Profile of mental disorders among the elderly United Arab Emirates population: sociodemographic correlates

Transcription

1 INTERNATIONAL JOURNAL OF GERIATRIC PSYCHIATRY Int J Geriatr Psychiatry 2004; 19: Published online in Wiley InterScience ( DOI: /gps.1101 Profile of mental disorders among the elderly United Arab Emirates population: sociodemographic correlates Rafia Ghubash 1,2, Omer El-Rufaie 1 *, Taoufik Zoubeidi 3, Qasim M. Al-Shboul 2 and Sufyan M. Sabri 1 1 Department of Psychiatry, Faculty of Medicine and Health Sciences, UAE University, Al-Ain, UAE 2 Department of Family & Community Medicine, College of Medicine and Medical Sciences, Arabian Gulf University, Manama, Kingdom of Bahrain 3 Department of Statistics, Faculty of Business and Economics, UAE University, UAE SUMMARY Objectives To investigate the prevalence, nature and sociodemographic correlates of mental disorders among the elderly United Arab Emirates (UAE) population. Study subjects and sample UAE nationals aged 60 years or more, were recruited from within a random sample of households representing the UAE national population, irrespective of the age of individuals in each household. Research Instruments (i) Geriatric Mental State Interview (GMS-A3): an Arabic version, using the AGECAT for analysis; (ii) A short questionnaire for relevant sociodemographic data. Procedure Purposely trained, Arabic speaking interviewers visited the targeted sample households to interview study subjects at their homes. Results The total number of screened subjects was 610: 166 (27.2%) in Al-Ain; 286 (46.9%) in Dubai and 158 (25.9%) in Ras Al-Khaimah. There were 347 (56.9%) male subjects and 263 (43.1%) female subjects. The mean age of the interviewed subjects was 68.6 (SD 8.3). The commonest diagnostic entities at the AGECAT syndrome case level were depression (20.2%), anxiety (5.6%), hypochondriasis (4.4%) and organic, mostly cognitive impairment with or without dementia (3.6%). Organic syndrome caseness, as an independent entity, showed significant correlation only to older age, while the rest of the mental disorders showed significant correlation with female gender, insufficient income and being single, separated, divorced or widowed. Conclusion The GMS-AGECAT package proved to be a useful tool for psychiatric assessment among the elderly in this Arabian culture. The prevalence rates of mental disorders among the elderly UAE population were, more or less, within the same range reported by other comparable worldwide studies. Copyright # 2004 John Wiley & Sons, Ltd. key words Geriatric Mental State Interview (GMS-A3); AGECAT; United Arab Emirates (UAE); mental disorders; old age INTRODUCTION *Correspondence to: Professor O. El-Rufaie, Department of Psychiatry and Behavioural Sciences, Faculty of Medicine and Health Sciences UAE University, P.O. Box 17666, Al-Ain, United Arab Emirates. Tel: Fax: elrufaie@uaeu.ac.ae Contract/grant sponsor: Sheikh Hamdan Bin Rashid Al Maktoum Award; contract/grant number: MRG-17/ Elderly people are at a particular risk of suffering from dementia, depression and other psychiatric disorders. Estimates of the prevalence of mental disorders among the elderly show considerable variations among the studied samples (Kay and Bergman, 1980; Henderson and Kay, 1984; Copeland et al., 1987a). It is probable that the methodological procedures adopted in the various studies account for much of the variations. Psychiatric research in the Gulf Region, in particular, and in the Arab world in general, is relatively scanty in comparison with the Western world and other developed countries. Studies of psychiatric morbidity among the elderly population in this geographic region are even more scanty. Among other objectives, this study is intended to establish Received 7 October 2003 Copyright # 2004 John Wiley & Sons, Ltd. Accepted 8 January 2004

