Although there is evidence of mild cognitive impairments for many
individuals with mild traumatic brain injury (mTBI) and posttraumatic
stress disorder (PTSD), little research evaluating the effectiveness of
cognitive training interventions has been conducted. This randomized
controlled trial examined the effectiveness of a 9-h group cognitive
training targeting higher-order functions, Strategic Memory Advanced
Reasoning Training (SMART), compared to a 9-h psychoeducational control
group in improving neurocognitive functioning in adults with mTBI and
PTSD. A sample of 124 adults with histories of mild TBI (n = 117) and/or current diagnoses of PTSD (n = 84) were randomized into SMART (n = 66) or Brain Health Workshop (BHW; n
= 58) and assessed at three time points: baseline, following training,
and 6 months later. Participants completed a battery of neurocognitive
tests, including a test of gist reasoning (a function directly targeted
by SMART) as well as tests of verbal, visual, and working memory and
executive functioning, functions commonly found to be mildly impaired in
mTBI and PTSD. The two groups were compared on trajectories of change
over time using linear mixed-effects models with restricted maximum
likelihood (LMM). Contrary to our hypothesis that SMART would result in
superior improvements compared to BHW, both groups displayed
statistically and clinically significant improvements on measures of
memory, executive functioning, and gist reasoning. Over 60% of the
sample showed clinically significant improvements, indicating that gains
can be found through psychoeducation alone. A longer SMART protocol may
be warranted for clinical samples in order to observe gains over the
comparison group.
Introduction
Approximately 1.7 million traumatic brain injuries (TBI) occur in the United States each year (1, 2).
The majority of those (75%) are mild traumatic brain injuries (mTBI),
which often involve physical, cognitive, and affective symptoms in the
acute phase followed by resolution of symptoms after ~1 month (3). However, an estimated 10–20% of patients continue to report symptoms that persist months to years after the injury (4, 5)
that have been associated with social and occupational dysfunction,
including under-employment, low income, and marital problems (6–9). As such, identifying efficacious interventions for cognitive deficits related to mTBI is a priority.
In addition, mTBI is highly comorbid with posttraumatic
stress disorder (PTSD), which represents a potential complicating factor
in recovery. Among veterans with histories of TBI, rates of PTSD range
from 33 to 65% (10, 11).
PTSD has been associated consistently with mild neurocognitive deficits
in a number of domains. Meta-analyses reveal significant differences
between individuals with PTSD compared to healthy and trauma-exposed
controls, representing medium to large effect sizes, in the domains of
verbal learning and memory, processing speed, attention/working memory,
and executive functions (12, 13). Moreover, patients with PTSD self-report cognitive problems with detrimental impacts on social and occupational functioning (14–16).
Research on neuropsychological functioning in mTBI is
less consistent, in part due to the heterogeneity in the criteria used
to define mTBI, populations sampled, time since injury, and mechanisms
of injury. Individuals with persisting post-concussive cognitive
complaints have shown impairments in sustained attention (17–19), divided attention (20), selective attention and inhibitory control (17, 21), cognitive flexibility and planning (8, 22, 23), processing speed (24), verbal memory (25–28), and visual memory (18).
In addition, even patients who report full recovery may continue to
experience cognitive problems under conditions of physical or
psychological stress (29).
The high comorbidity of mTBI and PTSD presents the potential for
greater impaired functioning. In studies examining mTBI and PTSD
concurrently, the majority found that while PTSD was related to
neuropsychological impairments, mTBI was not (30–32).
However, some studies have found a poorer performance profile in
individuals with both mTBI and PTSD, as compared to those with mTBI or
PTSD alone (21, 31). Given the overlap of structural and functional changes and neurocognitive deficits seen in both PTSD and mTBI [e.g., (33, 34)]
there is a critical lack of investigations that evaluate cognitive
rehabilitation approaches for these individuals. This paper attempts to
fill this void.
Brain regions particularly vulnerable to both mTBI and
PTSD are the frontal lobes, which are involved in learning and memory
operations, executive functioning, attention and working memory, and
reasoning abilities. The importance of frontal lobe function in
neurological recovery after TBI is reflected in functions such as
motivation, attention, and working memory that are prerequisites for
optimal rehabilitation. Difficulties in these areas are considered poor
prognostic indicators for TBI rehabilitation (35). Rehabilitation of frontal lobe functions is thus a crucial goal for enhancing recovery from brain injuries.
