Study Design and Setting
A cross-sectional design was used in the current study. Convenience sampling of 182 participants was used to obtain moderate size correlation (r = 0.35), given a power of 0.80, two-sided Type I error rate of p = 0.0167, based on nQuery calculation . The participants were recruited from 5 holistic nursing homes, including private and public nursing homes, located in the center of Jordan. Inclusion criteria were as follows: Jordanian participants, aged 60 years and older, and living in nursing homes. Exclusion criteria were neurological disorders, such as stroke or brain injury.
A demographic and health survey was developed by the principal investigator, including age, sex, education level, household income, marital status, impaired activities of daily living, the number of comorbidities, and the number of hospitalizations during the year before (See Appendix A). The Arabic version-Montreal Cognitive Assessment (MoCA) tool was used to detect both mild and moderate cognitive impairment based on its total score  (See Appendix B). It has 10 items about different cognitive domains (executive functions, memory, language, attention and concentration, conceptual thinking, calculations, visuoconstructional skills, and orientation). The total score of the test is a 30-point score and it takes 10 minutes to administer. A score of 26 or above indicates no cognitive impairment. The sensitivity of the cutoff score to detect MCI is 90% and to detect mild Alzheimer’s disease is 100%, and its specificity has been reported to be excellent (87%) . A total score of 18 to 25 was considered mild cognitive impairment, and a score of 10 to 17 was considered moderate cognitive impairment  or dementia. The MoCA tool was developed by Dr. Nasreddine and colleagues (2005), who has extensively used this tool in detecting cognitive impairment in nursing home residents. Using the MoCA has been well-documented in the literature to determine mild and moderate cognitive impairment . In conclusion, the MoCA is the most precise cognitive ability measure that differentiate between the stages of cognitive impairment and had excellent sensitivity and specificity values to detect MCI and mild Alzheimer’s disease  compared to other cognitive ability measures.
The Arabic version-geriatric depression scale (GDS) was used to detect depression and it has 15 items. The total score was 15 and its scores were distributed as follows: 0-4 normal, 5-8 mild depression, 9-11 moderate depression, and 12-15 severe depression . Greenberg’s study revealed that the GDS-15 had high sensitivity and specificity of 92% and 89%, respectively  (See Appendix C).
The Arabic version-HRQoL (Short Form (SF-36)) was used in the current study. The SF-36 health survey entails 2 main domains (physical and mental). The SF-36 health survey has 36 questions, including 8 categories: physical functioning (PF), general health (GH), vitality (V), social functioning (SF), role-physical (RP), bodily pain (BP), mental health (MH), and role-emotional (RE). Each category has a 0-100 scale calculated using a special software. Lower score indicates more disability, and a higher score shows less disability/higher HRQoL in its category . The scores of the Arabic-SF-36v2® were calculated using Qualtrics Software (Medical Outcomes Trust) (See Appendix D).
The Arabic version-Tilburg frailty indicator (TFI) was used in the current study to assess frailty among Jordanian older adults. The original TFI created by Gobbens and colleagues  has 15 items, entailing 3 domains: physical, psychological, and social domains. The reliability (KR 20= 0.77) and convergent and divergent validity of the Arabic version-TFI have been established in Jordanian population . The TFI is the only frailty instrument that has been translated and validated to use in Jordanian older adult population . It has only fifteen questions , which is easy to administer in older adult population (See Appendix I).
After institutional review board (IRB) approval (#20180229) from both the Jordan University of Science and Technology and targeted nursing homes, participants of the study were recruited through nursing home visits and asked to participate in the study. An overseeing mental health expert have ruled that all participants have been deemed capable of ethically and medically consenting for their participation in the study. Consent forms were delivered to or read aloud to the older adults prior to collecting data from them. To avoid drift resulting from different data collectors, a training session on how to gain information using the instruments used in the current study was conducted for data collectors.
Bivariate analysis, including t-test and ANOVA test, was used to examine the correlations between health variables and mental ability. Thereafter, multivariate analysis, including logistic and linear regression models was conducted, for the correlations between health variables and both mild and moderate cognitive impairment. The statistical analysis was performed using the Statistical Package for Social Sciences (SPSS) version 25 (SPSS, Inc, Chicago, Ill).