of all participating centers (S1 Text) approved the study, and patients or their legally authorized representative(s) provided written informed consent.

Clinical data collection

Using a standardized protocol [17], we collected demographic data, medication history, and details regarding vascular risk factors. Stroke subtypes were determined by consensus among experienced neurologists at each participating center, using a validated magnetic resonance imaging (MRI)-based algorithm [18] built on Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria as follows: large artery atherosclerosis (LAA), small vessel occlusion (SVO), cardioembolism (CE), undetermined, other-determined, and TIA. Admission National Institutes of Health Stroke Scale (NIHSS) score, prestroke modified Rankin Scale (mRS) score, and 3-month mRS score were collected prospectively [13,15].

Assessment of END

Attending neurologists assessed neurological status for each patient daily. Further assessments were done whenever neurological deterioration was noticed by patients themselves, caregivers, or nurses. Physicians and nurses who assessed NIHSS score were trained and certified in a standard manner through a web-based education system (http://www.stroke-edu.or.kr/). In addition, systematic audits, including monthly monitoring and on-site visits to review medical records, were performed by the outcome adjudication committee to assure data quality. There are 5 core members in the adjudication committee: 3 neurologists (one of whom is a coauthor of this paper; TH Park), 2 research nurses, and 1 research coordinator. They do not adjudicate their own hospital patients’ data. END was defined as any new neurological symptoms or signs or neurological worsening occurring within 72 hours after stroke onset, using the following criteria: (1) an increment in total NIHSS score of ≥2 points; (2) an increment in NIHSS consciousness score (1a to 1c) of ≥1; (3) an increment in NIHSS motor score (5a to 6b) ≥1; or (4) any new neurological deficit not assessed by the NIHSS [13,15,17,19]. Because of the descriptive nature of the study, we chose a low NIHSS cutoff for END to capture all possible events [20]. In addition, we selected the 72 hours criteria because ENDs beyond 72 hours after onset were less likely to be attributable to infarct growth. As previously described [13,15], the causes of END were classified as infarct growth, stroke recurrence, symptomatic hemorrhagic transformation, TIA, others, and unknown by consensus among experienced neurologists at each participating center.

Statistical analysis

Analyses evaluated circadian effects both broadly comparing night-onset (18:00 to 06:00) versus day-onset (06:00 to 18:00) strokes and at a more granular level stratifying by 4-hour time periods of onset. Baseline characteristics in the night-onset versus day-onset stroke groups were compared using the Student t test, the Wilcoxon rank-sum test, or the χ2 test according to variable types, as appropriate. Baseline characteristics among the groups stratified by the 4-hour time periods of onset were compared using one-way ANOVA or the Kruskal–Wallis test for continuous variables, and χ2 test for categorical variables, as appropriate.

Mixed-effects logistic regression models were used to investigate the association between stroke onset time and END while accounting for clustering by hospitals. The following predefined covariates that could be potentially associated with END were entered in the models: age, sex, prestroke mRS score, admission NIHSS score, previous stroke, hypertension, diabetes, hyperlipidemia, atrial fibrillation, smoking, stroke subtype, time from onset to hospital arrival, prestroke antiplatelet use, and prestroke statin use [20]. Considering seasonal variation in circadian rhythms [21], we also included stroke onset seasons as a covariate; we categorized patients into spring (March, April, and May), summer (June, July, and August), fall (September, October, and November), and winter (December, January, and February) groups.

To examine the association between stroke onset time and NIHSS score, we divided NIHSS score into 3 groups (0 to 1, 2 to 6, and ≥7), which allowed for each stratum to have a similar number of patients and performed mixed-effects ordered logistic regression analysis. In addition, we also used a mixed-effects negative binomial regression model with log link [22] to convey the sense of intergroup differences in stroke severity.

The effect of stroke onset time on dichotomized 3-month functional outcome (3-month mRS score 0 to 2 [favorable] versus 3 to 6 [unfavorable]) was also explored using a mixed-effects logistic regression model with adjustment for the same covariates in the models for stroke onset time versus END. The dichotomization was performed due to the violation of the proportional odds assumption.

As a post hoc analysis, stroke etiology-related differences in neurological severity, pathophysiological mechanisms, and circadian variation [7] were explored using statistical analyses after stratification by 3 stroke subtypes (LAA, SVO, and CE) [18]. The modifying effects of stroke subtypes were examined by entering an interaction term between stroke onset time and stroke subtypes into the model.

Sensitivity analyses were performed using the END criteria of the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST, increase of total NIHSS score ≥4). Moreover, to examine whether study results were confounded by changes in provision of care during off-hours, we examined the effect modification by weekdays versus weekend in the mixed-effects logistic regression model for the relationship between stroke onset time and END. The weekend included Korean national holidays as well as Saturdays and Sundays. In addition, we performed a sensitivity analysis on the relationship between stroke onset time and END after excluding a subset of patients with TIA.

The prospective analysis plan is available (S2 Text). The main analysis in the present report was consistent with the prospective analysis plan. The stroke subtype-related analyses were newly included at the data analysis stage. During the review process, we modified or added some more statistical analyses. First, we added the analysis investigating effect modification by weekdays versus weekend. Second, we categorized NIHSS score into 3 groups so as not to handle the score as a continuous variable. Third, we included stroke onset seasons as an additional covariate. Fourth, we added an additional sensitivity analysis that was performed after excluding patients with TIA. Data were analyzed using STATA software 16.0 (STATA Corp., Texas, USA). All significance tests were 2-sided, and p < 0.05 was considered statistically significant. In interaction analyses, which tend to have relatively low sensitivity, p < 0.10 was considered to indicate a potential interaction.

