17753

2014
NNRXXX10.1177/1545968313517753Neurorehabilitation and Neural Repairvan der Meulen et al

Clinical Research Article

Neurorehabilitation and

The Efficacy and Timing of Melodic

Neural Repair
2014, Vol. 28(6) 536­–544
© The Author(s) 2014
Intonation Therapy in Subacute Aphasia Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1545968313517753
nnr.sagepub.com

Ineke van der Meulen, PhD1,2, W.Mieke.E. van de

Sandt-Koenderman, PhD1,2, Majanka H. Heijenbrok-Kal, PhD1,2,
Evy G. Visch-Brink, PhD3, and Gerard M. Ribbers, MD, PhD1,2

Abstract
Background. Little is known about the efficacy of language production treatment in subacute severe nonfluent aphasia.
Although Melodic Intonation Therapy (MIT) is a language production treatment for this disorder, until now MIT effect
studies have focused on chronic aphasia. Purpose. This study examines whether language production treatment with MIT
is effective in subacute severe nonfluent aphasia. Methods. A multicenter, randomized controlled trial was conducted in
a waiting-list control design: patients were randomly allocated to the experimental group (MIT) or the control group
(control intervention followed by delayed MIT). In both groups, therapy started at 2 to 3 months poststroke and was given
intensively (5 h/wk) during 6 weeks. In a second therapy period, the control group received 6 weeks of intensive MIT. The
experimental group resumed their regular treatment. Assessment was done at baseline (T1), after the first intervention
period (T2), and after the second intervention period (T3). Efficacy was evaluated at T2. The impact of delaying MIT on
therapy outcome was also examined. Results. A total of 27 participants were included: n = 16 in the experimental group
and n = 11 in the control group. A significant effect in favor of MIT on language repetition was observed for trained
items, with mixed results for untrained items. After MIT there was a significant improvement in verbal communication
but not after the control intervention. Finally, delaying MIT was related to less improvement in the repetition of trained
material. Conclusions. In these patients with subacute severe nonfluent aphasia, language production treatment with MIT
was effective. Earlier treatment may lead to greater improvement.

