The present cross-sectional study was conducted from March 2023 to March 2024 in Justice K.S Hegde Charitable Hospital, Mangaluru, Karnataka, India. The study proposal was ethically approved by Nitte Institute of Physiotherapy’s Institutional Ethical Committee, Mangaluru, Karnataka, India on February 9^th 2023, with reference number: Ref: NIPT/IEC/Min//28/2022-2023. It was registered in the Clinical Trial Registry: India (CTRI) with the registration number: CTRI/2023/04/051545.

The target population for the study was poststroke patients and sample size being Time bound sampling. Informed consent was taken from the participants who agreed to participate in the present study.

Inclusion and Exclusion criteria: The study criteria included male and female stroke patients between the age of 18 to 80 years and above who are diagnosed with MCA stroke and have a Standarised Mini-mental State Examination (SMMSE) score greater than or equal to 24 [10,11]. They should be able to walk 10 feet without using any assistive device and be able to follow and respond to verbal instruction. In the exclusion criteria, patients with significant aphasia, recurrent stroke taking any psychiatric drug and who have any other neurological conditions or orthopaedic complications like amputation, fractures or orthopedic deformities.

After obtaining informed consent from the participants, the study measured patient’s walking speed while engaging in a CMDT [12,13]. Initially, participants were instructed to walk a distance of 10 feet at their regular pace, starting from a stationary position. Those who successfully completed this task were then asked to perform the CMDT.

The risk of falls among the patients involved in the study was evaluated using the Morse Fall Scale. This standardised scale assigns scores ranging from 0 to 24 for low fall risk, 25 to 44 for moderate fall risk, and 45 or higher for high fall risk, thereby categorising the participants based on their risk level [14]. In CMDT, the motor task is to perform Timed Up and Go (TUG) test [15]. To initiate the activity, the patient should be seated comfortably in a chair, and a line positioned 3 m away on the floor should be identified. Upon the therapist’s instruction of “Go,” the patient will rise from the chair, walk to the designated line at their usual pace, turn, walk back to the chair at their normal pace, and then resume a seated position. Recording of the time commences at the verbal cue “Go” and cease once the patient is seated back in the chair. Pedometer is attached to the patient during the task to measure the step count and time taken to complete the test is measured using stop watch. TUG test is divided into subtasks and the time taken to walk 3 m forward and retuning back is recorded excluding the time taken for getting up and sitting down in the chair [16]. Walking speed was calculated as distance by time (i.e., the time taken to walk the total distance of 6 m) [17].

In CMDT, the cognitive task was to perform the following task. Patients were asked to do serial three subtractions: Repeatedly subtracted three from a random number (e.g., 399, 396, 393, 390). Following the completion of the CMDT, the total number of cognitive responses and the correct number of cognitive responses was manually recorded [18].

Data collected was analysed by using the SPSS software version 29.0. The data was summarised using descriptive statistics: frequency, percentage; mean and standard deviation. To compare the data Independent sample t-test, One-way ANOVA was used. The multiple comparisons of data were analysed by using the Post-hoc analysis, Tukey test and Bonferroni was used.

The study was conducted among 23 post stroke patients with MCA stroke. The study participants consisted of an average age of 54.6±14.8 years. Among them, the mean duration of stroke was 1.3±1.6 years [Table/Fig-1]. In the 23 stroke patients, 7 (30.4%) had low fall risk, 9 (39.1%) had moderate fall risk and 7 (30.4%) had high fall risk. The majority were male 17 (73.9%) and female 6 (26.1%) were in minority [Table/Fig-2].

The statistical analysis revealed a significant difference in time taken, step count, total and correct number of cognitive responses among with risk groups [Table/Fig-3,4].

The statistical analysis revealed a significant difference in walking speeds among the risk groups, with an F value of 15.486 and a p-value less than 0.001 [Table/Fig-5]. Significant differences are noted in walking speed between the low and high-risk groups (p<0.001) and between the moderate and high-risk groups (p=0.010) after employing the Bonferroni correction for multiple comparisons [Table/Fig-6].

In the present study, significantly reduced walking speed was noted during dual-task walking. This finding aligns with previous research indicating a comparable decline in gait speed under dual-task conditions. Typically, this modulation in gait speed involves a decrease in step length and cadence, coupled with an increase in double support time [19]. Executing two tasks that require comparable cognitive or motor skills can result in slower performance in both tasks or delays specifically in the secondary task. This phenomenon is explained by theories such as capacity-sharing and bottleneck theory [20,21].