Results

A total of 60,364 patients with acute ischemic stroke or TIA were admitted within 7 days of symptom onset between May 2011 and July 2020 at the 11 participating centers. After excluding 42,903 patients based on the exclusion crteria, 17,461 patients were finally included in this study. A total of 1,893 (12.8%) patients were lost to follow-up and were excluded from the 3-month outcome analysis.

Baseline characteristics of study population: Day-onset versus night-onset stroke

The mean age of the 17,461 patients was 66.9 (SD 13.4) years, and 6,900 (39.5%) were women. Witnessed ischemic strokes and TIAs were more common in the morning hours (S1 Fig). Compared with the day-onset stroke patients (n = 12,449), the night-onset stroke patients (n = 5,012) were more likely to be younger, men, and smokers, and were less likely to have hypertension (Table 1). In addition, the night-onset stroke patients were likely to visit hospital earlier, receive revascularization therapy more frequently, and have a longer door-to-needle time and door-to-puncture time than the day-onset stroke patients. Grouping of stroke onset times by 4-hour intervals showed that the onset-to-arrival time was shortest between 22:00 and 02:00; revascularization therapy was more frequently performed between 18:00 and 22:00 and between 02:00 and 06:00 than other 4-hour time periods; and door-to-needle time and door-to-puncture time were longer between 22:00 and 02:00 and between 02:00 and 06:00 than other 4-hour time periods (S1 Table).

Table 1

Baseline characteristics.

Day-onset vs. night-onset ischemic stroke.

	06:00–18:00 (n = 12,449)	18:00–06:00 (n = 5,012)	p-value
Age, years	67.7 (13.1)	65.1 (13.9)	<0.001
Sex, women	4,985 (40.0)	1,915 (38.2)	0.025
Previous stroke	2,613 (21.0)	1,017 (20.3)	0.30
Hypertension	8,018 (64.4)	3,126 (62.4)	0.01
Diabetes	3,533 (28.4)	1,383 (27.6)	0.30
Hyperlipidemia	3,884 (31.2)	1,600 (31.9)	0.35
Current or recent* smoking	4,349 (34.9)	1,909 (38.1)	<0.001
Atrial fibrillation	3,139 (25.2)	1,320 (26.3)	0.12
Coronary artery disease	1,176 (9.5)	497 (9.9)	0.34
Prestroke mRS 0 or 1	11,229 (90.2)	4,546 (90.7)	0.31
Prestroke antiplatelet use	3,623 (29.1)	1,381 (27.6)	0.04
Prestroke statin use	2,658 (21.4)	1,076 (21.5)	0.86
Prestroke antihypertensive use	6,329 (50.8)	2,419 (48.3)	0.002
Prestroke antidiabetic use	2,650 (21.3)	1,032 (20.6)	0.31
Time from onset to hospital arrival, hour	2.0 (0.9 to 3.5)	1.4 (0.8 to 2.7)	<0.001†
Stroke subtype			<0.001
LAA	3,250 (26.1)	1,217 (24.3)
SVO	1,354 (10.9)	569 (11.4)
CE	3,097 (24.9)	1,270 (25.3)
Undetermined	2,714 (21.8)	1,015 (20.3)
Other-determined	355 (2.9)	139 (2.8)
TIA	1,679 (13.5)	802 (16.0)
Revascularization therapy	4,239 (34.1)	1,812 (36.2)	0.008
Intravenous	2,580 (20.7)	1,212 (24.2)
Intra-arterial	592 (4.8)	208 (4.2)
Intravenous + intra-arterial	1,067 (8.6)	392 (7.8)
Door-to-needle time, min	36 (27 to 49)	38 (28 to 52)	0.001†
Door-to-puncture time, min	98 (75 to 130)	112 (89 to 146)	<0.001†
Season			0.14
Spring	3,121 (25.1)	1,242 (24.8)
Summer	3,254 (26.1)	1,371 (27.4)
Fall	3,159 (25.4)	1,199 (23.9)
Winter	2,915 (23.4)	1,200 (23.9)
Symptomatic hemorrhagic transformation	161 (1.3)	72 (1.4)	0.46

Data are mean (SD), number (%), or median (interquartile range). Student t test was used for continuous variables, and χ2 test was used for categorical variables.

CE, cardioembolism; LAA, large artery atherosclerosis; mRS, modified Rankin Scale; NIHSS, National Institute of Health Stroke Scale; SVO, small vessel occlusion; TIA, transient ischemic attack.

*Quit smoking within 5 years of stroke onset.

†Rank-sum test was used.

END occurred in 2,219 (12.7%) patients. The leading cause of END was stroke progression (76.3%), followed by hemorrhagic transformation (8.9%), stroke recurrence (6.7%), unknown (4.0%), and others (2.8%). About two-thirds (65.1%) of END occurred on the first day after stroke onset, and 88.5% of END within 2 days after onset.