Keywords
aphasia, stroke, rehabilitation, language production treatment

Introduction is thought to be too frustrating, to our knowledge this claim

Aphasia is a common consequence of stroke; the incidence
in a stroke population ranges from 21% to 38%.1-3 The
recovery pattern from poststroke aphasia shows much vari-
ability: some patients recover quickly and experience no (or
only mild) language problems within a few weeks post-
stroke, whereas others remain severely nonfluent, that is,
they remain completely or almost unable to produce lan-
guage.3-6 In the past decades, the evidence for the effective-
ness of aphasia treatment has increased.7-11 However, most
studies focused on treatment effects in chronic populations.
Only a few examined aphasia treatment in the subacute
phase poststroke, when treatment interacts with spontane-
ous recovery processes.8,10,12-15
This study addresses language production treatment for
patients with severe nonfluent aphasia, persisting until 2 to
3 months poststroke. These patients generally receive a
combination of exercises aiming at language production,
language comprehension, and nonverbal communication
strategies. Although a focus on language production alone
has yet not been examined.
Melodic Intonation Therapy (MIT)16 is a language pro-
duction treatment for severe nonfluent aphasia. It is based
on the observation that these patients are often able to sing
words they cannot produce during speech. The treatment
involves repetitive singing of short sentences, while hand
tapping the rhythm. Originally, it was claimed that melody
activates language-capable regions in the right hemi-
sphere.16-19 However, recent evidence highlights the critical
role of rhythm and formulaic language in MIT.20 The
1
Rijndam Rehabilitation Center, Rotterdam, the Netherlands
2
Erasmus MC, University Medical Center, Department of Rehabilitation
Medicine, Rotterdam, the Netherlands
3
Erasmus MC, University Medical Center. Department of Neurology,
Rotterdam, the Netherlands
Corresponding Author:
Ineke van der Meulen, Rijndam Rehabilitation Center, Westersingel 300,
3015 LJ, Rotterdam 3015 LJ, the Netherlands.
Email: ivandermeulen@rijndam.nl
van der Meulen et al
Figure 1. Flow diagram.
contribution of the right hemisphere is still unclear: whereas
some report increased right hemisphere activation related to
MIT success,21,22 others suggest that MIT-induced language
recovery is related to reactivation of left perilesional
regions.23,24 Many studies have shown the beneficial effects
of MIT on language production.19,21,22,25,26 However, most
are single-case or case series studies in chronic patients.27,28
A recent pilot study investigated the effect of a modified
form of MIT in subacute patients with mild to severe apha-
sia and reported positive effects immediately after one short
therapy session.29 Thus, overall, the level of evidence for
MIT is low and little is known about its effect in early
phases poststroke.
To examine the efficacy of MIT as a therapy to improve
language production, this method was contrasted with a
control therapy not aiming at language production but using
linguistic tasks often trained in severe nonfluent aphasia,
such as written language production, language comprehen-
sion, and nonverbal communication strategies.
The aim of the present study was 3-fold. First, the effi-
cacy of MIT as a language production therapy for severe
nonfluent aphasia was evaluated in the subacute phase.
Second, we investigated whether the timing of MIT within
the subacute phase affects therapy outcome; it is suggested
537
that early aphasia intervention yields greater improvement
but, until now, the evidence for early treatment is not well
established.10,15 Third, we examined potential determinants
influencing MIT outcome.
Methods
Design
A waiting-list randomized controlled design was used
(Figure 1). Between baseline (T1) and the first intervention
period (T2), participants in the experimental group received
intensive MIT (6 weeks; 5 h/wk); no other language therapy
was allowed in this period. In the same way, participants in
the control group received intensive control treatment only
(6 weeks; 5 h/wk), thereby allowing comparison between
MIT and the control therapy at T2. After T2, patients allo-
cated to the control group received delayed MIT following
the same protocol (6 weeks; 5 h/wk), allowing to examine
the effect of timing of MIT. Patients in the experimental
group resumed their regular therapy after T2.
Patients were randomly allocated to either MIT or the
control group. For this, a computer-generated random allo-
cation sequence was used and the results placed in
538
consecutively numbered sealed envelopes. The study was
approved by the Medical Ethics Committee of Erasmus
University Medical Center. Written informed consent was
obtained from all participants or close relatives.
For obvious reasons, participants and speech-language
therapists (SLTs) could not be blinded for treatment condi-
tion. The researchers administering and scoring the assess-
ments at each test moment were blinded for group allocation.
In a few cases, blinding could not be maintained because
