Arabic version of the Hayling sentence completion test: Validation and factors associated with executive functions in a sample of the Lebanese adults

To validate the Arabic version of the Hayling Sentence Completion Test (HSCT) in healthy community-dwelling Lebanese adults, and identify factors that might affect executive functions (EF), mainly inhibitory control, in these participants.


Abstract Background
To validate the Arabic version of the Hayling Sentence Completion Test (HSCT) in healthy community-dwelling Lebanese adults, and identify factors that might affect executive functions (EF), mainly inhibitory control, in these participants.

Methods
Between August-December 2019, 350 participants were randomly selected. The Arabic version of the HSCT, divided into automatic and inhibition conditions, was used; in each condition, participants' response-time and number of errors commited were recorded.

Results
None of the scale items was removed. For the automatic condition, response-time items converged over one factor (αCronbach=0.905) and number of errors converged over seven factors (αCronbach=0.334). For the inhibition condition, response-time converged over one factor (αCronbach=0.943) and number of errors converged over four factors (αCronbach=0.728). Using electricity as a heating method inside the house was significantly associated with a lower response-time, whereas higher problematic mobile phone use was associated with higher response-time. Using wood as a heating system inside the house and higher problematic mobile phone use were associated with higher number of errors, while using Arabian incense (bakhour) inside the house was associated with lower number of errors.

Conclusion
We were able to validate the HSCT Arabic version for use in the Lebanese population.
Problematic mobile phone use was associated with lower executive functions in terms of reponse-time and number of errors. 3 Background Executive functions (EFs) include the production, preservation, and customizаtion of plаns aimed to accomplish the objectives of a given assignement 1 . It includes three main features: (1) inhibitory control, including self-control and interference control, (2) cognitive flexibility, also called mental flexibility, linked to creativity, and (3) working memory 1,2 . EFs are crucial skills for mental and physical well-being, school and life success, and for cognitive, social, and psychological growth and evolution 2 .
Inhibition represents one of the key skills of EFs. It involves the capacity to control one's attention, behavior, thoughts, and emotions, to suppress a strong internal tendency or external temptation, in order to do what's more relevant or essential 2 . In the absence of inhibition, we would be controlled by our impulses, settled tendencies of thought or behavior, by the environment that pulls us this way or the other 2 . Consequently, inhibition makes each one of us somehow unique, and allows us to proceed and behave deliberately, instead of being unintentional and impulsive creatures of tradition and routine 2 .
Inhibition has become crucial in several research domains within psychology and neurology. It is broadly recognized that inhibitory control is a multidimentional procedure encompassing at least three features, as defined by Bayard S et al.: « (1) Prepotent response inhibition (the ability to deliberatly suppress dominant, automatic, or prepotent responses), (2) Resistance to distractor interference (the ability to resist interference from distracting information in the external enviroment), (3) Resistance to proactive interference (the ability to resist the intrusions of irrelevant information, that were once considered relevant, into our memory). » 3 Taking into account inhibition as a multidetermined construct emphasizes the exigency to invent validated clinical tools for the evaluation of its features 3 . The Hayling Sentence Completion Test (HSCT), originally developed by Burgess & Shallice in 1996, is one of the available tools that evaluate inhibition. It is a measure of prepotent response inhibition that detects frontal lobe dysfunction. In this test, participants hear sentences in which the last word is absent. In the automatic condition, subjects are asked to complete sentences correctly (i.e., by a word solidly related to the sentence), reflecting the initiation of a semantically supported automatic response. In the inhibition condition, participants are asked to restrain themselves from saying the cueing word, and to complete the sentence with a totally unrelated word. To perform this task appropriately, subjects have to inhibit the related word and its semantic associates 4 .