Associations of night-onset strokes with higher neurological severity, more frequent END, and worse 3-month functional outcome

In the mixed-effects ordered logistic regression analysis, night-onset strokes had a higher likelihood of being more severe (adjusted common odds ratio = 1.08, 95% CI, 1.02 to 1.17). The mixed-effects negative binomial regression analysis (S2 Table) showed that night-onset strokes were associated with a higher presenting NIHSS score, compared with day-onset strokes (estimated mean NIHSS 6.1 versus 5.7; p < 0.001). After adjusting for covariates, END was more prevalent in night-onset strokes than in day-onset strokes (14.4% versus 12.8%; adjusted risk difference 1.6% [95% CI 0.4 to 2.7]; p = 0.006; Fig 1). This association was not modified by revascularization therapy (p for interaction = 0.57). Functional outcome data at 3 months were available in 15,568 (87.2%) patients. After adjustment for covariates, night-onset strokes had a significantly lower likelihood of favorable functional outcome (76.5% versus 77.6%; adjusted odds ratio, 0.88 [95% CI 0.79 to 0.98]; p = 0.03), compared with day-onset strokes.

Fig 1

Adjusted frequency of END stratified by stroke onset time (day-onset versus night-onset) and revascularization therapy.

Error bar indicates 95% confidence interval. Mixed-effects logistic regression models were used with adjustment for age, sex, previous stroke, prestroke mRS score, admission NIHSS score, hypertension, diabetes, hyperlipidemia, smoking, atrial fibrillation, prestroke antiplatelet use, prestroke statin use, revascularization, time from onset to hospital arrival, season of stroke onset, and stroke subtype. p for interaction by revascularization therapy = 0.57. Note that the revascularization population has more severe stroke than the nonrevascularization population at baseline. Revascularization therapy improves their outcomes but not to the level of patients with initially milder deficits. END, early neurological deterioration; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale.

When stroke onset times were grouped by 4-hour intervals, nighttime strokes had a higher likelihood of being more severe (Fig 2), rising from a nadir in the 06:00 to 10:00 time period to a peak in the 02:00 to 06:00 time period. Similar findings were observed in the mixed-effects negative binomial regression analysis in terms of the association between stroke onset time and admission NIHSS score (S2 Table). The 18:00 to 22:00 and 22:00 to 02:00 onset stroke patients were more likely to experience END than other time period onset stroke patients. At 3 months, a monotonic gradient in the proportions of favorable functional outcome was also noted, falling from a peak at 06:00 to 10:00 to a nadir at 22:00 to 02:00.

Fig 2

Multivariable associations of stroke onset time (at 4-hour intervals) with admission NIHSS score, END, and 3-month functional outcome.

Dots and error bars indicate data estimates and their 95% confidence intervals, respectively. For the relationship between stroke onset time and presenting stroke severity, NIHSS score was stratified into 3 groups (0–1, 2–6, and ≥7) with a similar number of patients in each stratum and mixed-effects ordered logistic regression was performed with adjustment for age, sex, previous stroke, prestroke mRS score, hypertension, diabetes, hyperlipidemia, smoking, atrial fibrillation, prestroke antiplatelet use, prestroke statin use, time from onset to hospital arrival, season of stroke onset, and stroke subtype. For the relationships of stroke onset time with END and 3-month functional outcome, mixed-effects logistic regression analysis was performed with adjustment for age, sex, previous stroke, prestroke mRS score, admission NIHSS score, hypertension, diabetes, hyperlipidemia, smoking, atrial fibrillation, prestroke antiplatelet use, prestroke statin use, revascularization, time from onset to hospital arrival, season of stroke onset, and stroke subtype. *p < 0.05 compared with the 06:00–10:00 group. END, early neurological deterioration; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale.

Stroke subtype-specific associations of stroke onset time with neurological severity, END, and 3-month functional outcome

A total of 4,467 (25.6%) patients had LAA stroke, 1,923 (11.0%) SVO stroke, and 4,367 (25.0%) CE stroke. After the subtype stratification, the association of night-onset (versus day-onset) strokes with END remained significant in LAA stroke (22.7% versus 18.7%, p = 0.005), but not in SVO (p = 0.13) or CE stroke (p = 0.95); p for interaction = 0.14 (S2 Fig). When stroke onset times were grouped by 4-hour intervals, these diurnal patterns were again most pronounced in LAA stroke, followed by SVO stroke (Table 2; see S3 Table for unadjusted comparisons). Patterns of NIHSS score variation were similar in LAA and SVO stroke, with the highest likelihood of more severe stroke in the 02:00 to 06:00 and 22:00 to 02:00 periods, respectively. CE stroke similarly had higher values at these times but also had a relatively high value in the 10:00 to 14:00, 10:00 to 14:00, and 18:00 to 22:00 periods compared with the 06:00 to 10:00 period (Table 2). In LAA stroke, END rates were higher in 18:00 to 22:00 and 22:00 to 02:00 time periods, but in SVO and CE stroke, END rates were not different across all 4-hour increments. Patterns of favorable functional outcome at 3 months were similar in LAA and CE stroke, with the lowest likelihood of favorable outcome at 22:00 to 02:00 for both subtypes. In contrast, for SVO stroke, the likelihood of favorable outcome was highest in the 22:00 to 02:00 period.

Table 2

Multivariable associations of stroke onset time (4-hour intervals) with admission NIHSS score, neurological deterioration, and 3-month functional outcome after stratification with stroke subtypes.