2 mental disorders among elderly uae population 345 preliminary base-line data on the mental health of the elderly UAE population. Over the last three to four decades the UAE society has been evolving over a period of rapid urbanisation, socio-economic advancement, modernisation and change in the quality of life. Many western qualities and ways of life have been introduced. It is inevitable that the integrity of both the nucleus and extended families will be affected by such quick and dramatic change. The shift in the family set up, and the society as a whole, from conservative Islamic society, governed to a great extent by the Bedouin values and culture to modern sophisticated life with Western rhythm and style, will inevitably reflect in various ways on the elderly people s welfare, health and prosperity. The specific aim of this study was to investigate the prevalence, nature and pertinent sociodemographic correlates of mental disorders among the elderly UAE population. In this study the broad spectrum of mental disorders in the elderly were investigated. Cognitive impairment and depression among the elderly have been extensively studied but other psychiatric disorders, including anxiety states, were not adequately investigated. The fact that the elderly are amongst the heaviest consumers of anxiolytics (Skegg et al., 1977; Catalan et al., 1988; Lindesay et al., 1989), is suggestive of the significance of anxiety and other neurotic disorders among this sector of the population (Lindesay et al., 1989). This is why we opted to investigate a wider range of mental disorders among the elderly, rather than restricting the study to depression and cognitive impairment. METHOD Study sample Study subjects were UAE nationals aged 60 years or more. Due to the high illiteracy rate among the elderly in this area, many of them do not know their exact age. The study interviewers were trained to crossvalidate the age stated by subjects with particular important events in their lifetime, and by information from close reliable relatives. The study sample was provided by the Central Department of Statistics, Ministry of Planning, Abu Dhabi. It consists of a random sample of 2000 households representing UAE local population irrespective of the age of the individuals in each household. The same sample was also used for the purpose of another child psychiatry study scheduled to be carried out simultaneously with this study. The sampling frame which was based on the UAE 1995 population census was also updated in 1998 to take into account the formation of new citizen households. Both the compilation of the sample frame and the drawing of the master sample were supervised by two sampling experts from the United Nations Statistical Office. The master sample was designed as a two-stage, stratified, cluster sample of approximately 4000 citizen households in 210 primary sampling units (PSUs). The PSUs consist of census enumeration areas in urban sectors and villages in rural sectors. The PSUs were stratified into six size categories, according to the number of citizen households per PSU. A cluster of PSUs was then selected from each stratum. The sampling scheme gave equal probability of selection to each household. For the purpose of this study 843 households were selected for screening. These households represent the population of two out of the seven Emirates, i.e. Dubai and Ras-Al-Khaimah, in addition to the population of Al-Ain, which constitutes the Eastern region of a third Emirate, i.e. Abu Dhabi. The sample included 843 households which were selected by randomly choosing approximately half the households within PSUs of the master sample that were located in the three Emirates. Institutions were not included in this sampling procedure. Further details regarding the study sample are presented in Table 1. Table 1. Sample sizes and response rates at various levels Al-Ain Dubai Ras-Al-Khaimah Total Households in the original sample frame Households without potential subjects (60 yrs or more) Total number of households which accepted to take part and subjects were screened Households, with potential subjects, which apologized 15 (10.2%) 40 (13.5%) 4 (2.5%) 59 (9.7%) from participation Number of subjects interviewed* 166 (27.2%) 286 (46.9%) 158 (25.9%) 610 (100%) Subjects apologized from within agreeable households 4 (2.4%) 12 (4.0%) 1 (0.6%) 17 (2.7%) *In some households more than one subject were interviewed.

