Original Paper

Neuroepidemiology Received: February 3, 2022

Accepted: March 9, 2022
DOI: 10.1159/000524206 Published online: March 23, 2022

Incidence of Aphasia in Ischemic Stroke

Angelina Grönberg a, b Ingrid Henriksson c Martin Stenman a
     

Arne G. Lindgren a, b  

aDepartment
of Clinical Sciences Lund, Neurology, Lund University, Lund, Sweden; bDepartment of Neurology,
Rehabilitation Medicine, Memory Disorders and Geriatrics, Skåne University Hospital, Lund, Sweden; cSpeech
and Language Pathology Unit, Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of
Gothenburg, Gothenburg, Sweden

Keywords 38 per 100,000 person-years) corresponding to a significant

Aphasia · Incidence · Ischemic stroke · Risk factors decrease of 30% between 2005–2006 and 2017–2018. The
decrease was significantly more pronounced in men. The ini-
tial severity of aphasia remained unchanged, with the major-
Abstract ity of patients having severe to global aphasia. No significant
Introduction: A decrease in ischemic stroke (IS) incidence differences between vascular stroke risk factors were noted
has been observed in high income countries during the last among stroke patients with or without aphasia. Conclusion:
decades. Whether this has influenced the occurrence of Even though the overall IS incidence rate has decreased dur-
aphasia in IS is uncertain. We therefore examined the inci- ing the first decades of the 21st century, the proportion of IS
dence rate and potentially related determinants of aphasia patients with aphasia at stroke onset remains stable at 30%.
in IS. Methods: We prospectively examined consecutive pa- Aphasia continues to be an important symptom that needs
tients admitted to hospital with first-ever acute IS between to be considered in stroke care and rehabilitation.
March 1, 2017, and February 28, 2018, as part of the Lund © 2022 The Author(s).
Stroke Register (LSR) Study, comprising patients from the Published by S. Karger AG, Basel

uptake area of Skåne University Hospital, Lund, Sweden. Pa-

tients were assessed with National Institutes of Health Stroke Introduction
Scale (NIHSS) at stroke onset. Presence of aphasia was evalu-
ated with NIHSS item 9 (language). We registered IS sub- Effective preventive stroke treatments and reduction
types and risk factors. To investigate possible temporal of stroke risk factors have led to a decrease in stroke inci-
changes in aphasia incidence, we made comparisons with dence in high income countries during the last decades
corresponding LSR data from 2005 to 2006. Incidence rates [1, 2]. Despite this, the number of patients living with
were calculated and adjusted to the European Standard stroke and the long-term effects of stroke are increasing
Population (ESP) and to the Swedish population. Results: due to an aging population and reduced stroke mortality
Among 308 included IS patients, 30% presented with apha- [1, 3] and is projected to grow in the coming decades [4].
sia (n = 91; 95% CI: 25–35), a proportion of aphasia in IS that This indicates that the burden of stroke will continue to
was similar to 2005–2006. The incidence rate of aphasia was remain high in society.
31 per 100,000 person-years adjusted to the ESP (95% CI: 25–

[email protected] © 2022 The Author(s). Correspondence to:

www.karger.com/ned Published by S. Karger AG, Basel Angelina Grönberg, angelina.gronberg @ med.lu.se
This is an Open Access article licensed under the Creative Commons
Attribution-NonCommercial-4.0 International License (CC BY-NC)
(http://www.karger.com/Services/OpenAccessLicense), applicable to
the online version of the article only. Usage and distribution for com-
mercial purposes requires written permission.
 Aphasia after stroke poses a major disability for the was determined by NIHSS item 9 “best language,” and we graded
patient and negatively impacts rehabilitation [5] and the severity of aphasia into 4 categories according to NIHSS item
9 (i.e., 0 = no aphasia, 1 = mild to moderate aphasia, 2 = severe
overall stroke outcome [6]. Moreover, aphasia is a condi- aphasia, 3 = global aphasia). Stroke severity according to NIHSS
tion that has one of the largest negative impacts on a per- was stratified into 3 groups: NIHSS score of 1–4 = mild stroke;
son’s health-related quality of life [7], with high risk of 5–14 = moderate stroke; and ≥15 = severe stroke [17].
depression and lower likelihood of returning to work [8]. We collected additional patient data by interviewing the pa-
Accurate knowledge of symptoms and factors associated tients and/or their next of kin, and reviewing the patients’ medical
records concerning: whether care was provided at a dedicated
with aphasia are essential to provide optimal care of pa- stroke unit, length of stay, discharge location, in-hospital death,
tients with this language disorder. and traditional stroke risk factors: hypertension, diabetes mellitus,
Recent changes in stroke incidence and advances in atrial fibrillation (AF), hypercholesterolemia, smoking, previous
stroke care may also affect the epidemiology of aphasia in TIA, heart disease and ischemic heart disease, as defined previ-
stroke [9]. Historically, the proportion of stroke patients ously [18, 19]. We also registered demographic characteristics: age,
gender, and education.
with aphasia has been reported to be approximately 30%, The evaluations were performed primarily during the patients’
although with a considerable range, varying between 19% hospital stay and missing data was subsequently complemented by
and 62% among patients with ischemic stroke (IS) [10– contacting the patients and/or next of kin by telephone after their
13]. However, up-to-date data on the incidence of aphasia discharge from hospital. A research physician determined patho-
after IS in the past decade are scarce, and knowledge of genetic stroke mechanism according to the Trial of Org in Acute
Stroke Treatment (TOAST) classification [20] and clinical stroke
current incidence and severity of aphasia is essential to syndrome by using The Oxfordshire Community Stroke Project
facilitate planning of stroke rehabilitation, improve de- (OCSP) definitions [21]. Corresponding data from LSR between
velopment of clinical guidelines for aphasia, and ulti- March 1, 2005, and February 28, 2006 (year 2005–2006), were used
mately estimate health care cost. The aims of this study to investigate possible temporal changes in aphasia incidence (pa-
were to: (1) report the current incidence and severity of tient inclusion criteria were the same as described above). We also
compared risk factors and demographic data between 2005–2006
aphasia after first-ever IS; (2) identify pathogenetic mech- and 2017–2018.
anisms and risk factors associated with aphasia in IS; (3)
investigate potential temporal changes of aphasia inci- Data Analysis
dence after IS. Age- and sex-standardized incidence rates were calculated by
using the direct method and per 100,000 person-years with 95%
confidence intervals. Incidence rates were age and gender stan-
dardized to the Swedish population as of December 31, 2017 [14],
Methods and to the European Standard Population (ESP) from 2013 [22].
We assessed the occurrence of aphasia across age and gender.
Data supporting the findings of this study are available from Categorical and binary variables were analyzed with the χ2 test. We
the corresponding author upon reasonable request. We consecu- applied Mann-Whitney U test to detect differences between year
tively included first-ever IS patients in the prospective study Lund 2005–2006 and year 2017–2018 in continuous variables (such as
Stroke Register (LSR), comprising the local uptake area of Skåne stroke severity according to NIHSS). Comparisons of categorical
University Hospital Lund (SUHL), Sweden between March 1, data with small sample size were examined using Fisher’s exact
2017, and February 28, 2018. SUHL is the only hospital in this area test.
designated for acute stroke care and stroke patients admitted to the Patients were divided into two subgroups according to the
hospital are identified for inclusion in LSR using “hot pursuit” presence or absence of aphasia. Associations between aphasia and
methods as described below. The total population of the uptake age (unadjusted and adjusted for stroke severity), gender, educa-
area was 284,003 inhabitants (all ages) as of December 31, 2017 tion, and risk factors were examined using logistic regression anal-
[14], with 8% ≥75 years and 50% females [14]. yses. Associations between aphasia and stroke severity were exam-
To identify first-ever IS patients for inclusion in the study dur- ined using the total NIHSS score excluding the NIHSS aphasia
ing their hospital stay, research nurses during weekdays screened component (NIHSS item 9). Associations between aphasia and
charts of patients admitted to SUHL for stroke symptoms. Stroke TOAST classification [20] were made including the 5 categories:
was defined according to the WHO criteria [15] and previous stud- (1) large-artery atherosclerosis (LAA), (2) cardioembolism (CE),
ies show that LSR includes approximately 94% of all patients with (3) small-artery occlusion (SAO), (4) stroke of other determined
first-ever stroke [16]. We included patients who after information etiology (OC), and (5) stroke of undetermined etiology (UND),
consented to participate in the study. and in addition, we performed calculations excluding SAO. The
A physician performed assessment with the National Institutes latter was done because a prerequisite for the TOAST category
of Health Stroke Scale (NIHSS) at stroke onset (hospital admis- SAO is that the patient should not have evidence of cortical dys-
sion, median day 0). NIHSS was assessed before any potential ad- function such as aphasia.
ministration of acute stroke recanalization treatment. When need- Confidence intervals and comparisons of incidence rates be-
ed, we performed additional clinical assessments and also reviewed tween year 2005–2006 and year 2017–2018 were performed using
data from the patients’ medical records. The diagnosis of aphasia Open Source Epidemiologic Statistics for Public Health [23]. All

2 Neuroepidemiology Grönberg/Henriksson/Stenman/Lindgren
DOI: 10.1159/000524206
Fig. 1. Study flow diagram of the LSR co-
hort. LSR, Lund Stroke Register; NIHSS,
National Institutes of Health Stroke Scale.