As far as we know, only five validated languages have been published for the HSCT in adults and elderly people: (1) English 5 , (2) Spanish 6 , (3) French 3 , (4) Italian 7 , and (5) Swedish 8 . There are no HSCT normative data for the Arabic speakers to date, so it can not be used in Arabic-speaking clinical practice and research groups. Therefore, this study intends to create an Arabic version of the HSCT that is standardized to the linguistic and cultural realities of the Lebanese population.
EFs are negatively influenced by many factors, among which cognitive aging, which usually weakens executive functioning 9 , and low family socioeconomic status (SES), which predicts a worse performance on tasks that evaluate them 10 . In contrast, high educational levels improve scores in the HSCT 3 , and multilingualism is associated with greater amount of controlled attention and inhibitory control, and can play a crucial role in protecting against the deterioration of executive functions with aging by helping to achieve cognitive reserve 11 . Furthermore, there is a positive correlation between healthy diet (e.g. whole grains, fish, fruits and/or vegetables) and executive functions, unlike unhealthy diet (e.g. fast food, sweetened beverages and red meat) that is associated with a decrease in executive functioning 12 . Moreover, a meta-analysis revealed a wide EFs deficit in overweight participants compared to healthy weight controls, specifically a deficit in inhibition and working memory in overweight participants 13 , while better physical fittness is shown to improve cognitive flexibilty 14 . On another note, a negative correlation was exhibited between EFs and constant exposure to polluted air and traffic noise 15 .
Studies concerning the relationship between long term mobile phone use and EFs are equivocal. Previous findings showed no evidence of a harmful effect of smartphone use on cognitive functioning. Instead, it suggests there could possibly be a favorable effect of phone use on cognitive performance and executive functioning, but the results of longitudinal analyses were confusing 16 . Another study indicated that a daily exposure (2 h/day for 4 weeks) to electromagnetic fields (EMF) emitted by mobile phones (MP) has no effect on executive function 17 19 . On another note, Angiotensin I receptor blockers (ARBs) improve memory and executive function in comparison to other antihypertensive drugs, and it attenuates the decline of cognition over time 20 . Moreover, aspirin use is also associated with a preventive effect against cognitive decline, particularly in people at risk for developing dementia 21 .
This study was conducted for several reasons, primarily because of the lack of an Arabic version of the HSCT that could be beneficial in our clinical practice. Another reason is the need to find potential relationships between various factors with executive functions, especially problematic mobile phone use as suggested by many previous studies, since smartphones have become, nowadays, a daily companion of people from all generations.
Thus, it is important to conduct this study in Lebanon to be adapted to the ideas, customs and social behaviour of the Lebanese citizens. Hence, the objectives of the current study are to use the Arabic version of the HSCT in healthy community-dwelling Arabic-speaking adults in Lebanon, to check its validity compared to other versions of the test, as well as to identify risk factors that might affect the executive funcions in these adults.