the patients spontaneously informed the researcher about
their therapy allocation.
Participants
Between 2009 and 2011, patients were recruited from 15
aphasia treatment services in hospitals, rehabilitation cen-
ters, and nursing homes in the Netherlands. Inclusion crite-
ria were: aphasic after left hemisphere stroke, time
poststroke 2 to 3 months, premorbidly right-handed, age 18
to 80 years, native language Dutch, and MIT candidate.
MIT candidacy was based on the MIT literature19,30 and
defined as follows: nonfluent aphasia (<50 words/min),
articulation deficits (Aachen Aphasia Test31 [AAT], sub-
score spontaneous language ≤3), repetition severely affected
(AAT subtest repetition ≤100), and moderate to good audi-
tory language comprehension (AAT subtest auditory com-
prehension ≥33; functional comprehension ≥5). Exclusion
criteria were: prior stroke resulting in aphasia, bilateral
lesion, intensive MIT prior to start of the study, severe hear-
ing deficit, and psychiatric history relevant to language
communication.
Interventions
All interventions were given by the patients’ SLT, experi-
enced in language rehabilitation in aphasia. MIT was
applied following the American manual.30,32 All SLTs were
trained to deliver MIT according to the therapy protocol.
The patient and the SLT sang short utterances together,
while hand tapping the rhythm. Gradually, the support from
the SLT decreased and singing was replaced by speaking.
The study protocol listed a set of utterances of increasing
complexity to be trained. The first utterances were selected
because of their frequent use in daily-life communication
(eg, “coffee please”). Later in the program, the utterances
became longer, more complex, and less frequent in daily
life. In addition, the SLT and the patient composed a set of
self-chosen utterances that were functionally relevant to the
patient, such as utterances related to hobbies. A minimum of
50% of the therapy time had to be spent on the utterances
provided in the protocol.
The control intervention did not emphasize spoken out-
put but focused on other linguistic modalities usually
trained in severe nonfluent aphasia (writing, language
Neurorehabilitation and Neural Repair 28(6)
comprehension, nonverbal communication strategies).
Spoken output was not discouraged but the therapists did
not provide feedback regarding patients’ verbal production
and offered no structural training of language production.
To ensure therapy intensity, homework assignments
were provided for both the MIT and the control group. We
developed an iPod application containing short videos of a
mouth singing the target utterances; patients could sing
along with the video or repeat the utterance afterwards.
Homework assignments for the control group included
paper-and-pencil tasks such as written sentence completion,
word–picture matching, and word categorizing tasks. The
minimum amount of face-to-face therapy time was 3 h/wk.
Therapists recorded therapy time per session, and patients
or a close relative recorded homework time per session.
Regular practice (T2-T3 for the early MIT group,
Figure 1) depended on the needs and capabilities of the
individual patient. Most patients received a combination of
language production therapy (eg, word-finding therapy),
semantic therapy, and nonverbal communication strategies.
Treatment intensity was not recorded.
Assessment
Prior to inclusion, the AAT was administered to establish
inclusion eligibility. Assessments were performed at base-
line (T1, within 2 weeks from the AAT), after the first treat-
ment phase of 6 weeks (T2), and 6 weeks later (T3)
(Figure 1). These assessments included the following tests:
the Sabadel story retelling task measuring information con-
tent in connected speech33; the Amsterdam Nijmegen
Everyday Language Test (ANELT) measuring verbal com-
munication in daily life34; the AAT subtests repetition and
naming; the MIT repetition task, a repetition task designed
for the present study including 11 utterances trained during
MIT and 11 matched untrained utterances; and the nonverbal
Semantic Association Task (SAT) measuring semantic disor-
ders.35 All language production tests were audio recorded.
Outcome measures were the Sabadel, the ANELT, the
AAT subtests repetition and naming, and the MIT repetition
task. The MIT repetition task allowed comparing the effects
for trained and untrained material, that is, the direct effect of
the MIT training on spoken repetition of the trained utter-
ances and the generalization to untrained material.
Generalization to the repetition of untrained material was
also examined by the AAT subtest repetition. The AAT
naming task was used to investigate further generalization
to word finding and word production. The ANELT and the
Sabadel were used to examine generalization to functional
language use, respectively, in everyday communicative sit-
uations and in story retelling. In the ANELT, the researcher
presents an everyday communicative situation, for exam-
ple, a doctor’s visit. The patient’s task is to produce an ade-
quate, verbal reaction. The Sabadel evaluates the production
van der Meulen et al 539