3 346 r. ghubash ET AL. Case definition and research instruments A short preliminary questionnaire. For recording basic socio-demographic data not included in the GMS-A3, schedule. Geriatric Mental State Interview (GMS-A3). The Geriatric Mental State Interview (GMS), a standardised, semi-structured interview for examining and recording the mental state in elderly subjects. The first version was developed in 1976 (Copeland et al., 1976). Later, shortened versions were developed (GMS-6, GMS-A1, GMS-A3). The reliability and validity of the GMS have been established in inpatient, out-patient and community samples. Studies established consistently high level of agreement between psychiatrists and non-psychiatrist raters. The schedule was translated and validated in over 17 languages. For the purpose of this study the GMS-A3, designed mainly for community studies, was used. It is the most recent version for the full range of mental illnesses (183 items). It seeks to establish symptoms experienced by the subjects in the four weeks before the interview. Diagnoses were made by means of a computer-assisted system, the Automated Geriatric Examination for Computer Assisted Taxonomy (AGECAT). The AGECAT was also developed by Copeland et al. (1986) as a computerised system which could be applied to data derived from the GMS by grouping the symptoms to form a pattern recognised by a psychiatrist as illness, usually identified as syndrome case (Copeland et al., 1986; Copeland et al., 1987b). The AGECAT diagnostic system is divided into a number of logical stages. It brings together items from the GMS and other schedules into symptom components, which are then condensed to form eight diagnostic clusters. Each cluster consists of groups representing symptoms important for diagnosis. Scores on each group determine the level of diagnostic confidence for each subject on each diagnostic cluster. The diagnostic clusters include organic, schizophrenic, manic, depressive (psychotic and neurotic), hypochondrical, obsessional, phobic and anxious. In the second stage the levels are compared across clusters so as to determine whether or not the symptoms rank as a syndrome case. All clusters with diagnostic confidence level of 3, 4 and 5 represent syndrome case levels, while 1 and 2 sub-case levels. When a subject reaches the case level in several syndromes, then the AGECAT will select one of these as the overall diagnosis, which is referred to as diagnostic syndrome. For the purpose of this study we report cases as identified by the AGECAT at the syndrome case level. This may result in over-diagnosis of cases which may not lead to psychiatric intervention if judged by a psychiatrist. However, there is evidence of agreement between psychiatrists and AGECAT diagnoses for both institutionally based and community based subjects (Copeland et al., 1986; Copeland et al., 1987b). Translation and training of interviewers One Arabic speaking doctor and seven arabic speaking nurses, were recruited as interviewers for this study. All of them had previous experience with community psychiatric research. The Egyptian translation of the GMS-A3 into Arabic was used. The bilingual psychiatrist who led the group of translators, and one of his colleagues were invited to train our interviewers on the methods of administration. During the training videotaped and live interviews, simulated and real patients, were used. The training emphasised conveying the exact original meaning of each item and the standardisation of administration. Procedure In each of Al-Ain, Dubai and Ras-Al-Khaimah, the simultaneous child psychiatry study team, using the same sampling frame, visited households according to the sample frame map. An additional assignment for the team was to report, for the purpose of this study, about presence of subjects 60 years or over within each household visited. Households with potential study subjects (60 