Study design
This study was conducted between August and December 2019. Participants were randomly chosen from the general population across Lebanon. Individuals included in this study were aged above 18 years old, having no known history of neurological or psychological impairement. Excluded patients were those with abnormal performances on the HSCT, which have been described in a wide variety of neurological, psychiatric, and neurodevelopmental conditions, such as Alzheimer's disease and mild cognitive impairment 22-24 , brain traumatic injury 25 , cerebrovascular accidents 26 , Parkinson's disease 27 , amyotrophic later sclerosis 28 , frontotemporal dementia 28 , schizophrenia 29 , bipolar disorder 30 , and autism spectrum disorder 31 . Patients aged above 55 years old 32 underwent a mini mental state examination (MMSE); those obtaining a score of 24 or more were included in the study 33 . Minimal sample size calculation According to Comrey and Lee 34 , a minimal sample of 5-10 observations is needed per item of the scale in order to validate a scale. Therefore, a minimum of 300 participants was needed for adequate statistical power, since the assessment of the automatic and inhibition conditions included 15 items each.

Questionnaires and variables
Data was collected via a personal interview, and subjects were tested with only one examiner in a quiet office. The first section assesses the socio-demographic characteristics, including age, gender, region, the number of rooms in the household and the number of persons living in it, the level of education, the number of mastered languages, the monthly income, and the medical history and chronic treatments.
The second section evaluates the diet of each participant, taking into consideration vegetables, fruits, dairy products, breakfast, eating 5 meals per day, sweets consumption, red meat, beverages, and fast food. Each variable is scored between 1 and 5 to get a minimum score of 9 and a maximum score of 45; higher scores indicate a healthier diet. This scale was inspired by the constituants of the DASH diet (Dietry Approaches to Stop Hypertension) 35 .
The third section is the Arabic translation of the short form of International Physical Activity Questionnaire (IPAQ) 36 that asks about three specific types of activities: walking, moderate-intensity activities, and vigorous-intensity activities. Then participants are classsified to have low, moderate, or high physical acitivity.
To be classified as 'moderate', participants should match one of the following: a) 3 days of vigorous-intensity activity for a minimum of 20 minutes per day. OR b) 5 days of moderate-intensity activity and/or walking for a minimum of 30 minutes per day. OR c) 5 days of any combination of walking, moderate-intensity or vigorous intensity activities, achieving at least a total physical activity of 600 minutes/week.
To be classified as 'high', participants should match one of the following: a) vigorous-intensity activity 3 days, achieving at least 1500 minutes/week. OR b)7 days of any combination of walking, moderate-intensity or vigorous-intensity activities of at least 3000 minutes/week.
Participants who did not meet the above criteria were considered to have a 'low' physical activity level 36 .
The fourth section assesses the exposure to polluted air; participants are asked about the environment where they live, the heating system used (wood, gas, electricity), if they are living or working next to factories or power stations, and their exposure to cigarette smoking.
The fifth section represents the Arabic translation of the Short Version of the Problematic Mobile Phone Use Questionnaire (PMPUQ-SV), containing 15 items that are scored from 1 ('I strongly agree') to 4 ('I strongly disagree'), except for 8 items that are reversly scored.
Overall scores range from 15 to 60, with higher scores indicating a higher risk for problematic mobile phone use 37 .

The Hayling Sentence Completion Test
The HSCT was translated to Arabic and adapted to the linguistic and cultural realities of the Lebanese population. The A-HSCT consists of two conditions (аutomatic and inhibition), for which two different groups of 15 sentences are given. In both conditions, the interviewer reads out loud the unfinished sentences, and the pаrticipant has to complete each sentence with one word 4 .

Automatic condition
The participants are requested to give a word that is related to the beginning of the sentence, and should do so as quickly аs possible. For example, "He mailed the letter without а . . . (participant says) stamp." Time latency in automatic condition measures the participant's rapidity in initiating an automatic response. According to the scoring system, three error points are scored when participants provide an incorrect word, one point when the answer is semantically related to the sentence, and no error point if the correct word is given. Higher error score corresponds to a lower performance 4 .

Inhibition condition
The participants are requested to complete the sentence аs fast as they can with a word that is completely unlinked to it, which makes no sense at all in the context of the sentence. For example, "The captain wanted to stay with the sinking . . . (pаrticipant says) apple." If at any time during this condition the subject completes the sentence correctly instead of using an unrelated word, s/he is told thаt the word is too linked to the sentence and is retold the task instructions. Time latency in inhibition condition gives information about the time needed to inhibit the correct response and find an incorrect one 4 .
According to the scoring system, three points are given when the sentence is completed with the answer that fits with it. One point is given when a subject gives an antonym, a semantically related word, or a word that makes a vague reference to the sentence.
Participants receive zero points when a totally unrelated word is provided. A higher error score indicates a lower performance 4 .
The total time to complete both conditions of the A-HSCT was approximately five minutes.
The automatic condition was tested prior to the inhibition section. Two practice sentences were initially presented prior to each condition. Time of response latency was measured and collected using a stopwatch in both conditions; the timing began soon after the tester finished the sentence, and was stopped soon after the participant began their answer. with an Eigenvalue higher than one were retained. The Student t-test was used to compare two means, whereas the Pearson correlation was used to study the association between two continuous variables. Multivariable linear regression models were done to explore factors associated with the response-time and the number of errors taken as dependent variables and taking all variables that showed a p<0.05 in the bivariate analysis as independent variables. A p<0.05 was considered significant. Reliability was assessed using Cronbach's alpha.