other statistical calculations were performed with the SPSS soft- cantly more severe strokes, were older (Table 1), and had
ware package 25. Values of p < 0.05 were considered statistically longer hospital stays in comparison to stroke patients
significant. The study was approved by the Regional Ethical Re-
view Authority in Lund (registration number 2016/179). without aphasia (median 8 days vs. 4 days; OR, 1.08; 95%
CI: 1.04–1.12). These factors remained significantly as-
sociated with aphasia also after adjusting for NIHSS
scores excluding the aphasia component (NIHSS item 9).
Results Patients with aphasia were more likely to be discharged
to a short-term care facility (25% vs. 13%; p = 0.01), and
In total, 338 patients were diagnosed with first-ever IS the overall in-hospital mortality was higher for patients
between March 1, 2017, and February 28, 2018. Among with aphasia (18%) in comparison to those without apha-
these, 308 patients were included in the study (Fig. 1). The sia (2%, OR, 9.05; 95% CI: 3.20–25.54). However, when
median age of the included patients was 76 years (IQR adjusting for stroke severity, the abovementioned factors
69–82 years) and 152 (49%) were female. Baseline char- were no longer significant among patients with or with-
acteristics are shown in Table 1. out aphasia. There were no significant differences regard-
The incidence rate of first-ever IS was 108 per 100,000 ing the prevalences of stroke risk factors (hypertension,
person-years (95% CI: 97–121 per 100,000 person-years) diabetes mellitus, AF, hypercholesterolemia, smoking, is-
adjusted to the ESP. The overall proportion of IS patients chemic heart disease, heart disease) between patients with
with aphasia was 30% (95% CI: 25–35%) according to NI- or without aphasia. The proportion of stroke patients
HSS item 9 (n = 91) in the acute phase of stroke onset. The treated at a dedicated stroke unit was high (94%), and
overall crude incidence rate of first-ever IS aphasia there was no difference between patients with or without
amounted to 32 per 100,000 person-years (95% CI: 26–39 aphasia (94% and 92%, respectively; OR, 1.01; 95% CI:
per 100,000 person-years). The age- and sex-standard- 0.34–2.99).
ized incidence rate adjusted to the Swedish population (of
December 2017) was 35 per 100,000 person-years (95% Aphasia in Relation to Stroke Severity and Age
CI: 33–49 per 100,000 person-years). Adjusted to the ESP, The prevalence of aphasia increased significantly with
the overall incidence rate of aphasia after IS was 31 per stroke severity, as measured by NIHSS (after excluding
100,000 person-years (95% CI: 25–38 per 100,000 person- the aphasia component, NIHSS item 9, p < 0.001). Each
years). There was no significant difference between males 1-point increase on NIHSS (excluding item 9) increased
and females in aphasia incidence rate adjusted to ESP year the odds of aphasia by 19% (OR, 1.19; 95% CI: 1.13–1.26),
2017–2018 (p = 0.92). Patients with aphasia had signifi- after adjusting for age. The prevalence of aphasia in-

Severity, Etiology, and Temporal Change Neuroepidemiology 3

DOI: 10.1159/000524206
 Table 1. Baseline characteristics of IS patients with and without aphasia year 2017–2018

Patients with first-ever ischemic stroke

variable patients without patients with all patients OR
aphasia (n = 217) aphasia (n = 91) (n = 308) (95% CI)*

Age, years, median (IQR) 74 (68–81) 78 (72–86) 76 (69–82) 1.04 (1.02–1.06)

Female gender, n (%) 101 (47) 51 (56) 152 (49) 1.46 (0.90–2.40)
NIHSS at baseline, median (IQR) 3 (1–5) 10 (4–19) 4 (2–7) 1.25 (1.18–1.32)
Stroke risk factors†
Hypertension 172 (79) 70 (77) 242 (79) 0.87 (0.48–1.57)
Diabetes mellitus 72 (33) 28 (31) 100 (33) 0.90 (0.53–1.52)
AF 62 (29) 34 (37) 96 (31) 1.49 (0.90–2.50)
Hypercholesterolemia 123 (57) 47 (52) 170 (55) 0.82 (0.50–1.33)
Smoking 38 (18) 11 (12) 49 (16) 0.66 (0.32–1.35)
Previous TIA 43 (20) 14 (15) 57 (19) 0.74 (0.38–1.42)
Ischemic heart disease 50 (23) 24 (26) 74 (24) 1.20 (0.68–2.10)
Heart disease 97 (45) 48 (53) 145 (47) 1.38 (0.85–2.26)
Educational level, n (%)
Low ≤9 years 107 (49) 49 (54) 156 (51) Ref
Middle ≥10 ≤12 years 53 (24) 18 (20) 71 (23) 0.74 (0.39–1.40)
High ≥12 years 57 (26) 24 (26) 81 (26) 0.92 (0.51–1.66)
TOAST classification, n (%)
CE 52 (24) 33 (36) 85 (28) 1.81 (1.06–3.06)
LAA 22 (10) 16 (18) 38 (12) 1.89 (0.94–3.80)
SAO 33 (15) 2 (2) 35 (11) 0.13 (0.03–0.53)
OC 6 (3) 0 6 (2) –
UND 104 (48) 40 (44) 144 (47) 0.85 (0.52–1.39)
OCSP, n (%)
LACI 55 (25) 4 (4) 59 (19) 0.14 (0.05–0.39)
PACI 93 (43) 44 (49) 137 (45) 1.25 (0.76–2.04)
POCI 52 (24) 10 (11) 62 (20) 0.39 (0.19–0.81)
TACI 17 (8) 33 (36) 50 (16) 6.69 (3.48–
12.87)