	06:00–10:00	10:00–14:00	14:00–18:00	18:00–22:00	22:00–02:00	02:00–06:00
LAA
Number of patients	1,047	1,254	949	731	268	218
Admission NIHSS score*
Common odds ratios (95% CI)	Reference	0.97 (0.83 to 1.14)	1.13 (0.95 to 1.34)	1.01 (0.83 to 1.22)	1.20 (0.92 to 1.56)	1.28 (0.96 to 1.71)
p-value	Reference	0.73	0.18	0.92	0.18	0.09
END
Adjusted incidence (95% CI)†, %	15.7 (13.1 to 18.3)	19.0 (16.4 to 21.6)	15.9 (13.3 to 18.6)	20.8 (17.4 to 24.1)	22.1 (16.8 to 27.3)	17.2 (12.0 to 22.3)
Adjusted risk difference (95% CI), %	Reference	3.3 (0.2 to 6.4)	0.3 (−2.9 to 3.5)	5.1 (1.4 to 8.8)	6.4 (0.9 to 11.9)	1.5 (−3.9 to 6.9)
p-value	Reference	0.038	0.88	0.006	0.014	0.58
Favorable outcome‡
Adjusted incidence (95% CI)†, %	79.1 (75.8 to 82.3)	78.2 (75.0 to 81.4)	78.0 (74.6 to 81.5)	76.7 (72.8 to 80.5)	73.8 (68.1 to 79.5)	75.9 (70.2 to 81.5)
Adjusted outcome difference (95% CI), %	Reference	−0.9 (−3.8 to 2.0)	−1.1 (−4.2 to 2.1)	−2.4 (−5.9 to 1.1)	−5.2 (−10.6 to −0.1)	−3.2 (−8.6 to 2.2)
p-value	Reference	0.55	0.51	0.17	0.044	0.23
SVO
Number of patients	438	493	423	339	131	99
Admission NIHSS score*
Common odds ratios (95% CI)	Reference	0.87 (0.67 to 1.13)	0.85 (0.65 to 1.12)	0.99 (0.74 to 1.32)	1.14 (0.76 to 1.71)	0.94 (0.60 to 1.48)
p-value	Reference	0.29	0.26	0.93	0.54	0.79
END
Adjusted incidence (95% CI), %	13.5 (9.2 to 17.8)	13.8 (9.5 to 18.0)	11.4 (7.4 to 15.4)	16.2 (10.9 to 21.5)	14.3 (7.3 to 21.4)	18.8 (10.3 to 27.2)
Adjusted risk difference (95% CI), %	Reference	0.3 (−4.2 to 4.7)	−2.1 (−6.5 to 2.4)	2.7 (−2.5 to 8.0)	0.8 (−6.3 to 8.0)	5.3 (−3.0 to 13.5)
p-value	Reference	0.91	0.36	0.30	0.82	0.18
Favorable outcome‡
Adjusted incidence (95% CI)†, %	91.2 (88.0 to 94.3)	93.1 (90.5 to 95.7)	92.7 (89.8 to 95.6)	93.5 (90.4 to 96.5)	96.5 (93.2 to 99.7)	93.0 (88.1 to 97.9)
Adjusted outcome difference (95% CI), %	Reference	1.9 (−1.0 to 4.9)	1.5 (−1.7 to 4.7)	2.3 (−1.1 to 5.8)	5.3 (1.2 to 9.3)	1.8 (−3.3 to 6.9)
p-value	Reference	0.20	0.35	0.20	0.048	0.51
CE
Number of patients	970	1,145	982	825	252	193
Admission NIHSS score*
Common odds ratios (95% CI)	Reference	1.24 (1.03 to 1.50)	1.27 (1.05 to 1.54)	1.25 (1.02 to 1.54)	1.57 (1.16 to 2.12)	1.78 (1.26 to 2.50)
p-value	Reference	0.02	0.015	0.03	0.004	0.001
END
Adjusted incidence (95% CI)†, %	13.7 (11.3 to 16.1)	14.6 (12.3 to 16.9)	14.1 (11.7 to 16.5)	14.5 (11.8 to 17.2)	14.2 (9.7 to 18.8)	14.8 (9.6 to 20.2)
Adjusted risk difference (95% CI), %	Reference	0.9 (−2.1 to 3.8)	0.4 (−2.7 to 3.5)	0.8 (−2.4 to 4.1)	0.6 (−4.4 to 5.5)	1.2 (−4.4 to 6.7)
p-value	Reference	0.56	0.79	0.61	0.82	0.67
Favorable outcome‡
Adjusted incidence (95% CI)†, %	62.8 (58.8 to 66.9)	62.0 (58.1 to 66.0)	62.2 (58.2 to 66.2)	59.7 (55.4 to 63.9)	55.7 (49.7 to 61.7)	59.8 (53.6 to 66.1)
Adjusted outcome difference (95% CI), %	Reference	−0.8 (−4.1 to 2.6)	−0.6 (−4.0 to 2.8)	−3.1 (−6.8 to 0.5)	−7.1 (−12.7 to −1.5)	−3.0 (−8.8 to 2.8)
p-value	Reference	0.64	0.73	0.09	0.012	0.31

CE, cardioembolism; CI, confidence interval; END, early neurological deterioration; LAA, large artery atherosclerosis; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; SVO, small vessel occlusion.