years or over) were visited by a representative of this study to deliver a letter which explained the nature of the study and requested consent to participate. A team of two interviewers visited each household, usually late in the afternoon and normally after a telephone call. After summarising the nature of the study and ensuring consent both the sociodemographic questionnaires and the Arabic version of the GMS-A3 were administered. Statistical analysis Data was entered in the AGECAT computerized diagnostic system. The system analyzed the data into a number of logical stages: first symptoms components then symptom clusters. The outcome of the latter together with the socio-economic variables were coded, re-entered and analyzed using the Statistical

4 mental disorders among elderly uae population 347 Package for Social Sciences (SPSS). The inter-rater reliability was assessed using 40 subjects. Each pair of our four pairs of interviewers assessed ten subjects. Each interviewer recorded his own findings independently. Calculation of the inter-rater reliability was based on the chance-corrected agreement coefficient AC1 (Gwet, 2001), which avoids the well-known weaknesses of the Kappa coefficient (Cohen, 1968). The values of AC1 range between 0 and 1, where 1 represent a perfect agreement while a 0 represents no agreements beyond those made by chance. The chi-squared test was used to determine the sociodemographic variables that were significantly correlated to mental disorder syndrome cases. Simple logistic regression analysis was used to assess the individual effect of the variables on syndrome caseness. The level p < 0.05 was considered to be the cut-off value for significance. RESULTS Statistical analysis revealed high levels of agreement among the interviewers in all the studied syndrome cases. The mean and range of AC1 over the four pairs of interviewers were: 0.86 (0.77 1) for organic syndrome caseness, 0.92 (0.9 1) for schizophrenic syndrome caseness, 1 (1 1) for manic syndrome caseness, 0.89 (0.77 1) for depressive syndrome caseness, 0.97 (0.89 1) for hypochondriasis syndrome caseness, 0.94 (0.88 1) for phobic syndrome caseness, and 0.83 (0.66 1) for anxiety syndrome caseness. The sample size in the three research cites and the response rate at the levels of the households and individual subjects are presented in Table 1. A range of relevant sociodemographic data (except for the occupational status) is presented in Table 2. None of the females interviewed had a regular occupation. Out of 347 men, 104 (30%) were working at the time of interview. Among those, 55 (53%) were unskilled labourers, ten (10%) were in business or trade and the rest were distributed in small proportion in other jobs, including skilled labour (4.8%), army and police (3.8%), clerical (3.8%), professionals (2.9%), teachers (1%) and fishermen (1%). The diagnostic entities as identified by AGECAT, at the syndrome case level are presented in Table 3. Comorbidity of hypochondriasis with anxiety, depression and comorbid anxiety and depression was investigated further more. The results are outlined in Table 4. An additional further analysis was also performed for the category of organic syndrome caseness, by estimating the prevalence among the sample population aged 65 or more, which turned out to be 5.1%, in comparison to 3.6% for those aged 60 or more (Table 3). Simple logistic regressions between each of five pertinent variables and caseness in each of the eight categories of mental disorders show that gender, marital status, and income sufficiency were significantly correlated with mental disorder. A stepwise logistic regression was also fitted between the same group of variables and caseness in each of the mental disorders categories. Again, gender, marital status, and income sufficiency showed statistically significant correlation (Table 5). To assess the association of the same variables with