Prior studies have demonstrated that training-based rehabilitation therapy helps patients with neurological damage (36–39).
However, a major limitation of many rehabilitation studies is the lack
of a theoretical foundation based on known mechanisms of brain function,
which can serve to guide treatment development. The proliferation of
computer-based technology over the past decade has led to the rise of
the rehabilitative models that employ repeatable tasks and mass
training. Despite their popularity, results on the efficacy of these
restorative training programs have been mixed, and considerable debate
remains regarding how to effectively restore cognitive capacities
following TBI.
To date, randomized controlled trials (RCTs) aimed at
improving cognitive functioning in patients with mTBI have shown limited
effectiveness (40–42).
The research literature examining cognitive rehabilitation (CR) for
mTBI has been limited by a lack of well-designed and sufficiently
powered studies that fail to include control groups and functional
outcomes (41, 43).
RCTs aimed at treating cognitive symptoms in the post-acute or chronic
stage are particularly lacking. A recent exception is a study that
compared psychoeducation, computerized brain training, therapist-led CR,
and a therapist-led CR/psychotherapy hybrid (40).
The four interventions were equivalent in improving cognitive
functioning, with between 23 and 33% of participants showing reliable
change on the primary working memory outcome. The therapist-led CR and
the integrated groups showed significantly greater improvements on a
self-report of functional cognitive and behavioral difficulties (23 and
19%, respectively, in the two groups, showed reliable change) compared
to psychoeducation and computerized brain training. However, these
interventions were resource-intensive, with treatment consisting of
daily therapy for 6 weeks.
Research examining CR for PTSD-related cognitive
impairments is lacking. Recently, researchers tested the effectiveness
of a computerized cognitive training program, a hybrid of Lumosity and
MyBrainSolutions, in improving neurocognitive functioning in a sample of
primarily motor vehicle accident survivors recruited from emergency
rooms (44).
Compared to the control group that engaged in computer games, card
games, and matching tasks, the CR group showed significant improvements
(Cohen's d = 0.58) in cognitive flexibility after 1 month of CR,
assessed 3 months following the trauma. This study lends preliminary
support for the use of cognitive training for PTSD, particularly in the
acute phase, although less is known about the treatment of long-term
cognitive impairments related to PTSD.
Researchers have argued that for rehabilitative
interventions to be successful, they must target skills that are
directly applicable to daily functioning, particularly for patients with
more mild impairment levels, as is the case with mTBI and PTSD (45, 46).
In addition, given the importance of frontal lobe functioning in both
mTBI and PTSD, cognitive training must address higher-order, frontal
lobe-mediated cognitive skills.
The development of Strategic Memory and Reasoning Training [SMART; (47, 48)]
addressed this need, with the goal of targeting higher-order functions
found to be crucial for the recovery following brain injury (49).
Prior research has shown that when these specific brain functions are
targeted, such as the ability to focus on a task while ignoring
irrelevant information, brain changes are more significant (49–51).
SMART emphasizes top-down processing by targeting focused attention,
assimilation of information, mental flexibility, and innovation, all
higher-order cognitive functions driven by the frontal lobes. Other
top-down cognitive training programs have demonstrated effectiveness in
improving cognitive and daily functioning in individuals reporting more
severe brain injuries (50–53); however, limited research has been devoted to milder brain injuries.
The goal of SMART is to teach metacognitive strategies
to enhance time and cognitive resource management through goal setting
and the inhibition of distracting or irrelevant stimuli. In addition, it
prioritizes deeper level synthesis of information to obtain the “gist”
while encouraging fluid and flexible thinking (54, 55).
Training in gist reasoning, or “the ability to strategically comprehend
and convey generalized, core meaning(s) from complex information,” is a
primary component of the SMART protocol [54, p. 2]. Strong gist
reasoning minimizes the cognitive overload of competing stimuli in the
environment and focuses on constructing meaning rather than remembering
details. Gist reasoning impairments have been found in adults and
adolescents with mild and moderate TBI (56, 57).