*Admission NIHSS score was categorized into 3 groups (0–1, 2–6, and ≥7). Mixed-effects ordered logistic regression was used. Adjusted for age, sex, previous stroke, prestroke mRS score, hypertension, diabetes, hyperlipidemia, smoking, atrial fibrillation, prestroke antiplatelet use, prestroke statin use, revascularization, season of stroke onset, and time from onset to hospital arrival.

†Adjusted for age, sex, previous stroke, prestroke mRS score, admission NIHSS score, hypertension, diabetes, hyperlipidemia, smoking, atrial fibrillation, prestroke antiplatelet use, prestroke statin use, revascularization, season of stroke onset, and time from onset to hospital arrival.

‡3-month mRS score 0–2 versus 3–6 (unfavorable).

Sensitivity analyses

When END was defined using the SITS-MOST criteria [23], night-onset strokes were again more frequently associated with END (5.7%, 95% CI, 5.1% to 6.4%), compared with day-onset strokes (4.9%, 95% CI, 4.6% to 5.3%) after adjusting for covariates (p = 0.03; S3 Fig). There was no statistically significant weekdays versus weekend-related interaction (p for interaction = 0.50; S4 Table). In addition, the sensitivity analysis for the non-TIA patients showed that night-onset stroke patients had END more frequently compared with day-onset stroke patients (S5 Table).

Discussion

In this nationwide multicenter study on 17,461 witnessed acute ischemic stroke or TIA with the onset-to-arrival time being less than 6 hours, night-onset stroke patients had a higher admission NIHSS score, a higher likelihood of experiencing END, and a lower likelihood of favorable 3-month outcome than day-onset stroke patients. Circadian variation patterns differed among ischemic stroke subtypes, with LAA patients particularly showing more frequent END and worse 3-month functional outcome with night-onset, CE patients showing more severe presenting deficits and worse 3-month functional outcome with night-onset, and SVO patients not evidencing night compared with day differences. To the best of our knowledge, this is the first large-scale study demonstrating circadian effects on END as well as ischemic stroke severity and 3-month functional outcome.

In a recent preclinical study using rodent models of MCA occlusion, inactive phase (daytime in rodents) onset strokes were associated with more active cell death and infarct growth, compared with active phase onset strokes [8]. The present large-scale clinical study demonstrated that inactive phase (nighttime) onset strokes in human patients were similarly associated with more severe deficits at presentation and a higher likelihood of END, which was mostly (in approximately 75%) due to infarct growth. As a possible confounder, we considered potential differences in care quality between duty hours and off-hours. However, the circadian effect on presenting stroke severity and END was not modified when stroke onset time was further stratified by weekdays versus weekend. Accordingly, the diurnal variations in severity and progression appear to be biologically driven, rather than due to variations in systems of care. Similar care quality on weekdays and weekends is in line with recent reports [24,25]. Moreover, after applying the SITS-MOST criteria of END, the independent association between stroke onset time and END was retained, which supports the robustness of our results. In addition, the incidence of END in our population (12.7%) was comparable with previous reports [26,27].

Night-onset LAA strokes were more prone to END than day-onset LAA strokes, whereas the initial neurological severity of night-onset CE strokes was higher compared with day-onset CE strokes. A possible explanation for this pattern is that LAA strokes tend to have better collateral flow than CE strokes after acute large vessel occlusion [28]. As a result, LAA stroke may have slower initial infarct growth and thus a higher likelihood of continued progression after arrival, resulting in more in-hospital END in LAA stroke. Conversely, CE stroke may have more rapid initial lesion growth, producing more neurological worsening before hospital arrival and a higher stroke severity at admission.

The current large-scale study of ischemic stroke clinical presentation and course is consonant with recent smaller studies of ischemic stroke imaging. An analysis of CT perfusion imaging data in patients with anterior large vessel occlusion stroke found larger initial ischemic core volume with night-onset compared with day-onset strokes [29]. Core growth from onset to imaging proceeded was also faster in night-onset than day-onset strokes. The present study findings in ischemic stroke align with a post hoc analysis [30] of intracerebral hemorrhage in the Intensive Blood Pressure Reduction in an Acute Cerebral Hemorrhage Trial (INTERACT) that demonstrated night-onset was associated with a higher likelihood of worse Glasgow Coma Scale score (8 or less) at admission, compared with day-onset, despite a lack of significant association with the volume of intracerebral hemorrhage [30]. These findings suggest that the pathophysiologic mechanisms of ischemic and hemorrhagic stroke are affected in a consonant manner by circadian variation.

The current study results also align with findings from investigations of myocardial ischemia. In 165 patients who had ST-elevation myocardial infarction (STEMI) with known ischemic times between 1 and 6 hours, the greatest myocardial injury occurred at 1 AM onset of ischemia and 5 AM onset of reperfusion with primary percutaneous coronary intervention [31]. A study in 1,548 consecutive patients with STEMI found poorer myocardial perfusion and larger infarct size in the early morning. As a potential causal factor, platelet aggregation was highest between 4 AM and 8 AM [32]. Other studies showed the greatest decrease in left ventricle function in myocardial infarction begin at approximately 1 AM [31,33]. These findings in ischemic heart disease are largely in line with the 4-hour interval data in the present study on ischemic stroke. Further research is required to compare the mechanisms underlying circadian variation of ischemic heart injury with those of ischemic brain injury.