organic syndrome caseness (as an independent entity, without comorbidity with other mental disorder), simple logistic regressions were fitted with each of the variables. Except for age, none of the variables had a significant correlation (Table 6). DISCUSSION The overall response rate among the sample households, with potential subjects, was 90.3%. Among individual subjects the response rate was even better at 97.3%. The sample screened, 610 subjects out of 843 households, appears to be smallish. This was an inevitable consequence of the fact that, due to financial and other constraints, we limited our screening to only two Emirates and a region of a third one, out of the total seven Emirates in the country. The commonest identified AGECAT syndrome case categories among the entire sample were depression (20.2%), anxiety (5.6%), hypochondriasis (4.4%) and organic (3.6%). This is in conformity with the results of other relevant studies in identifying depression and organic, mostly cognitive impairment, as consistently common mental health problems among the elderly. Also the prevalence rate of anxiety disorder immediately follows these two conditions in many studies. However, the identification of hypocondriasis in a small proportion among our syndrome cases, appears to be related to the psychometric properties of the GMS, which is designed to generate this entity, rather than an indication of the clinical significance of hypochondriasis as an independent diagnostic entity among the sample. A significant finding in this regard, was the firm association and comorbidity between hypochondriasis in one hand, and anxiety, depression or comorbid anxiety and depression on the other hand (Table 4). Organic, mostly cognitive impairment with or without dementia, was estimated as 3.6%. A similar figure of 3.5% was reached in an Indian community

5 348 r. ghubash ET AL. Table 2. Relevant economic and sociodemographic data Al-Ain Dubai Ras-Al-Khaimah Total n (%) n (%) n (%) n (%) Gender Male 100 (60.2) 147 (51.4) 100 (63.3) 347 (56.9) Female 66 (39.8) 139 (48.6) 58 (36.7) 263 (43.1) Age groups (57.8) 173 (60.5) 89 (56.3) 358 (58.7) (25.3) 79 (27.6) 48 (30.4) 169 (27.7) (12.7) 28 (9.8) 17 (10.8) 66 (10.8) 90 and above 7 (4.2) 6 (2.1) 4 (2.5) 17 (2.8) Mean Median SD Personal monthly income Less than 5000 Dhs* 59 (35.5) 188 (65.7) 120 (75.9) 367 (60.2) Dhs 58 (34.9) 41 (14.3) 32 (20.3) 131 (21.5) Dhs 8 (4.8) 19 (6.6) 1 (0.6) 28 (4.6) Dhs 1 (0.6) 2 (0.7) 1 (0.6) 4 (0.7) More than Dhs 0 (0.0) 6 (2.1) 0 (0.0) 6 (1.0) Unreported 40 (24.1) 30 (10.5) 4 (2.5) 74 (12.1) Income sufficiency, according to individual subjects Excellent 60 (36.1) 51 (17.8) 18 (11.4) 129 (21.1) Acceptable 86 (51.8) 173 (60.5) 101 (63.9) 360 (59.0) Not sufficient 16 (9.6) 61 (21.3) 36 (22.8) 113 (18.5) Unreported 4 (2.4) 1 (0.3) 3 (1.9) 8 (1.3) Marital status Single 1 (0.6) 1 (0.3) 0 (0.0) 2 (0.3) Married 118 (71.1) 174 (60.8) 107 (67.7) 399 (65.4) Divorced 7 (4.2) 12 (4.2) 1 (0.6) 20 (3.3) Widowed 38 (22.9) 98 (34.3) 50 (31.6) 186 (30.5) Unreported 2 (1.2) 1 (0.3) 0 (0.0) 3 (0.5) Family (household) compound Husband and/or wife and their children 72 (43.4) 152 (53.1) 109 (69.0) 333 (54.6) Extended family 60 (36.1) 84 (29.4) 37 (23.4) 181 (29.7) Extended family and others 25 (15.1) 34 (11.9) 11 (7.0) 70 (11.5) Living alone 7 (4.2) 14 (4.9) 0 (0.0) 21 (3.4) Unreported 2 (1.2) 2 (0.7) 1 (0.6) 5 (0.8) Education Illiterate 141 (84.9) 233 (81.5) 138 (87.3) 512 (83.9) Elementary 19 (11.4) 40 (14.0) 17 (10.8) 76 (12.5) Preparatory 3 (1.8) 6 (2.1) 2 (1.3) 11 (1.8) Secondary 2 (1.2) 3 (1.0) 0 (0.0) 5 (0.8) University 1 (0.6) 4 (1.4) 0 (0.0) 5 (0.8) Post graduate 0 (0.0) 0 (0.0) 1 (0.6) 1 (0.2) *US Dollar ¼ 3.68 Dirhams. Table 3. Diagnostic entities at the syndrome case level in Al-Ain, Dubai and Ras-Al-Khaimah Al-Ain (n ¼ 166) Dubai (n ¼ 286) Ras-Al-Khaimah (n ¼ 158) Total (n ¼ 610) n (%) n (%) n (%) n (%) Organic 6 (3.6) 12 (4.2) 4 (2.5) 22 (3.6) Schizophrenia 0 (0.0) 1 (0.3) 3 (1.9) 4 (0.7) Mania 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Depression 19 (11.4) 84 (29.4) 20 (12.7) 123 (20.2) Obsession 1 (0.6) 7 (2.4) 3 (1.9) 11 (1.8) Hypochondriasis 5 (3.0) 15 (5.2) 7 (4.4) 27 (4.4) Phobia 0 (0.0) 1 (0.3) 0 (0.0) 1 (0.2) Anxiety 2 (1.2) 24 (8.4) 8 (5.1) 34 (5.6)