AF, atrial fibrillation; TOAST, Trial of Org in Acute Stroke Treatment [20]; NIHSS, total score on National Institutes
of Health Stroke Scale; IQR, interquartile range; OCSP, The Oxfordshire Community Stroke Project [21]; CE,
cardioembolism; LAA, large-artery atherosclerosis; SAO, small-artery occlusion; OC, other determined etiology;
UND, undetermined etiology; LACI, lacunar infarct; PACI, partial anterior circulation infarct; POCI, posterior
circulation infarct; TACI, total anterior circulation infarct. * Comparisons between patients without aphasia and
patients with aphasia according to NIHSS item 9. † Definitions of stroke risk factors according to [18, 19].

creased by 4% per each year of age among the IS patients justing for stroke severity (i.e., NIHSS excluding item 9),
(OR, 1.04; 95% CI: 1.02–1.06). However, after adjusting the difference of TOAST categories between stroke pa-
for stroke severity, this increase was barely significant tients with and without aphasia was no longer significant
(OR, 1.03; 95% CI: 1.00–1.05). (OR, 1.18; 95% CI: 0.61–2.29). There was no change in
association between stroke severity and aphasia also when
Stroke Classification excluding patients with SAO, a TOAST category that has
The most common TOAST category for all 308 IS pa- a prerequisite of no cortical dysfunction such as symp-
tients, was undetermined (n = 144, 47%) followed by CE toms of aphasia. Patients with aphasia more often pre-
(n = 85, 28%). The pathogenetic mechanism for 36% of sented with the OCSP syndrome total anterior circulation
patients with aphasia (n = 33) was CE, which was signifi- stroke (36% vs. 8%, p < 0.001) in comparison to stroke
cantly higher compared with stroke patients without patients without aphasia.
aphasia (OR, 1.81; 95% CI: 1.06–3.06). However, after ad-

4 Neuroepidemiology Grönberg/Henriksson/Stenman/Lindgren
DOI: 10.1159/000524206
 Fig. 2. Incidence rate of first-ever IS patients with and without aphasia across age groups year 2005–2006 and
2017–2018.

Temporal Changes – Incidence, Aphasia Severity, and (95% CI: 23–41 per 100,000 person-years) (p = 0.22).
Mortality between 2005–2006 and 2017–2018 Though not significant (p = 0.08), there was a trend to-
The overall IS incidence rate decreased with 36% ad- ward a higher proportion of women with aphasia in com-
justed to the ESP between 2005–2006 and 2017–2018, parison to men (26% men vs. 34% women) in 2017–2018.
from 169 per 100,000 person-years in 2005–2006 (95% In 2005–2006, there was no difference between men and
CI: 154–186 per 100,000 person-years) to 108 per 100,000 women (27% vs. 27%) with aphasia. In 2017–2018, stroke
person-years in 2017–2018 (95% CI: 97–121 per 100,000 severity (NIHSS) was higher for women in comparison to
person-years; Fig. 2). The proportion of stroke patients men (p = 0.04), whereas in 2005–2006 the stroke severity
with aphasia in the acute phase of stroke onset was 27% did not differ between men and women (p = 0.19). Com-
year 2005–2006 (95% CI: 23–32%) as compared to 30% paring the total cohort of patients, stroke severity accord-
year 2017–2018 (95% CI: 25–35%), indicating that there ing to NIHSS (including item 9) remained stable between
was no significant temporal change in aphasia incidence 2005 and 2006 (median NIHSS = 4) and 2017–2018 (me-
(p = 0.45). The total aphasia incidence rate adjusted to the dian NIHSS = 4; p = 0.44). There was no difference be-
ESP decreased with 30% (p = 0.01) from 44 per 100,000 tween aphasia severity year 2005–2006 and year 2017–
person-years in 2005–2006 (95% CI: 37–54 per 100,000 2018 (p = 0.35), likewise aphasia severity between genders
person-years) to 31 per 100,000 person-years in 2017– was equivalent 2005–2006 (p = 0.71) and year 2017–2018
2018 (95% CI: 25–38 per 100,000 person-years). The in- (p = 0.69). Among 91 patients with aphasia in 2017–2018,
cidence rate for males was 59 per 100,000 person-years in 36 had mild to moderate aphasia (39%; 95% CI: 30–50),
2005–2006 (95% CI: 46–74 per 100,000 person-years) and 27 had severe aphasia (30%; 95% CI: 29–40), and 28 had
32 per 100,000 person-years in 2017–2018 (95% CI: 23– global aphasia (31%; 95% CI: 22–41). The corresponding
42 per 100,000 person-years), indicating a significant de- figures in 2005–2006 were 49 with mild to moderate
crease in incidence rate for men with aphasia (p < 0.001). aphasia (50%; 95% CI: 40–60%), 23 with severe aphasia
The equivalent for women was 40 per 100,000 person- (23%; 95% CI: 16–33%) and 26 with global aphasia (27%;
years in 2005–2006 (95% CI: 30–53 per 100,000 person- 95% CI: 19–36%; Fig. 3). The proportion of all patients
years) and 31 per 100,000 person-years in 2017–2018 with acute IS who died at hospital during the acute phase