In addition, gist reasoning is associated with frontal lobe activation
and draws upon functions of inhibitory control, working memory,
cognitive flexibility, abstract reasoning, and fluency (56, 58), domains often impaired in both TBI and PTSD.
The effectiveness of SMART has been tested in a number
of studies of adults and adolescents with TBI. The typical SMART
training consists of 15 h of training conducted over 10 group sessions
in the first 5 weeks and a final 3 h of training at spaced intervals
over the next 3 weeks. Vas et al. (59)
conducted an RCT comparing SMART to a psychoeducational control (Brain
Health Workshop; BHW) in adults with TBI histories of >2 years and
moderate functional impairment. The majority of participants' brain
injuries were not specified as mild, moderate, or severe. SMART was
associated with significantly greater improvements in gist reasoning
compared to psychoeducation controls. Generalized improvements were also
seen in working memory and participation in functional activities,
domains that were not directly targeted by the SMART training. These
gains were maintained 6 months post-training.
A subsequent study with children and adolescents who had
received a mild, moderate, or severe closed-head TBI at least 6 months
prior to study participation also demonstrated positive findings. These
participants, who demonstrated below average gist reasoning skills at
baseline, completed either a shorter SMART training protocol of eight
45-min sessions or a memory training (60). The SMART participants displayed significant improvements in their ability to abstract meanings (d = 1.41) and recall facts (d
= 0.77) compared to the control group. The SMART participants also
demonstrated significant improvements in the untrained executive
functions of working memory (d = 0.94) and inhibition (d =
0.73), whereas the control group participants did not. In a larger RCT
of adults with a history of unclassified TBI who were experiencing mild
cognitive impairments at the time of the training, Vas et al. (57)
compared receiving at least 18 h of SMART to BHW over 8 weeks. They
found greater improvements for SMART participants on measures of gist
reasoning, set shifting, and self-reported psychological health and
daily function. These studies demonstrate the effectiveness of SMART in
samples of individuals with a range of brain injury severity. One of the
purposes of the present study was to assess its effectiveness in a
sample of adults with milder brain injuries.
Notably, SMART is also effective in improving cognitive functioning in cognitively healthy individuals (54, 58, 61–63),
which suggests that SMART may show benefits for individuals with mTBI
and PTSD who have less impaired, or even average, functioning. Lack of
impairment is not uncommon for many individuals with mTBI or PTSD [e.g.,
(25, 64–70)], yet appraisals of cognitive functioning are often negative and not aligned with objective performance (16, 71–74).
As a result, targeting cognitive functions via an approach that
emphasizes neuroplasticity and psychoeducation may additionally improve
expectancies and appraisals.
The developers of SMART recently introduced a shortened
SMART training of three, 3-h sessions that has not yet been tested with
mTBI. Similarly shortened protocols have shown gains in higher-order
reasoning, working memory, and immediate and delayed memory in
adolescents and adults with chronic mTBI (75).
To our knowledge, SMART has never been tested with patients with PTSD, a
population that struggles with cognitive problems with limited existing
cognitive rehabilitation research. The overlap of both structural and
functional changes and neurocognitive deficits seen in both PTSD and
mTBI [e.g. (33, 34)]
and the high rates of comorbidity associated with poorer functional
outcomes, highlights the need for cognitive rehabilitation research that
addresses both conditions alone and together. The purpose of the
current study was to investigate the effectiveness of a shortened SMART
training program, compared to a psychoeducation control, in improving
neurocognitive functioning in patients with mTBI and/or PTSD. We
hypothesized that participation in SMART, compared to the control group,
would result in improved gist reasoning as well as improved performance
on tests of generalized cognitive functions (working memory, verbal
memory, visual memory, and executive functioning).
Beatrice Ottiger
Dirk Lehnick
Tobias Pflugshaupt
Tim Vanbellingen
Thomas Nyffeler
Kristin W. Samuelson
Krista Engle1, 
Joshua Jordan
Charles Benight
Xiang Li
YuKai Liu
FuSang Wang
XiaoHan Zheng
Chao Sun
JunShan Zhou
JianJun Zou