of connected speech. The patient's task is to retell a story Table 1. Baseline Characteristics of the 2 Study Groups.
immediately after it has been presented by the researcher, Experimental (n = 16), Control (n = 11),
supported by pictures. Mean (SD) Mean (SD)

Age, years 53.1 (12.0) 52.0 (6.6)

Gender, % male 25% 63.6%
Statistical Analysis Educationa 5.0 (2.0) 6.5 (1.6)
Time poststroke, weeks 9.3 (2.0) 11.9 (5.9)
The power analysis was based on a small Sabadel pilot
Stroke type
study.36 From this, we calculated that a sample size of 15 Hemorrhagic, % 6.3% 9.1%
patients per group was needed to provide 80% power Ischemic, % 87.5% 81.8%
(α = .05, β = .20) to detect a mean difference in improve- Unknown, % 6.3% 9.1%
Stroke localization, % LH 100% 100%
ment of 11.5 (standard deviation [SD] = 12.42) content
Handedness, % right handedb 100% 100%
information units (CIUs) on the Sabadel with an expected AAT Token Test31 38.8 (12.5) 38.4 (9.3)
effect size of .90 (Cohen’s d). CIUs are words that are ade- AAT language repetition 44.8 (33.1) 38.2 (27.0)
quate and comprehensible in relation to the target story and AAT auditory comprehension 40.0 (5.9) 42.3 (7.2)
AAT naming 5.6 (15.0) 2.7 (5.7)
are often used to assess communicative efficiency in apha- ANELT34 13.0 (5.1) 12.7 (5.9)
sia.37 We aimed to recruit 20 patients per intervention group, Sabadel, CIUs37 15.5 (22.3) 12.6 (19.5)
thereby taking into account that not all patients would com- MIT repetition 26.7 (26.7) 21.1 (29.2)
plete the intervention.
Abbreviations: SD, standard deviation; LH, left hemisphere; AAT, Aachen Aphasia
Analyses were performed on an intention-to-treat basis. Test; ANELT, Amsterdam Nijmegen Everyday Language Test; CIU, content
Independent t tests for continuous data and χ2 tests for cat- information units; MIT, Melodic Intonation Therapy.
a
Level of education: 1 = lowest (primary school), 8 = highest (university).
egorical data were used to test group differences at b
Handedness before stroke (Edinburgh Handedness Inventory and/or medical
baseline. information).
To evaluate the efficacy of MIT at T2, univariable linear
regression analyses, adjusted for baseline, were used for all
outcome measures. Furthermore, the proportion of partici- MIT. Information on the proportion of screened patients eli-
pants in each group showing a clinically relevant improve- gible for inclusion in the study was only available from the
ment on the ANELT (>7) between T1 and T2 was compared main study center (64% eligible for inclusion); here, the
by means of a χ2 test. reasons for nonparticipation were failure to meet the inclu-
To examine the impact of timing, we used a linear mixed sion criteria (66.7%), early discharge (22.2%), and refusal
model analysis with repeated measurements, analyzing pos- to participate (11.1%).
sible between-groups differences over the total intervention A total number of 27 patients were included in the study:
time (T1, T2, T3), taking into account correlations within 16 were allocated to the experimental group and 11 to the
subjects. control group. Four patients withdrew from MIT after 1 or
For analysis of the determinants, the data on all patients 2 weeks, because they felt uncomfortable with the therapy
pre- and post-MIT were taken into account (T1-T2 early or were disappointed by their progress. Thus, the required
MIT, T2-T3 delayed MIT). Potential determinants, that is, number of 15 patients per group was not achieved. Figure 1
age, gender, severity of the aphasia (score AAT Token Test), presents the CONSORT diagram of patient flow.
treatment intensity, time poststroke at start of MIT, patients’ Table 1 presents patient characteristics. Except for gen-
linguistic profile at the start of MIT: preMIT scores on AAT der (χ2 = 4.03, p = .045), at baseline there were no signifi-
language repetition, AAT auditory comprehension, and the cant differences between the 2 groups.
nonverbal SAT were examined for all outcome measures by
means of univariable linear regression analyses. Efficacy of MIT
Furthermore, scores on all outcome measures were dichoto-
mized into 2 groups: responders (improvement >10 on MIT There was no significant difference in treatment intensity
repetition, >14 on AAT repetition, >16 on AAT naming, >7 between the 2 groups (MIT: mean = 6.52 h/wk [SD = 3.55];
on the ANELT, >0 on the Sabadel) and nonresponders. control: mean 5.67 h/wk [SD = 1.41]; t = −.71, p = .49).
Group differences were examined using χ2 tests and inde- Table 2 presents the difference scores of both groups
pendent t tests. All analyses were performed using the SPSS between T1 and T2. At T2, the MIT group showed signifi-
version 18.0. cant improvement on all tasks, except for the Sabadel task.
The control group showed significant improvement only on
the repetition of untrained MIT items.
Results The linear regression analysis revealed a significant dif-
ference in improvement at T2 between the 2 groups for the
Participants MIT repetition test (trained items) and on the AAT subtest
In each center, all aphasia patients were screened by the repetition. Furthermore, a trend was observed for one func-
SLT to establish whether they fit the clinical picture for tional task: the ANELT (Table 2). Because of the gender
540 Neurorehabilitation and Neural Repair 28(6)

Table 2. Changes Over the Intervention Period T1 to T2 on All Outcome Measures and Group Comparisons Adjusted for Baselinea.

Intervention Period, T1-T2

Experimental Group: MIT Control Group: Non-MIT

(Δ T2-T1) (Δ T2-T1) Group Comparison

Mean SD p Mean SD p β p
Sabadel 6.1 13.9 .13 5.2 11.6 .17 1.2 .82
ANELT34 6.6 6.9 <.01 2.3 5.4 .20 4.1 .07
Naming (AAT)31 20.5 20.1 <.01 5.0 18.7 .48 15.6 .10
Repetition (AAT) 28.5 21.6 <.001 11.8 17.4 .06 17.2 .05
MIT-repetition 27.5 17.9 <.001 8.1 11.8 .05 18.3 <.01
trained items 17.6 11.8 <.001 2.3 5.6 .16 15.0 <.01
untrained items 9.9 7.8 <.001 5.8 7.3 .03 3.3 .25

Abbreviations: MIT, Melodic Intonation Therapy; SD, standard deviation; ANELT, Amsterdam Nijmegen Everyday Language Test; AAT, Aachen Aphasia
Test.
a
Positive mean scores represent an increase in outcome score over the intervention period. A positive β indicates more effect in the experimental
group than in the control group. A negative β indicates more effect in the control group than in the experimental group. Significant values are
represented in bold.