Severity, Etiology, and Temporal Change Neuroepidemiology 5

DOI: 10.1159/000524206
 Fig. 3. Proportions of aphasia severity relative to stroke severity 2005–2006 and 2017–2018.

after stroke onset was 4% (n = 14) in 2005–2006 com- ed in previous studies [12, 24], however, in our study,
pared to 6% (n = 19) year 2017–2018 (no significant dif- when adjusting for stroke severity (NIHSS excluding the
ference, p = 0.17). There was also no significant difference aphasia item 9), cardioembolism as the underlying stroke
in mortality in the group of stroke patients with aphasia mechanism according to TOAST was no longer signifi-
between year 2005–2006 and 2017–2018 (p = 0.08). How- cant. The probable risk of aphasia after stroke may there-
ever, patients with aphasia had significantly higher mor- fore be more related to stroke severity rather than the un-
tality in comparison to stroke patients without aphasia (p derlying stroke mechanism. Patients with isolated sub-
= 0.01), and this association remained between 2005– cortical infarcts may sometimes also have aphasia (e.g., if
2006 and 2017–2018. the lesion is in the thalamus), and this could be of interest
to investigate in more detail in future studies.
The risk of aphasia in IS increases with age [24], and we
Discussion can confirm that the odds of having aphasia increased by
4% per each year of age of stroke patients (OR, 1.04; 95%
The new data in this prospective study imply that de- CI: 1.02–1.06). However, in contrast, when adjusting for
spite a significant decrease in IS incidence rates during stroke severity (NIHSS excluding the aphasia item), we
the past decade, the proportion of patients with aphasia found that the risk of aphasia is primarily associated with
in acute IS remains stable at approximately 30%. Aphasia stroke severity (OR, 1.25; 95% CI: 1.18–1.32) rather than
was significantly associated with more severe strokes, an age. Nonetheless, aphasia is more frequent among older
association that remained after removing the aphasia than younger stroke patients; only every seventh person
item from the total NIHSS score (p < 0.001). This was re- with aphasia in our study was of working age (≤65), which
lated to higher mortality and longer hospital stays of per- is in line with data from previous studies [12]. The pres-
sons with aphasia and is in accordance with previous re- ence of vascular risk factors for stroke was equivalent be-
search [9]. tween stroke patients with or without aphasia even though
A higher proportion of patients with aphasia present- patients with aphasia had a tendency towards more often
ed with cardio-embolic stroke. This has also been report- having AF. AF is a strong risk factor for severe strokes [25]