Results
Out of 364 participants approached, 350 (96.15%) accepted to enroll in this study. The mean age of the participants was 45.72 years, with 177 (50.6%) of them being females.
Other sociodemographic and characteristics of the participants are summarized in Table 1.

Bivariate analysis
The  (Table 3). Higher age was significantly associated with lower response-time (better performance), whereas higher problematic mobile phone use was significantly associated with higher response-time and higher number of errors (worse performance) ( Table 4).

Multivariable analysis
The results of a first linear regression, taking the response-time as the dependent variable, showed that using electricity as a heating method inside the house (B=-0.3) was significantly associated with lower response-time (better performance), whereas a higher problematic mobile phone use (B = 0.03) was significantly associated with higher response-time (worse performance) ( Table 5, Model 1).
The results of a second linear regression, taking the number of errors as the dependent variable, showed that using wood as a heating system inside the house (B = 0.41) and higher problematic mobile phone use (B = 0.02) were significantly associated with higher number of errors (worse performance), whereas using Arabian incense (bakhour) inside the house (B=-0.29) was significantly associated with a lower number of errors (better performance) ( Table 5, Model 2).  On another note, Bakhour is a widespread spiritual practice in the Middle East region, composed of a vast range of chemical compounds and metals. Opposed to the upstated results concerning indoor air pollution, the use of Arabian incense was associated to a decreased number of errors commited during the A-HSCT. We hypothesize that this could be attirubted to the spirituality inferred by the use of Bakhour among the participants, which is the subject of many studies that elaborate the benefits of spirituality on executive functions, and its capacity in attenuating the severity of depression 43 .

Other factors
In the bivariate analysis, participants who are older, use aspirin, have cardiac diseases, diabetes, or hypertension exhibited better performance in terms of response-time in the inhibition condition of the A-HSCT. These results contradict the literature that underlines the negative correlation between executive functions and aging. 9 However, they could be explained by the fact that many of the elderly participants reported taking metformin, aspirin, atorvastatin, and angiotensin converting enzyme inhibitors (ACEi) or ARBs, which, when taken as a combination, could potentiate the anti-oxidant effect of each drug, enhancing their neuroprotective effects on the hippocampus 44 .

Clinical implications
The A-HSCT is a clinical tool that measures inhibitory control as part of executive functions, reflecting mainly frontal cortex function. The obtained score could be used in patients with many neurological or psychological disorders to follow them up throughout the years, and to detect amelioration or deterioration of their scores. Moreover, by comparing the patients' scores, clinicians can detect the efficacy of their treatment and medications.

Limitations
Our study is based on the questionnaire that was applied on the participants during an interview. Thus, some of these participants might have felt pressured or shy while being interviewed, leading them to give wrong answers or to perform poorly in the given tasks.

Conclusion
The Arabic version of the Hayling Sentence Completion Test is now a valid tool that can be used by clinicians to measure response inhibition in adults and elderly patients in the Lebanese population. We were also able to suggest an association between many factors and executive functions. The most significant was problematic mobile phone use and the deterioration of executive functions, mainly inhibitory control, in terms of both responsetime and number of errors in the A-HSCT. As far as we know, the present study is the first to create normative data, let alone validate the A-HSCT in the Lebanese population.
Furthermore, it should be noted that additional studies are needed to prove a possible harmful effect of mobile phone use on frontal cortex, since it became a primarily used feature in our daily life. The study protocol was approved by the Holy Spirit University of Kaslik (USEK) ethics committee. A written informed consent was obtained from each participant.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analyzed during this study are not publicly available to maintain the privacy of the individuals' identities. The dataset supporting the conclusions is available upon request to the corresponding author.

Competing interests
The authors have nothing to disclose.

Funding
None.

Author contributions
SB was responsible for the data collection and designed the study; SB drafted the manuscript; SH carried out the analysis and interpreted the results, assisted in drafting and reviewing the manuscript; EAH and JM helped in the drafting the paper. SH and HM were the project supervisors. All authors reviewed the final manuscript and gave their consent.