This study has several strengths, including the large sample size, consecutive enrollment of patients with clear onset of ischemic stroke, prospective data collection, and regular data audits in monthly investigator meetings and annual adjudication committee meetings. END was prospectively determined during hospitalization by experienced stroke nurses and validated by the neurologists participating in the study, as part of a weekly institutional quality-of-care monitoring program for stroke patients [20]. Brain imaging was used to confirm and categorize END. There are also limitations that must be considered when interpreting our data. First, using quantitative penumbral imaging data would more directly demonstrate the circadian effects on infarct growth. Second, we did not obtain information on sleep disorders such as obstructive sleep apnea, which can affect both circadian rhythms and stroke [34]. Third, we did not obtain information on patient work/activity schedules that would allow probing of shift work effects. Fourth, further investigation with a larger sample size is required for a higher level of data granularity, which could demonstrate the association between stroke onset time and outcomes at relatively narrow time intervals. Fifth, we have followed the medical literature approach in use of the term “circadian” for a phenomenon that would be termed “diurnal” in the preclinical literature. Sixth, wake-up stroke was shown to be associated with a greater initial severity and a worse functional outcome at 3 months, compared with wakefulness stroke [5]. In another study, daytime-unwitnessed stroke patients were more likely to receive reperfusion therapy due to earlier hospital arrival after symptom recognition, compared with wake-up stroke patients [35]. Thus, the exclusion of unwitnessed stroke in the present study might have biased toward the null, potentially leading to an underestimation of the association between stroke onset time and outcomes.

In conclusion, compared with day-onset strokes, night-onset strokes are associated with more frequent END as well as higher initial neurological severity and worse 3-month functional outcome. We suggest that the circadian factor should be considered in designing future neuroprotection trials.

STROBE checklist.

(DOCX)

Click here for additional data file.

Baseline characteristics by stroke onset time (4-hour intervals).

(DOCX)

Click here for additional data file.

Mixed-effects negative binomial logistic regression analysis between admission NIHSS score and stroke onset time.

NIHSS, National Institutes of Health Stroke Scale.

(DOCX)

Click here for additional data file.

Unadjusted associations of stroke onset time (4-hour intervals) with admission NIHSS score, neurological deterioration, and 3-month functional outcome after stratification with stroke subtypes.

NIHSS, National Institutes of Health Stroke Scale.

(DOCX)

Click here for additional data file.

Multivariable associations between stroke onset time and END: Weekdays versus weekend. END, early neurological deterioration.

(DOCX)

Click here for additional data file.

Multivariable associations between stroke onset time and END after exclusion of TIA patients.

END, early neurological deterioration; TIA, transient ischemic attack.

(DOCX)

Click here for additional data file.

Distribution of the study population by stroke onset time.

(DOCX)

Click here for additional data file.

Adjusted incidence of END stratified by stroke onset time and stroke subtype.

END, early neurological deterioration.

(DOCX)

Click here for additional data file.

Multivariable associations between stroke onset time and END defined by the SITS-MOST criteria.

END, early neurological deterioration; NIHSS, National Institutes of Health Stroke Scale; SITS-MOST, Safe Implementation of Thrombolysis in Stroke-Monitoring Study.

(DOCX)

Click here for additional data file.

List of institutional review boards.

(DOCX)

Click here for additional data file.

Prospective analysis plan (translated in English).

(DOCX)

Click here for additional data file.

24 Sep 2021

Dear Dr Ryu,

Thank you for submitting your manuscript entitled "Circadian Time of Ischemic Stroke Onset Affects Presenting Severity, Acute Progression, and Long-term Outcome" for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff as well as by an academic editor with relevant expertise and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by Sep 28 2021 11:59PM.

Login to Editorial Manager here: https://www.editorialmanager.com/pmedicine

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Callam Davidson

Associate Editor

PLOS Medicine

27 Oct 2021

Dear Dr. Ryu,

Thank you very much for submitting your manuscript "Circadian Time of Ischemic Stroke Onset Affects Presenting Severity, Acute Progression, and Long-term Outcome" (PMEDICINE-D-21-04039R1) for consideration at PLOS Medicine.

Your paper was evaluated by an associate editor and discussed among all the editors here. It was also discussed with an academic editor with relevant expertise, and sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We expect to receive your revised manuscript by Nov 17 2021 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Callam Davidson,

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

To facilitate review, please include continuous line numbering in the margin of your manuscript.

Please revise your title according to PLOS Medicine's style. Your title must be nondeclarative and not a question. It should begin with main concept if possible. "Effect of" should be used only if causality can be inferred, i.e., for an RCT. Please place the study design (“a cohort study”) in the subtitle (ie, after a colon).

Please remove the ‘Funding’ information from your acknowledgements section and instead please enter this into the submission form under the ‘Financial Disclosure’ section.

The Data Availability Statement (DAS) requires revision. If the data are owned by a third party but freely available upon request, please note this and state the owner of the data set and contact information for data requests (web or email address). Note that a study author cannot be the contact person for the data.

Abstract Methods and Findings:

* Please ensure that all numbers presented in the abstract are present and identical to numbers presented in the main manuscript text.

* Please include the setting, year in which the study took place, length of follow up, and main outcome measures.

* Please include the important dependent variables that are adjusted for in the analyses.