6 Neuroepidemiology Grönberg/Henriksson/Stenman/Lindgren
DOI: 10.1159/000524206
and may contribute to the stable initial stroke severity (NI- As strengths, we used “hot pursuit methods” to pro-
HSS) of patients in our cohort between 2005–2006 and spectively include patients from a well-defined popula-
2017–2018 and the steady aphasia incidence at 30%. tion, during a defined time period. A high proportion of
The incidence rate of aphasia in first-ever IS was 31 per hospitalized IS patients was included in our study (91%).
100,000 person-years after adjusting to the ESP [22]. The In addition, the local uptake area has only one hospital for
incidence rate of aphasia (ESP) decreased with 30% (i.e., acute stroke admissions and the rate of hospital admis-
from 44 to 31 per 100,000 person-years) between 2005– sion for IS is high [16].
2006 and 2017–2018, which follows the decreasing stroke The use of NIHSS item 9 to identify aphasia can be dis-
incidence rate reported in this study and in our region cussed. However, we have previously validated NIHSS
[16]. The decreased incidence rate of aphasia in IS was item 9 for this purpose and shown that NIHSS has accept-
significantly more pronounced in men (decrease by 46%), able diagnostic accuracy for detecting aphasia after acute
while the observed trend towards decrease in women was stroke [30]. More comprehensive aphasia test batteries
nonsignificant (23%). This difference between men and are often too demanding for the acute stroke patient and
women is of concern; however, the wide 95% CI urge for have long administration times, making them difficult to
cautious interpretation and needs to be further investi- implement in the acute setting and leaving some patients
gated and confirmed in future studies. In contrast to pre- not being evaluated regarding aphasia.
vious epidemiological studies of aphasia [24, 26], we ob- We did not include recurrent stroke patients and pa-
served a trend toward a higher proportion of women hav- tients with aphasia caused by other mechanisms, conse-
ing aphasia in year 2017–2018 as compared to 2005–2006. quently our findings may underestimate the total overall
This may be explained by the change in demographics incidence rate of all patients with aphasia in the popula-
combined with the temporal change of stroke severity; tion. There have been temporal changes in acute stroke
where women had significantly more severe strokes in care in Sweden between 2005 and 2017 with an increase
year 2017–2018 in comparison to men. Future studies in- of recanalization treatment from 3% to 15% [31]. This
vestigating aphasia in relation to gender are warranted. may have reduced the long-term prevalence of aphasia
The in-hospital mortality after stroke was similar 2005– after stroke, as well as affected stroke morbidity. We did
2006 and 2017–2018 (p = 0.17), despite substantial chang- not study possible changes of stroke morbidity in detail
es in acute stroke treatment under the same time period. but the NIHSS in the acute phase was median 4 (IQR 2–8)
This may be related to the observed stable initial stroke in 2005–2006 and median 4 (IQR 2–7) in 2017–2018, in-
severity as discussed above. dicating that acute stroke severity did not differ between
The initial severity of aphasia remained unchanged be- the two periods. The Swedish Stroke Register (Riksstroke),
tween 2005–2006 and 2017–2018, with the majority of however, reports that at 3 months after stroke onset, 22%
patients (50–60%) suffering severe or global aphasia (NI- were dependent in their activities of daily living (ADL) in
HSS score on item 9 ≥ 2). This is of concern because initial 2005 compared to 17% year 2017 [31]. Even though not
severity of aphasia strongly predicts outcome [27, 28] and influencing the incidence of aphasia at stroke onset, cur-
aphasia represents one of the most devastating conse- rent advanced acute stroke treatments may therefore
quences after stroke with subsequent impact on quality of have effect on the subsequent outcome of aphasia and
life [28], and has even been reported to be a marker for studies examining this and how aphasia may be related to
unfavorable outcome in patients with mild stroke [29]. stroke morbidity and prognosis are warranted.
Our study has limitations: it is hospital-based and not In recent years, the effects of aphasia therapy on recov-
population-based which might infer a potential bias be- ery and the importance of intensive aphasia therapy on
cause aphasia may increase the probability of hospital ad- outcome have been highlighted [32, 33]. For the first
mission after IS stroke onset [12]. However, population- time, intensive aphasia therapy and communication part-
based studies have reported similar incidence of aphasia ner training was stressed in the stroke guidelines from
in stroke [12], even though there may have been difficul- Swedish National Board of Health and Welfare [34] in
ties with evaluating the presence of aphasia in patients not 2017–2018. Future research on improved aphasia treat-
being clinically examined at hospital. Exclusion of 30 pa- ment in relation to a potential reduced aphasia prevalence
tients may have affected our incidence; however, we tried after stroke is needed.
to mitigate the potential risk of selection bias with con- In conclusion, the incidence rate of aphasia after IS has
secutive inclusion of patients with a comprehensive range decreased, yet the proportion of IS patients with initial
of stroke symptoms. aphasia remains unchanged. The continued high inci-

Severity, Etiology, and Temporal Change Neuroepidemiology 7

DOI: 10.1159/000524206
dence of aphasia emphasizes the importance of consistent Funding Sources
language screening of stroke patients in the acute phase
This project is funded by The Swedish Research Council (2019-
to ensure optimal management. Future studies providing 01757), The Swedish Government (under the “Avtal om Läkarut-
data on prevalence and recovery of aphasia after the acute bildning och Medicinsk Forskning, ALF”), The Swedish Heart and
phase are warranted to ascertain up-to-date knowledge Lung Foundation, The Swedish Stroke Association, Region Skåne,
on the long-term prognosis and burden of aphasia. Lund University, Skåne University Hospital, and Sparbanksstif-
telsen Färs och Frosta, Fremasons Lodge of Instruction Eos in
Lund. Dr. Lindgren reports personal fees from Bayer, NovoNord-
isk, Astra Zeneca, BMS Pfizer, and Portola outside the submitted
Acknowledgment work.
We would like to thank statistician Anna Åkesson for her sup-
port in data analysis for this manuscript.
Author Contributions

