The Arabic Translation and Cross-Cultural Adaptation of the Bristol Activity of Daily Living Scale

INTRODUCTION

The World Health Organization (WHO) considers dementia as one of the major causes of disability affecting elderly [1]. Twelve percent of the years lived with disability in adults are thought to be caused by dementia [1]. The prevalence of dementia increases with age, such that dementia affects 4.7% of people older than 65 years living in high-income countries. Albeit there is limited information on the burden and prevalence of dementia in the Arab world, with estimates of 2.6–6% of older adults living in the Arab region having dementia [2–6], a recent report suggested prevalence could reach as high as 18.5% in the oldest age groups [2]. The cost of caring for dementia globally is estimated to surpass one trillion US dollars with the greatest impact on high-income countries [7, 8]. Early diagnosis of dementia is cost-effective in enabling earlier provision of treatment of behavioral symptoms, education of caregivers, and, thus, prevention of institutionalization [9].

Accurate diagnosis of dementia requires a detailed assessment of the patient’s current and previous functional performance [10, 11]. Functional measures of daily living are essential components of the comprehensive assessment of and management of patients with cognitive decline, and it is crucial that the measures used are relevant to the patient being assessed [12]. For example, the Barthel Activity of Daily Living Index was originally validated in patients with stroke [13], thereby limiting its utility in patients with dementia given the different patterns of disability between the two groups [14]. On the other hand, the Bristol Activity of Daily Living Scale (BADLS) was designed and validated for use in patients with dementia [15]. The scale developers sought input from caregivers of patients diagnosed with dementia when undergoing initial scale development [15]. The scale consists of 20 items and responses are graded to describe the person’s increasing dependency in performing basic and instrumental activities of daily living. The items assess the following daily activities: drink preparation, use of the telephone, food preparation, housework, communication, shopping, eating, orientation to space, games and hobbies, dental care, hygiene, bathing, dressing, using the toilet, drinking, mobility, transferring, orientation to time, driving, using public, transport, and managing finances. Each item is scored 0 (independent) to 3 (fully dependent), with the total score ranging from 0 to a maximum score of 60. The Cronbach’s alpha reliability coefficient of the original total scale was 0.82 indicating good reliability. The BADLS scale has a moderate convergent validity (r = 0.65) against the participant’s actual function at home as rated by a nurse [12]. Additionally, the BADLS scale is strongly associated (r = –0.67) with the commonly used screening cognitive test, the Mini-Mental State Examination (MMSE) [12, 15].

There are few Arabic language scales for patients with dementia. Several studies have attempted to translate and validate cognitive and functional assessment and screening tools to the Arabic language. One example is the recent Arabic validation of the Katz Activity of Daily Living Index [16]. Katz ADL scores are strongly associated with MMSE scores (r = 0.7) and with Clinical Dementia Rating scale scores (r = –0.8) [17–19]. However, the scale by Katz et al. assesses basic activities of daily living, which only become impaired in advanced stages of dementia [18, 19], reducing the usefulness of the scale for assessing dementia in the early stages, or for detecting early signs of functional decline.

The original BADLS was developed for English-speaking people residing in western countries [15], therefore, possibly introducing bias if used to assess patients with different language or culture without proper translation and validation. Given the importance of psychometrically sound scales in the Arabic language to assess the functional performance of Arabic people with dementia [20], we aimed to translate, cross-culturally adapt, and validate the BADLS to the Arabic language for people with neurocognitive decline and dementia.

METHODS

RESULTS

The Arabic BADLS assessment was completed by the participant’s primary caregiver for a total of 69 individuals; 94% of participants were recruited from the outpatient setting in Saudi Arabia and Egypt. The median age of the patients was 77 years (IQR 68, 82), with 21.7% less than 65 years, and 32 (46.4%) were male (Table 1). The study participants had diverse educational background, forty-five percent of theparticipants were either had no formal education or received 4 years of education which is typical of older people in Arabic countries. Around 67% of the female participants were illiterate (had no formal education) or received 4 years of education. While around 33% were college or college equivalent graduates, most of those these were male (70%) and they tended to be younger with mean age of 58 years. More than half (52.2%) of the caregiver informants completing the assessments were the participant’s daughter.

The mean total score for the BADLS was 21.5±17.7, ranging from 0–56. The Arabic BADLS had excellent internal consistency, Cronbach’s alpha 0.95 (95% CI 0.93–0.96) and the corrected item total correlations for the BADLS items were all > 0.48 (Table 2).