6 mental disorders among elderly uae population 349 Table 4. Comorbidity of hypochondriasis with anxiety, depression and comorbid anxiety and depression Diagnostic entity From the Comorbidity with total sample hypochondriasis n (%) n (%) Anxiety 34 (5.6) 7 (20.6) Depression 123 (20.2) 16 (13.0) Comorbid anxiety 30 (4.9) 7 (23.3) and depression study, also using the GMS (Bhatnagar and Frank, 1997). However, the estimate of 3.6% tends to be in the lower border, in comparison to the estimates of most other comparable studies. A large Liverpool sample of subjects aged over 65 years, also using the GMS, determined a prevalence rate of 5.2% at the diagnostic syndrome level (Copeland et al., 1987a). A prevalence rate of 4.6% of cognitive impairment was determined in another UK study, using CARE organic brain syndrome scale (Lindesay et al., 1989). However, it should be noted that the subjects of this study were younger than those in the previous studies. The entry age to our study was 60 years (mean age 68.6) while the entry age for the other studies was 65. This was substantiated by computing the prevalence among those aged 65 or more within our sample, which turned out to be 5.1%. Evidence is well established that the prevalence of all types of cognitive impairment, including dementias, increases with age (Graham et al., 1997) and probably with no gender differences (Fichter et al., 1995). This was also confirmed by our finding that age was the only variable, among others, to demonstrate significant correlation with organic syndrome caseness (Table 6). The AGECAT syndrome depression among the entire study sample was 20.2%. This appears to be relatively higher than the prevalence estimated in comparable studies. In Copeland and colleagues Liverpool study the overall AGECAT depression was 11.3% (Copeland et al., 1987a), while in a Taiwan community study using the GMS, depressive neurosis was 15.3%, and major depression was 5.9% (Chong et al., 2001). A range from 5 to 19% was determined among elderly people in black and ethnic minorities in the UK (McCracken et al., 1997). A Canadian study of elderly community residents, determined 11.4% as AGECAT depression (Newman et al., 1998). It may be interesting to note that the AGECAT depression in our Al-Ain sample was 11.4%, in Ras-Al-Khaimah 12.7%, in Dubai 29.4% and the overall prevalence was 20.2%. The prevalence rate of depression was within the same range in both Al-Ain and Ras-Al-Khaimah, while significantly higher in Dubai. The result of hierarchical logic regression between depression syndrome caseness in the three sites, after controlling for age, marital status and income sufficiency, confirmed the clinical significance of the elicited difference between Dubai and the other two regions. This is suggestive of the need Table 5. Correlation of the group of subjects with identified caseness in any of the range of mental disorders, with five pertinent variables (n ¼ 155) Variable Non-syndrome cases Syndrome cases p-value Odds ratio n (%) n (%) Gender Male 275 (79.3) 72 (20.7) Female 180 (68.4) 83 (31.6) Age groups (76.5) 109 (23.5) (71.2) 32 (28.8) and above 22 (61.1) 14 (38.9) Marital status Currently married 318 (79.7) 81 (20.3) Currently single, divorced, separated, widowed 135 (64.1) 73 (35.1) Income sufficiency Excellent 109 (84.5) 20 (15.5) Acceptable 273 (75.8) 87 (24.2) 1.7 Not sufficient 66 (58.4) 47 (41.6) Education Illiterate 377 (73.6) 135 (26.4) Less than secondary 70 (80.5) 17 (19.5) Secondary and above 8 (72.7) 3 (27.3) p-values computed simple logistic regression.