difference between the 2 groups, we controlled for this
potentially confounding variable; in addition we controlled
for aphasia severity. In both cases, this did not alter the
results.
The mean improvement of 6.6 points in the MIT group
approaches the clinically relevant improvement of >7 points
as specified in the ANELT34 at the individual level. In the
experimental group, 35.7% of the participants showed an
improvement >7 on the ANELT, considerably more than in
the control group (9.1%). However, this difference did not
reach significance (χ2 = 2.39, p = .12, Fisher’s exact test
p = .18).
Differences Over Time
The linear mixed model analysis showed a main effect of
time on all outcome measures: Sabadel: F = 5.49, p = .011;
ANELT: F = 7.82, p = .003; AAT naming: F = 11.37,
p = .001; AAT repetition: F = 16.33, p < .001; MIT repeti-
tion trained items: F = 26.62, p < .001; MIT repetition
untrained items: F = 17.19, p < .001. This effect of time was
present on the repetition tasks for both groups, but only for
the experimental group on the more functional tasks:
Sabadel (experimental group F = 5.30, p = .02; control
group F = 1.46, p = .28), ANELT (experimental group
F = 8.81, p = .004; control group F = 1.21, p = .34) and
naming (experimental group F = 19.92, p < .001; control
group F = 1.77, p = .27). Thus, whereas both groups
improved over time on language repetition, only the experi-
mental group showed significant improvement over time on
the functional tasks. The analysis revealed an interaction
between time and intervention group for repetition of
trained items (F = 8.89, p = .001). Over time, the experi-
mental group improved more on the repetition of trained
items than the control group.
Although the analysis does not show interaction on any
of the other variables, visual inspection of the improvement
patterns (Figure 2) shows that the results are in favor of the
experimental group for all measures.
The differences do not reach significance because the
study is underpowered.
Determinants for Therapy Outcome
Of all potential determinants, only treatment intensity and
time post onset had an impact on one or more outcome vari-
ables. Treatment intensity predicted outcome on the repeti-
tion of trained items, MIT task (β = .04, p = .02). Time
poststroke at the start of MIT predicted outcome on
untrained items, MIT task (β = −.68, p = .01), on AAT rep-
etition (β = −1.54, p = .02), and on the ANELT (β = −.46,
p = .04). The earlier MIT was started, the greater the
improvement on these outcome measures. No significant
differences were found between the groups of responders
and nonresponders.
Discussion
This study shows that training language production with
MIT has a beneficial effect on language production in
severe nonfluent aphasia in the subacute phase poststroke.
The experimental group, receiving early MIT, showed sig-
nificant improvement on all outcome measures except for
the Sabadel. Furthermore, their improvement in language
repetition was significantly greater than that in the control
group, receiving control therapy of the same intensity from
the same time poststroke. This effect was present for trained
(MIT test) and untrained material (AAT subtest repetition),
indicating a generalization to untrained material. Finally,
the considerable difference between the MIT and the
van der Meulen et al
Figure 2. Improvement over time: mean score ± 1 SD.
control group in improvement on verbal communication
(as measured by the ANELT) provides support for general-
ization of these capabilities to verbal communication in
daily life.
The study is too underpowered to obtain significant
effects on most of the outcome measures, which is a clear
limitation of the study. However, all observed differences,
541
although not significant, are in favor of the MIT group. This
suggests that in a larger sample size significant differences
would be found. A larger study verifying our observations is
therefore worthwhile.
The role of treatment intensity is worth considering.
Several studies have shown a relation between treatment
intensity and treatment effect: higher intensity yields larger
542
treatment effects.7,8,10,12 In this study, we chose a high treat-
ment intensity that is clinically feasible in the subacute
stage poststroke, both in terms of patient burden and in the
context of a rehabilitation program that entails other thera-
pies (eg, physiotherapy) as well. As such, the results of the
study are relevant for clinical practice. However, it is pos-
sible that with higher treatment intensity larger treatment
effects and generalization to daily life communication
would have been observed.
These results provide support for the efficacy of lan-
guage production training with MIT in subacute severe non-