All authors contributed to the study conception. Acquisition of

Statement of Ethics data was performed by Angelina Grönberg and Martin Stenman,
and analysis and interpretation of data was performed by all au-
The study was approved by the Regional Ethical Review Au- thors. Supervision was performed by Ingrid Henriksson and Arne
thority in Lund (registration number 2016/179). Written consent G. Lindgren. An original draft was composed by Angelina Grön-
was obtained from patients or from their next of kin to participate berg, and all authors revised previous versions of the manuscript
in the study. and approved the final manuscript.

Conflict of Interest Statement Data Availability Statement

The authors have no conflicts of interest to disclose. Data supporting the findings of this study are available from
the corresponding author upon reasonable request.

References
 1 Feigin VL, Krishnamurthi RV, Parmar P,   7 Lam JM, Wodchis WP. The relationship of 60 13 Tsouli S, Kyritsis AP, Tsagalis G, Virvidaki E,
Norrving B, Mensah GA, Bennett DA, et al. disease diagnoses and 15 conditions to prefer- Vemmos KN. Significance of aphasia after
Update on the global burden of ischemic and ence-based health-related quality of life in first-ever acute stroke: impact on early and
hemorrhagic stroke in 1990–2013: the GBD Ontario hospital-based long-term care resi- late outcomes. Neuroepidemiology. 2009;
2013 Study. Neuroepidemiology. 2015; 45(3): dents. Med Care. 2010;48(4):380–7. 33(2):96–102.
161–76.   8 Ali M, Lyden P, Brady M; VISTA Collabora- 14 SCB. Statistical database [Internet]. Stock-
  2 Modig K, Talbäck M, Ziegler L, Ahlbom A. tion. Aphasia and dysarthria in acute stroke: holm: population statistics; 2022. [cited 2020
Temporal trends in incidence, recurrence and recovery and functional outcome. Int J Stroke. Mar 27]. Available from: www.scb.se.
prevalence of stroke in an era of ageing popu- 2015;10(3):400–6. 15 Hatano S. Experience from a multicentre
lations, a longitudinal study of the total Swed-   9 Flowers HL, Skoretz SA, Silver FL, Rochon E, stroke register: a preliminary report. Bull
ish population. BMC Geriatr. 2019;19(1):31. Fang J, Flamand-Roze C, et al. Poststroke World Health Organ. 1976;54(5):541–53.
  3 Li L, Scott CA, Rothwell PM; Oxford Vascular aphasia frequency, recovery, and outcomes: a 16 Aked J, Delavaran H, Norrving B, Lindgren A.
Study. Trends in stroke incidence in high-in- systematic review and meta-analysis. Arch Temporal trends of stroke epidemiology in
come countries in the 21st century: popula- Phys Med Rehabil. 2016;97(12):2188–201.e8. Southern Sweden: a population-based study
tion-Based Study and systematic review. 10 Kadojic D, Bijelic BR, Radanovic R, Porobic on stroke incidence and early case-fatality.
Stroke. 2020;51(5):1372–80. M, Rimac J, Dikanovic M. Aphasia in patients Neuroepidemiology. 2018;50(3–4):174–82.
  4 Stevens E, Emmett E, Wang Y, McKevitt C, with ischemic stroke. Acta Clin Croat. 2012; 17 Anemaet WK. Using standardized measures
Wolfe C. The burden of stroke in Europe. 51(2):221–5. to meet the challenge of stroke assessment.
Stroke Alliance for Europe; 2017. p. 131. 11 Flowers HL, Silver FL, Fang J, Rochon E, Mar- Top Geriatr Rehabil. 2002;18(2):47–62.
  5 Teasell R, Bitensky J, Salter K, Bayona NA. tino R. The incidence, co-occurrence, and 18 Starby H, Delavaran H, Andsberg G, Lövkvist
The role of timing and intensity of rehabilita- predictors of dysphagia, dysarthria, and apha- H, Norrving B, Lindgren A. Multiplicity of
tion therapies. Top Stroke Rehabil. 2005; sia after first-ever acute ischemic stroke. J risk factors in ischemic stroke patients: rela-
12(3):46–57. Commun Disord. 2013;46(3):238–48. tions to age, sex, and subtype – a study of
 6 Nesi M, Lucente G, Nencini P, Fancellu L, 12 Engelter ST, Gostynski M, Papa S, Frei M, 2,505 patients from the lund stroke register.
Inzitari D. Aphasia predicts unfavorable out- Born C, Ajdacic-Gross V, et al. Epidemiology Neuroepidemiology. 2014;42(3):161–8.
come in mild ischemic stroke patients and of aphasia attributable to first ischemic stroke: 19 Jackson CA, Hutchison A, Dennis MS, Ward-
prompts thrombolytic treatment. J Stroke incidence, severity, fluency, etiology, and law JM, Lindgren A, Norrving B, et al. Differ-
Cerebrovasc Dis. 2014;23(2):204–8. thrombolysis. Stroke. 2006;37(6):1379–84. ing risk factor profiles of ischemic stroke sub-
types. Stroke. 2010;41(4):624–9.