* In the last sentence of the Abstract Methods and Findings section, please describe the main limitation(s) of the study's methodology.

At this stage, we ask that you include a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract. Please see our author guidelines for more information: https://journals.plos.org/plosmedicine/s/revising-your-manuscript#loc-author-summary

Please temper the primacy claim in paragraph two of the introduction (‘To date, circadian effects on post-stroke END has never been investigated yet’) by adding ‘to our knowledge’.

Please provide the names of the institutional review boards that provided ethical approval.

Did your study have a prospective protocol or analysis plan? Please state this (either way) early in the Methods section.

a) If a prospective analysis plan (from your funding proposal, IRB or other ethics committee submission, study protocol, or other planning document written before analyzing the data) was used in designing the study, please include the relevant prospectively written document with your revised manuscript as a Supporting Information file to be published alongside your study, and cite it in the Methods section. A legend for this file should be included at the end of your manuscript.

b) If no such document exists, please make sure that the Methods section transparently describes when analyses were planned, and when/why any data-driven changes to analyses took place.

c) In either case, changes in the analysis-- including those made in response to peer review comments-- should be identified as such in the Methods section of the paper, with rationale.

Citations should appear before punctuation throughout the manuscript.

Please define "lost to follow-up" as used in this study. Other reasons for exclusion should be defined.

In Figures 1 and 2, please show the Y-axis beginning at zero. If this is not possible, please show a break in the axis.

Please remove the ‘competing interests’ and ‘author contributions’ from the end of the main text – in the event of publication, this information will be published as metadata based on your responses to the submission form questions.

Thank you for providing your STROBE checklist. Please replace the page numbers with paragraph numbers per section (e.g. "Methods, paragraph 1"), since the page numbers of the final published paper may be different from the page numbers in the current manuscript.

Comments from the reviewers:

Reviewer #1: The authors are to be commended for using a large prospective national multicenter stroke registry in South Korea to extend data on circadian patterns for stroke incidence to explore day vs night onset and adverse outcomes defined by early neurological deterioration (over 72 hours) on the popular NIHSS score and standard 90-day functional outcome on the mRS score. The rationale is that the ischemic penumbra may be more vulnerable in a 'resting state', as evident from basic science studies. The do confirm that neurological deterioration is greater with night onset, as well as variable differences in time to presentation and management, which were taken into account in various models with adjustment for multiple confounders which was offered by the large dataset involving 17,461 patients with either acute ischemic stroked or TIA out of the full registry database of 60,000+ cases who presented within 7 days. The data are interesting but concerns are raised over the approach to analysis and naturally, residual confounding.

1) my particular concern is the use of NIHSS to decimal places, which over-inflates the granularity of this measure. It is actually a categorical scale, so measures of 6.1 vs. 5.7 do not make any sense. All analyses need to be re-done accordingly.

2) while 'night' vs. 'day' can be a simple comparison, the analysis for 4 hourly time periods should be undertaken as a p for trend rather than categories against a control group. A significant time epoch against control could purely be due to chance from multiple testing

3) while adjustments have been made for hospital clustering and day of the week, there is also likely to be a seasonal component which also needs to be taken into account. Moreover, this also determines change in physiological variables, in particular systolic blood pressure but also blood glucose level, which may be particularly important on the ischemic penumbra and could be included in the models of baseline variables.

4) given variability in diagnosis (despite involvement of neurologists) and potential pathophysiology, I would like to see sensitivity analysis confined to acute ischemic stroke cases

5) Figure 1 should relate to 'frequency' rather than 'incidence'

Reviewer #2: See attachment

Michael Dewey

Reviewer #3: In their manuscript, Ryu et al. analyze the diurnal variation in a large national cohort of well defined and times stroke patients, and, importantly, have the ability to assess the severity of the stroke by outcome measures.

The authors confuse the difference between circadian (endogenous, and without any external time cue) and diurnal rhythms. Their data is in entrained conditions. Therefore, using the term circadian is wrong. This should be replaced by diurnal, time-of-day,... This is not a purely academic difference. Circadian refers to endogenous ("biological") effects, whereas there are a number of exogenous ("circumstances") effects, too. This study cannot resolve this dichotomy and, as the authors partly acknowledge by listing some limitations, can only point to correlation but not causation.

It is therefore, in my view, not warranted to conclude in the Discussion that the "These associations appeared to be biologically driven." when pointing to the association between mice and human patients. Sleep between the species is very different, as are number of other characteristics important for stroke.

Moreover, the authors do not comment on how there might be a biased introduced by their "unwitnessed stroke" exclusion criteria that makes them exclude about the same number of patients than they included. Maybe some time of day has a lower likelyhood of strokes being witnessed. This should be discussed.

The data is drawn throughout a period of about 9 years and there is nothing shown for seasonal effects nor discussed. This should be added to the caveats already mentioned (sleep, shift-work, chronotype).

In conclusion, I think this is an important study to make the medical community aware of daytime specific changes in disease and implications for treatment. There are a number of strengths here, but a few more caveats should be considered.

Any attachments provided with reviews can be seen via the following link:

[LINK]

Submitted filename: ryu.pdf

Click here for additional data file.

19 Nov 2021

Submitted filename: Response_to_reviewers_Final.docx

Click here for additional data file.