fluent aphasia. Contrary to the belief of many clinicians that
a focus on language production is too frustrating for patients
with subacute severe nonfluent aphasia, this study shows
that intensive language production training is possible and
effective in this population. However, our study does not
indicate that MIT is the best way to achieve improved lan-
guage production in this group. A direct comparison
between different language production interventions is
needed to resolve this issue. Similarly, it is possible that an
adaptation of the MIT technique might yield better results.
Several therapeutic elements of MIT may be responsible for
its effects on language production: melody, rhythm, hand
tapping, or reduction of speed in singing versus speak-
ing.20,21,38,39 A recent study comparing the effect of melody
and rhythm on language production in nonfluent aphasia
showed that melody had no additional effect over rhythm.20
In the present study, it was impossible to unravel the impact
of the MIT components, since we used the original MIT
technique. Similarly, our study allows no conclusions
regarding the role of formulaic language.20 MIT involved
both formulaic (eg, “How are you?”) and nonformulaic
utterances (eg, “The ministers are talking nonsense”).
Our 2 tasks for functional language (the Sabadel and
ANELT) showed different results. In contrast to the
improvement of everyday life communication seen with the
ANELT, the Sabadel failed to show improvement in either
of the intervention groups. However, it is possible that this
task is not suitable for measuring verbal communication in
people with severe nonfluent aphasia; storytelling is known
to be extremely difficult for severely aphasic patients.40
In this study, many participants (48%) were unable to pro-
duce more than one adequate word on the Sabadel story
retelling, both before and after the intervention period. In
contrast, these same patients were able to produce adequate
words and utterances on the ANELT.
The observed contrast between trained and untrained
material in the MIT repetition task is clinically relevant.
From the start of MIT it was emphasized that personally
relevant utterances should be trained. The results of this
study underline the importance of carefully selecting the
target utterances. One of the limitations of this study is that
we did not examine patients’ use of the trained utterances in
their daily life communication. This would require partner
Neurorehabilitation and Neural Repair 28(6)
questionnaires, which are not very reliable. Anecdotic
reports from partners suggest that patients did benefit from
the improvement on trained utterances in daily life; this is in
line with the results of Stahl et al.20
Remarkably, the present study also showed that a small
difference in the timing of MIT had considerable conse-
quences for its effect. A delay of merely 6 weeks was related
to less improvement. Although the difference between early
and delayed MIT was only significant for the repetition of
trained items, the overall larger improvement in the early
MIT group (Figure 2) suggests that timing does affect ther-
apy outcome. The earlier application of MIT may have had
more interaction with the processes of spontaneous recov-
ery, which mainly occur during the first 3 months after
stroke.3,4,6,41 This is a challenging idea on the effect of tim-
ing of an intervention on neuromodulatory effects during
recovery. The timing of aphasia treatment remains an
important but unresolved question. A meta-analysis showed
that aphasia treatment in the first 3 months after stroke
yields larger effect sizes than treatment in later phases.10 In
contrast, a recent study reported no additional beneficial
effect of aphasia treatment in the first 4 months after
stroke.15 However, this latter study examined aphasia treat-
ment in an unselected group of aphasic patients, with het-
erogeneous and low-intensity treatment paradigms. To our
knowledge, besides the present study, no other study has
evaluated the effect of delaying a specific treatment; never-
theless, clinically, this is a highly relevant issue.
All participants fit the reported criteria for MIT candi-
dacy, that is, nonfluent aphasia after left hemisphere lesion,
language repetition severely disordered, and relatively good
auditory comprehension.19,30 Nevertheless, large individual
differences with respect to MIT success were observed. To
implement MIT more effectively in clinical practice, we
examined potential determinants influencing therapy out-
come. However, we were unable to detect any determinants
in patients’ profile before MIT. Earlier studies reported age
and initial aphasia severity to be important predictors for
language recovery; however, these results are inconsistent
and the relation between these factors and the type of inter-