Confirming concurrent validity, the Arabic BADLS was significantly and negatively correlated with the Arabic Katz ADL. Likewise, the Arabic BADLS also had strong convergent validity with the MoCA (see Table 3) indicating that those with better cognition were more independent in their ADL function. Also as expected, older patients and those with less education were more dependent on the Arabic BADLS, with women also being more dependent, especially when older. In univariate regression analysis, the BADLS score increased by 2.7 points for every 5-year increase in age above 65 years, and by 8.2 for females, and decreased by 16.1 for individuals with more than 6 years of education (Table 4). In our sample, younger age and male sex were associated with higher education attainment. The VIF for education, gender, and education was 1.15, 1.14, and 1.03, respectively, indicating no significant collinearity. The multivariable model for BADLS score adjusted for age and sex showed increased dependency scores by 8.7 points only for females when adjusted for age more than 65 years. The model explained 33% of variability in the data (Table 5).

Moreover, sex differences were apparent in terms of item endorsement on the BADLS scale (Table 6). Males were more likely to endorse the “not applicable” option for the “food preparation”, “housework”, and “drink preparation” items. On the other hand, females were more likely to endorse the “not applicable” option for the “transportation” and “games and hobbies” items.

DISCUSSION

This study is the first to translate and cross-culturally adapt the original BADLS to the Arabic language in individuals with varying degrees of cognitive function or dementia. The translated scale demonstrated excellent internal consistency. Despite making changes to some items to make them more culturally appropriate, all items, whether culturally modified or not, had strong item-total correlations suggesting they were still tapping into the same underlying construct. There was also evidence of excellent validity, supported by strong correlations with the Katz ADL, MoCA, and age, implying that the Arabic BALDS is likely able to discriminate between dementia disorders of different severity.

In contrast with the original BADLS [15], the scores of the Arabic BADLS were associated with sex and education, as demonstrated in the higher dependency scores for females over 65. This interesting demographic observation is likely related to the government not mandating education for women until the mid 1970s.

All the four culturally modified items had good face validity; these modifications were necessary to accommodate for modern changes to modern life (around finances) and to what is considered normal for Arabic culture. Arabic speaking countries are mostly developing countries with rich cultural and traditional characteristics that vary significantly from western countries [2, 26–28]. For example, people growing up or living in Saudi Arabia or Egypt still traditionally use their hand when eating instead of using utensils [29]. In contrast, eating by hand is considered a decline in function for people living in western countries, and this is clearly reflected in the original scoring of the BADLS [15]. Another example of an important difference is the process of self-hygiene [30, 31]. Most of the bathrooms in traditional homes do not utilize bathtubs, rather, bathing buckets or showers are more common [31, 32]. The act of bathing in a bathtub was not commonly practiced by previous generations before modern homes were built [31, 32]. All these items were adapted in the Arabic BADLS to reflect cultural aspects of these common activities of daily living for people originating from Arabic speaking countries.

The high frequency of “not-applicable” responses to food or drink preparation and housework, particularly for men is worth noting. This clearly contrasts with the original development of the BADLS where less frequent “not applicable” response options were chosen [15]. One explanation is cultural gender roles in Arabian societies, with very few females engaging in the work force during their youth in the preceding generations, devoting their time completely to household duties such as meal preparation and housework activities, whereas men used transportation more frequently for work and income generation [33]. Likewise, the great majority of primary caregivers who completed the questionnaires were theparticipants’ children, and in two thirds of the cases this child was the daughter. Certain cultural and religious values in society encourages providing continuous care for elders at home, with an automatic transfer of responsibilities to younger individuals in the family, thereby relieving elderly individuals early on from performing many responsibilities they would be normally doing had they been living in an environment with less family support [33, 34].

Functional assessment scales are essential when assessing patients with dementia and cognitive impairment in the clinic. Hence, there is a huge potential for using the BADLS in this population, especially considering that it has already demonstrated an ability to track changes over time, including being sensitive to response to treatment [35]. Given the emergence of novel therapies in treating Alzheimer’s disease, we hope the Arabic BADLS will be a useful tool in monitoring response to therapy [36]. The scale will be similarly valuable in allowing for early recognition of dementia or cognitive impairments.

Our study has several strengths: the translation process was rigorous, and we elicited feedback from patients coming from various Arab countries. In addition, we performed regression analysis to help inform policymakers in tracking the changes in BADLS score while planning for preventative strategies and screening programs for dementia. However, we were not able to perform confirmatory factor analysis because of the small sample size. Also, the psychometric properties of the scale might change over time because of the rapid changes in the literacy level in Saudi Arabia. In our sample, the literacy level was lower than the previously reported 50% in those over 65 years of age or older [37].