Lee KB et al., the result of the study demonstrates that age, Morse fall scale and TUG could identify fallers and non fallers [22]. The MFS has the ability to predict a patient’s risk of falling using a cut-off score of 32.5 points which come under moderate fall risk in the current study and the TUG test proved to be the most reliable assessment of walking ability for predicting falls in patients with neurological conditions [22]. Turning during walking, such as in the TUG test, is a task that requires considerable attention from healthy individuals. It includes slowing down gait speed, adjusting gait parameters, and reorienting body segments in a sequence [23-25]. A prior study indicated that both anterior-posterior and medial-lateral ground reaction forces undergo changes before individuals decelerate their walking speed to initiate a turn [26]. Step lengths and step widths become uneven in order to shift the Center of Mass (COM) towards the new direction during turning [24]. Therefore, successfully executing a turn requires integrated control of multiple systems, including motor, sensory, and postural control [27].

The current study showed that the time and step count taken to complete CMDT increased with increase in risk of fall and the total and correct number of cognitive responses was reduced with increase in risk of fall. The relationship between working memory and executive functions is well-established. Hence, the greater gait decrement observed during the counting dual task among individuals at a high-risk of falls, compared to those at low-risk, might be attributed to impaired executive functioning [28]. Prior studies have already outlined an independent association between executive functioning and balance as well as mobility in stroke patients [28].

Executive functions refer to higher cognitive processes that occur in both the anterior and posterior regions of the brain. These processes are responsible for modulating and generating effective, goal-directed actions, as well as managing attentional resources [29].

In the study done by Hyndman D et al., showed reduced stride length while walking on flat ground while performing a memory task, which distinguished fallers from non-fallers. The research revealed that patients with strokes had increased walk time and poorer cognitive task performance these findings were consistent with the current study [30].

In a study conducted by Tsang CS and Pang MY, it was discovered that among various single-task and dual-task walking assessments, the degree of motor interference during obstacle crossing coupled with a categorical naming task exhibited greater efficacy in predicting future falls [31]. The study’s findings revealed that individuals who experienced falls demonstrated significantly longer walking times compared to non-fallers, both during single-task and dual-task obstacle crossing, which aligns with the results of the current study. According to the study, in situations where dual task walking presents attention competing contexts, patients will slow down in order to maximise their postural stability [31].

Baetens T et al., study revealed that fallers may be discriminated from non-fallers by a reduction in the stride length and the non paretic step length when walking while counting back. Fall-prone stroke patients can be identified by a decrease in their spatial gait features during performing a working memory task [32]. In contrast to the study gait parameters like cadence, velocity, and paretic step length did not show any significant difference. However, distinct gait decrement percentages, particularly pertaining to stride length and non-paretic step length, serve as distinguishing factors between fallers and non-fallers [32]. Moreover, the decrement percentage associated with paretic step length can differentiate between individuals experiencing multiple falls versus those experiencing a single fall [32].

The study conducted by Plummer-D’Amato P et al., demonstrated a decrease in gait decrements in dual-task scenarios among post-stroke individuals, a finding consistent with the present study [33]. Notably, significant dual-task effects were observed for various gait parameters including gait speed, stride time, average stride length, and cadence. Additionally, the study revealed that speech performance was significantly influenced by concurrent walking tasks [33].

In the study conducted by Al-Yahya E et al., it was found that individuals who had experienced a stroke exhibited an increased requirement for prefrontal cortex activity during walking [34]. This demand intensified when performing a concurrent cognitive task. Under dual-task conditions, there was heightened activation not only in the prefrontal cortex but also in the inferior temporal cortex and basal ganglia [34]. These findings provide evidence regarding the neural mechanisms involved in real-life walking under dual-task conditions, highlighting how elevated cognitive demands, as indicated by increased prefrontal cortex activity, could potentially restrict walking abilities in everyday life after a stroke [34].

Comparatively smaller sample size of the study may affect the generalisability of the result. The study included both acute and chronic stroke and less female patients compared to male with may affect results of study. Considering the present findings, future studies can be conducted focusing on impact of walking speed on risk of fall in stroke patients other than MCA stroke using CMDT. With reference to the current results, the study can be extended to a stroke population belonging to Anterior Cerebral Artery (ACA) and Posterior Cerebral Artery (PCA) type of stroke and other spatial and temporal gait parameters could be assessed using an electronic walkway connected to a computer with specific software.

The study showed that walking speed will impact the risk of fall in stroke patients while performing CMDT. In conclusion, the study demonstrates that walking speed in stroke patients, assessed using CMDT, is significantly associated with the risk of falls. The findings indicate that patients at higher risk of falls exhibit slower walking speeds and poorer performance in CMDT activities. Factors such as age, duration since stroke onset, and gender contribute to fall risk, with older patients and those with shorter durations since stroke showing higher fall risk, particularly among females.