vention is not clear.3,5,11,41 Neurological variables, such as
size and location of the lesion, may play an important
role.3,5,41,42 A limitation of the present study is that lesion
characteristics were not documented in detail. Although
information on lesion was collected from the participants’
medical records (scans and/or scan reports), these records
often lacked information on the exact size and location of
the lesion. Because routine scans were made shortly after
the stroke, at the moment of hospitalization many of the
scans showed no structural damage, and no reliable infor-
mation on the lesion was available at the start of MIT.
Another limitation is the lack of follow-up measure-
ments. Stahl et al20 suggested that the effects of MIT are
stable during a 3-month period after treatment completion.
van der Meulen et al
In conclusion, the present study shows that intensive oral
language production training is possible and effective in
subacute severe nonfluent aphasia.
Acknowledgments
We thank all therapists and evaluators in the participating centers
for their participation and commitment.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest
with respect to the research, authorship, and/or publication of this
article: Van der Meulen and Van de Sandt-Koenderman are pre-
paring a Dutch version of the MIT treatment program, to be pub-
lished by Bohn Stafleu van Loghum, the Netherlands. The
publisher has had no influence on the data collection, methods,
interpretation of the data, and final conclusions.
Funding
The author(s) disclosed receipt of the following financial support
for the research, authorship, and/or publication of this article: This
study was supported by the Stichting Rotterdams Kinderrevalidatie
Fonds Adriaanstichting (Grant No. 2007/0168 JKF/07.08.31
KFA).
References
1. Dickey L, Kagan A, Lindsay M, Fang J, Rowland A, Black S.
Incidence and profile of inpatient stroke-induced aphasia in
Ontario, Canada. Arch Phys Med Rehabil. 2010;91:196-202.
2. Engelter S, Gostynski M, Papa S, et al. Epidemiology of
aphasia attributable to first ischemic stroke. Stroke. 2006;37:
1379-1384.
3. Pedersen PM, Jorgensen HS, Nakayama H, Raascha HO,
Olsen TS. Aphasia in acute stroke: incidence, determinants,
and recovery. Ann Neurol. 1995;38:659-666.
4. Bakheit A, Shaw S, Carrington S, Griffiths S. The rate
and extent of improvement with therapy from different
types of aphasia in the first year after stroke. Clin Rehabil.
2007;21:941-949.
5. Lazar R, Speizer A, Festa J, Krakauer J, Marshall R.
Variability in language recovery after first-time stroke.
J Neurol Neurosurg Psychiatry. 2008;79:530-534.
6. El Hachioui H, Lingsma H, Van de Sandt-Koenderman WME,
Dippel D, Koudstaal P, Visch-Brink EG. Recovery from
aphasia after stroke: a one year follow-up study. J Neurol.
2013;260:166-171.
7. Bhogal SJ, Teasell R, Speechley M. Intensity of aphasia ther-
apy, impact on recovery. Stroke. 2003;34:987-993.
8. Brady M, Kelly H, Godwin J, Enderby P. Speech and language
therapy for aphasia following stroke. Cochrane Database Syst
Rev. 2012;(5):CD000425. doi:10.1002/14651858.CD000425.
pub3.
9. Cicerone KD, Dahlberg C, Malec JF, et al. Evidence-
based cognitive rehabilitation: Updated review of the lit-
erature from1998 through 2002. Arch Phys Med Rehabil.
2005;86:1681-1692.
543
10. Robey RR. A meta-analysis of clinical outcomes in the treat-
ment of aphasia. J Speech Lang Hear Res. 1998;41:172-187.
11. Salter K, Teasell R, Bhogal SJ, Zettler L, Foley N. Evidence-
based review of stroke rehabilitation. Module 14: Aphasia.
http://www.ebrsr.com/uploads/Aphasia-SREBR-SREBR-
15_1.pdf. Accessed July 2, 2013.
12. Bakheit A, Shaw S, Barrett L, et al. A prospective, random-
ized, parallel group, controlled study of the effect of intensity
of speech and language therapy on early recovery from post-
stroke aphasia. Clin Rehabil. 2007;21:885-894.
13. de Jong-Hagelstein M, Van de Sandt-Koenderman WME,
Prins N, Dippel D, Koudstaal P, Visch-Brink EG. The effi-
cacy of early cognitive-linguistic treatment and communica-
tive treatment in aphasia after stroke: a randomized controlled
trial (RATS-2). J Neurol Neurosurg Psychiatry. 2011;82:
399-404.
14. Doesborgh S, Van de Sandt-Koenderman WME, Dippel D,
Van Harskamp F, Koudstaal P, Visch-Brink EG. Effects of
semantic treatment on verbal communication and linguistic
processing in aphasia after stroke: a randomized controlled
trial. Stroke. 2004;35:141-146.
15. Bowen A, Hesketh A, Patchick E, et al. Effectiveness of
enhanced communication therapy in the first four months
after stroke for aphasia and dysarthria: a randomised con-
trolled trial. BMJ. 2012;345:e4407.
16. Albert ML, Sparks RW, Helm NA. Melodic Intonation
Therapy for aphasia. Arch Neurol. 1973;29:130-131.