17 Dec 2021

Dear Dr. Ryu,

Thank you very much for re-submitting your manuscript "Association of circadian time of ischemic stroke onset with presenting severity, acute progression, and long-term outcome: A cohort study" (PMEDICINE-D-21-04039R2) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by three reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We hope to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Dec 24 2021 11:59PM.

Sincerely,

Callam Davidson,

Associate Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

Title: Please consider updating to ‘Association of ischemic stroke onset time and presenting severity, acute progression, and long-term outcome: A cohort study’.

Data Availability Statement: Please provide a contact email address for the steering committee in addition to the URL already provided.

Author Summary: Please revise your author summary in line with our guidelines (https://journals.plos.org/plosmedicine/s/revising-your-manuscript). More specifically, the summary should be non-technical and suitable for non-expert readers (including scientists and non-scientists). Avoid the use of complex terminology wherever possible and please make bullet points single sentences. Assertions of causality should also be avoided given the observational nature of the study (refer instead to associations).

Line 111: Remove ‘sample-sized’.

Line 150: Rather than providing the name of one representative institutional review board, please provide all IRB names in the supporting information and cite here (e.g. S1 Supporting Information). See here for more details regarding supporting information files (https://journals.plos.org/plosmedicine/s/supporting-information).

Line 230: Please include the original (unchanged) prospective analysis plan and relocate the details of additional or revised analyses performed during the revision process to your manuscript methods instead.

Table 1: In the legend, please indicate the statistical test used to generate p-values. Please also present descriptive statistics consistently (e.g., seasonal data has ‘%’ in parentheses while other percentages do not).

Figure 1: Please indicate the statistical test used in the legend.

Table 2: Please provide the unadjusted comparisons as well as the adjusted comparisons.

Lines 415-421: Please consider whether this content would be better placed in the limitations discussion in the paragraph below.

Comments from Reviewers:

Reviewer #1: The authors have adequately addressed Reviewer comments

Reviewer #2: The authors have dealt with all my points.

Michael Dewey

Reviewer #3: I have read the rebuttal and new manuscript of the authors.

Specifically

- While I am wouldn't want to stop the publication of this manuscript over this technicality: Circadian is incorrect. If there are many other articles that have been published including the same mistake, it remains a mistake and should be corrected. It becomes an important distinction especially in the clinical context once the authors would want, as they do, assign a "biological driven" association. I am still unconvinced of this and really, especially in such retrospective data collection, they cannot reasonably concluded this.

- Regarding the seasonal variation: While I think the analysis the authors did make is useful, I fear I did not make myself clear: For example, seasonal changes in photoperiod lead to changes in wake-up time (see papers by Roenneberg, for example), and, thus, as a hypothesis, there should be a shift in the timing of stroke according to this. This would be another argument for a "biologically driven" phenomenon.

I have no further comments.

Any attachments provided with reviews can be seen via the following link:

[LINK]

29 Dec 2021

Submitted filename: Response_to_Editor.docx

Click here for additional data file.

5 Jan 2022

Dear Dr. Ryu,

Thank you very much for re-submitting your manuscript "Association of ischemic stroke onset time with presenting severity, acute progression, and long-term outcome: A cohort study" (PMEDICINE-D-21-04039R3) for review by PLOS Medicine.

Please address the remaining editorial and production issues at the end of this email. I look forward to receiving the revised manuscript by Jan 12 2022 11:59PM.

Please email me (cdavidson@plos.org) if you have any questions or concerns.

Sincerely,

Callam Davidson,

Associate Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from the editor:

Please revise the Author Summary as follows:

* Line 76: Please update this bullet to read 'The diurnal pattern in the distribution of ischemic stroke suggests an influence of circadian rhythms on stroke incidence, but it is unclear whether circadian rhythms may also affect the clinical severity of stroke, and whether time-of-day of stroke occurrence may affect acute clinical worsening after stroke onset.'

* Line 83: Please update 'ask' to 'investigate'.

* Line 86: Please begin a new bullet here (i.e. beginning 'Night-onset stroke patients...').

* Line 88: Please begin a new bullet here (i.e. beginning 'These patients were also found to have...').

* Lines 90-95: Please remove this bullet point as this level of detail is not required in the Author Summary.

* Line 97: Please update 'demonstrate' to 'suggest'.

* Line 99: Please remove this bullet point as this level of detail is not required in the Author Summary.

Please update the text at lines 222-229 to read as follows:

‘The prospective analysis plan is available (S2 Text). The main analysis in the present report was consistent with the prospective analysis plan. The stroke subtype-related analyses were newly included at the data analysis stage. During the review process, we modified or added some statistical analyses. First, we added the analysis investigating effect modification by weekdays vs. weekends. Second, we categorized NIHSS scores into 3 groups so as not to handle the scores as a continuous variable. Third, we included stroke onset seasons as an additional covariate. Fourth, we added an additional sensitivity analysis that was performed after excluding patients with TIA.'

11 Jan 2022

Dear Dr Ryu,

On behalf of my colleagues and the Academic Editor, Dr Joshua Willey, I am pleased to inform you that we have agreed to publish your manuscript "Association of ischemic stroke onset time with presenting severity, acute progression, and long-term outcome: A cohort study" (PMEDICINE-D-21-04039R4) in PLOS Medicine.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Sincerely,

Callam Davidson

Associate Editor

PLOS Medicine