8 Neuroepidemiology Grönberg/Henriksson/Stenman/Lindgren
DOI: 10.1159/000524206
20 Adams HP Jr, Bendixen BH, Kappelle LJ, Bill- 25 Zulkifly H, Lip GYH, Lane DA. Epidemiology 31 The Swedish Stroke Register (RiksStroke).
er J, Love BB, Gordon DL, et al. Classification of atrial fibrillation. Int J Clin Pract. 2018; The Riksstroke annual report 2005 and 2017
of subtype of acute ischemic stroke. Defini- 72(3):e13070. [Internet]. 2022 [cited 2022 Mar 2]. Available
tions for use in a multicenter clinical trial. 26 Inatomi Y, Yonehara T, Omiya S, Hashimoto from: https: //www.riksstroke.org/sve/for-
TOAST. Trial of org 10172 in acute stroke Y, Hirano T, Uchino M. Aphasia during the skning-statistik-och-verksamhetsutveckling/
treatment. Stroke. 1993;24(1):35–41. acute phase in ischemic stroke. Cerebrovasc rapporter/arsrapporter/.
21 Bamford J, Sandercock P, Dennis M, Burn J, Dis. 2008;25(4):316–23. 32 Breitenstein C, Grewe T, Flöel A, Ziegler W,
Warlow C. Classification and natural history 27 Lazar RM, Minzer B, Antoniello D, Festa JR, Springer L, Martus P, et al. Intensive speech
of clinically identifiable subtypes of cerebral Krakauer JW, Marshall RS. Improvement in and language therapy in patients with chronic
infarction. Lancet. 1991;337(8756):1521–6. aphasia scores after stroke is well predicted by aphasia after stroke: a randomised, open-la-
22 Eurostat. Revision of the European standard initial severity. Stroke. 2010;41(7):1485–8. bel, blinded-endpoint, controlled trial in a
population report of Eurostat’s task force. 28 Osa García A, Brambati SM, Brisebois A, health-care setting. Lancet. 2017; 389: 1528–
Luxembourg: Publications Office of the Euro- Désilets-Barnabé M, Houzé B, Bedetti C, et al. 38.
pean Union; 2013. Predicting early post-stroke aphasia outcome 33 Brady MC, Kelly H, Godwin J, Enderby P,
23 Dean AG, Sullivan KM, Soe MM. OpenEpi: from initial aphasia severity. Front Neurol. Campbell P. Speech and language therapy for
open source epidemiologic statistics for pub- 2020;11:120. aphasia following stroke. Cochrane Database
lic health [Internet]. 2021 [cited 2021 Nov 19]. 29 Kremer C, Kappelin J, Perren F. Dissociation Syst Rev. 2016. 2016(6):CD000425.
Available from: www.OpenEpi.com. of severity of stroke and aphasia recovery ear- 34 The National Board of Health and Welfare
24 Dickey L, Kagan A, Lindsay MP, Fang J, Row- ly after intravenous recombinant tissue plas- (Socialstyrelsen). National stroke care guide-
land A, Black S. Incidence and profile of inpa- minogen activator thrombolysis. J Clin Neu- lines [Internet]. 2022 [cited 2022 Mar 2].
tient stroke-induced aphasia in Ontario, Can- rosci. 2014;21(10):1828–30. Available from: https: //www.socialstyrelsen.
ada. Arch Phys Med Rehabil. 2010;91(2):196– 30 Grönberg A, Henriksson I, Lindgren A. Ac- se/kunskapsstod-och-regler/regler-och-rik-
202. curacy of NIH stroke scale for diagnosing tlinjer/nationella-riktlinjer/riktlinjer-och-ut-
aphasia. Acta Neurol Scand. 2021; 143(4): varderingar/stroke/.
375–82.

Severity, Etiology, and Temporal Change Neuroepidemiology 9

DOI: 10.1159/000524206