17. Berlin CI. On: Melodic Intonation Therapy for aphasia by
R. W. Sparks and A. L. Holland. J Speech Hear Disord.
1976;41:298-300.
18. Helm-Estabrooks N. Exploiting the right hemisphere for
language rehabilitation: Melodic Intonation Therapy.
In: Perecman E, ed. Cognitive Processing in the Right
Hemisphere. New York, NY: Academic Press; 1983:229-240.
19. Sparks R, Helm N, Albert M. Aphasia rehabilitation resulting
from Melodic Intonation Therapy. Cortex. 1974;10:303-316.
20. Stahl B, Henseler I, Turner R, Geyer S, Kotz S. How to
engage the right hemisphere in aphasic without even sing-
ing: Evidence for two paths of speech recovery. Front Hum
Neurosci. 2013;7:35.
21. Schlaug G, Marchina S, Norton A. From singing to speak-
ing: why singing may lead to recovery of expressive language
function in patients with Broca’s aphasia. Music Percept.
2008;25:315-323.
22. Schlaug G, Marchina S, Norton A. Evidence for plasticity in
white-matter tracts of patients with chronic Broca’s aphasia
undergoing intense intonation-based speech therapy. Ann N Y
Acad Sci. 2009;1169:385-394.
23. Belin P, Van Eeckhout P, Zilbovicius M, et al. Recovery from
nonfluent aphasia after Melodic Intonation Therapy: a PET
study. Neurology. 1996;47:1504-1511.
24. Breier J, Randle S, Maher L, Papanicolaou A. Changes in
maps of language activity activation following Melodic
Intonation Therapy using magnetoencephalography: two case
studies. J Clin Exp Neuropsychol. 2010;32:309-314.
25. Bonakdarpour B, Eftekharzadeh A, Ashayeri H. Melodic
Intonation Therapy in Persian aphasic patients. Aphasiology.
2003;17:75-95.
544
26. Goldfarb R, Bader E. Espousing Melodic Intonation Therapy
in aphasia rehabilitation: a case study. Int J Rehabil Res.
1979;2:333-342.
27. Van der Meulen I, Van de Sandt-Koenderman WME, Ribbers
GM. Melodic Intonation Therapy: present controversies and
future opportunities. Arch Phys Med Rehabil. 2012;93:46-52.
28. Benson DF, Dobkin BH, Gonzalez Rothi LJ, Helm-
Estabrooks N, Kertesz A. Assessment: Melodic Intonation
Therapy. Neurology. 1994;44:566-568.
29. Conklyn D, Novak E, Boissy A, Bethoux F, Chemali K. The
effects of modified Melodic Intonation Therapy on nonflu-
ent aphasia: a pilot study. J Speech Lang Hear Res. 2012;55:
1463-1471.
30. Sparks R. Melodic Intonation Therapy. In: Chapey R, ed.
Language Intervention Strategies in Aphasia and Related
Neurogenic Communication Disorders. Baltimore, MD:
Lippincott Williams & Wilkins; 2008:837-851.
31. Graetz P, de Bleser R, Willmes K. Akense Afasie Test [Aachen
Aphasia Test] (Dutch version). Lisse, the Netherlands Swets
& Zeitlinger; 1991.
32. Helm-Estabrooks N, Nicholas M, Morgan A. Melodic
Intonation Therapy. Austin, TX: Pro-Ed; 1989.
33. Van Eeckhout P. Sabadel: Histoires insolites pour faire par-
ler. Paris, France: Medsi; 1982.
34. Blomert L, Koster C, Kean ML. Amsterdam-Nijmegen Test
voor Alledaagse Taalvaardigheid [Amsterdam-Nijmegen
Everyday Language Test]. Lisse, the Netherlands Swets &
Zeitlinger; 1995.
Neurorehabilitation and Neural Repair 28(6)
35. Visch-Brink EG, Stronks DL, Denes G. De semantische
associatie test [Semantic Association Test]. Amsterdam, the
Netherlands Harcourt Assessment; 2005.
36. Paul C, Pijnenburg V. Het sabadelonderzoek: Een semi-spon-
tane taalanalyse in het kader van het rats-onderzoek [The
Sabadel Study: A Semi-Spontaneous Language Analysis
in the Rats Study]. Unpublished Master Thesis, Radboud
University. 2002.
37. Nicholas LE, Brookshire RH. A system for quantifying the
informativeness and efficiency of the connected speech of
adults with aphasia. J Speech Hear Res. 1993;36:338-350.
38. Boucher V, Garcia LJ, Fleurant J, Paradis J. Variable efficacy
of rhythm and tone in melody-based interventions: implica-
tions for the assumption of a right-hemispheric facilitation in
non-fluent aphasia. Aphasiology. 2001;15:131-149.
39. Norton A, Zipse L, Marchina S, Schlaug G. Melodic
Intonation Therapy: shared insights on how it is done and why
it might help. Ann N Y Acad Sci. 2009;1169:431-436.
40. Weinrich M, McCall D, Boser KI, Virata T. Narrative and
procedural discourse production by severely aphasic patients.
Neurorehabil Neural Repair. 2002;16:249-274.
41. Laska A, Hellblom A, Murray V, Kahan T, von Arbin M.
Aphasia in acute stroke and relation to outcome. J Intern Med.
2001;249:413-422.
42. Marchina S, Zhu L, Norton A, Zipse L, Wan C, Schlaug G.
Impairment of speech production predicted by lesion load of
the left arcuate fasciculus. Stroke. 2011;42:2251-2256.
Copyright of Neurorehabilitation & Neural Repair is the property of Sage Publications Inc.
and its content may not be copied or emailed to multiple sites or posted to a listserv without
the copyright holder's express written permission. However, users may print, download, or
email articles for individual use.