7 350 r. ghubash ET AL. Table 6. Correlation of organic syndrome caseness with five pertinent variables (n ¼ 11) Variable Non-organic syndrome cases Organic syndrome cases p-value Odds ratio n (%) n (%) Gender Male 340 (98.0) 7 (2.0) Female 259 (98.5) 4 (1.5) Age groups (99.4) 3 (0.6) (96.4) 4 (3.6) and above 32 (88.9) 4 (11.1) Marital status Currently married 393 (98.5) 6 (1.5) Currently single, divorced, separated, widowed 203 (97.6) 5 (2.4) Income sufficiency Excellent 128 (99.2) 1 (0.8) Acceptable 351 (97.5) 9 (2.5) Not sufficient 112 (99.1) 1 (0.9) Education Illiterate 501 (97.9) 11 (2.1) Less than secondary 87 (100.0) 0 (0.0) Secondary and above 11 (100.0) 0 (0.0) p-values computed using simple logistic regression. for further rigorous research to identify the factors behind this observation. Perhaps, the fast rhythm of change in all aspects of life in urban Dubai compared to the rural and simpler life style in other two places, may contribute, directly or indirectly, to this finding. The total prevalence rate of anxiety in this study was 5.6%. Again this is higher than the rates reported in similar studies, especially those using the GMS/ AGECAT package. The Copeland et al. New York and London comparative study determined by far smaller rates for all range of neuroses (Copeland et al., 1987b). Another community study, also using the same methodology, in a sample of elderly Asian immigrants living in the UK, made an estimate of 4% for anxiety neurosis among the studied sample (Bhatnagar and Frank, 1997). However, there is accumulating evidence suggesting that anxiety and anxiety disorders are not prominent health problems in old age. A relatively recent German study indicated that the prevalence rate of anxiety in the younger old (70 84 yrs) was 4.3%, while in the older old ( yrs) was 2.3%. There were more phobic symptoms in the younger age group, and females showed significantly higher anxiety (Schaub and Linden, 2000). Further more, the extremely low rate of phobic disorder reported in this study, compared to other relevant studies, may be related to the nature of the phobic questions in the GMS which may exclude some fears if they are thought to be normal for elderly people. It is also claimed that clinical anxiety among the elderly seems to be confined to those who also suffer clinical depression (Yohannes et al., 2000). It is argued that aging is associated with intrinsic reduction in the susceptibility to anxiety and depression, and this was attributed to the possibility of reduced emotional responsiveness, increased emotional control and psychological immunisation to stressful life events in old age (Jorm, 2000). For confirming the validity of the comparison made between our anxiety estimates and the estimates of other cited relevant studies, it is crucial to take into consideration the level of severity used to discriminate between syndrome caseness and non-caseness, so as to match the diagnostic confidence level of 3 or more criteria used in this study. A comparison was made between the groups with identified mental disorders (syndrome cases) with the group without evidence for mental disorder (non-syndrome cases), in relation to five variables (Table 5). Mental disorders were significantly higher among the female compared to the male subjects. This is consistent with the findings of similar studies (Catalan et al., 1988; Yohannes et al., 2000). Married subjects living with their spouses were compared with the single, divorced, separated or widowed, i.e. living alone. In conformity with other relevant published works, there was significant increase of mental disorders among those living alone (Baker et al., 1996). On comparing income sufficiency, as reported by the subjects themselves, there was significant increase of

8 mental disorders among elderly uae population 351 mental disorder among those who had reported insufficient income. There was no significant correlation between the level of education and the occurrence of mental disorder in old age. Apparently the association of last two variables, income and education, with mental disorder among the elderly was not researched enough to make a meaningful comparison. Although the Arabic version of the GMS-A3 used in this study proved to be feasible for use among our sample, the main drawback was the long time needed for administration. It took each interviewer about minutes to interview a study subject. However, as the study progressed, the time needed diminished as the interviewers gained experience and mastered the technique of application. Perhaps, a drawback to the AGECAT diagnoses was in connection with the category of organic syndrome caseness which, theoretically, includes all degrees of cognitive impairment, with and without dementia, in addition to other organic brain dysfunctions. The fact that the GMS- AGECAT package is not designed to identify independent diagnostic entities of dementia or cognitive impairment is a limitation to this particular research methodology. ACKNOWLEDGEMENT The authors would like to express their sincere thanks to the Sheikh Hamdan Bin Rashid Al Maktoum Award for Medical Sciences who supported this project by the research grant (MRG-17/ ). Authors would like also to thank Dr Emad Hamdi Ghoz and Dr Youserya Amin for their continuous advice, and Mr D. Ranganathan for preparation of the manuscript. REFERENCES Baker FM, Okwumabua J, Philipose V, Wong S Screening African-American elderly for the presence of depressive symptoms: a preliminary investigation. J Geriatr Psychiatry Neurol 9(3): Bhatnagar K, Frank J Psychiatric disorders in elderly from the Indian sub-continent living in Bradford. Int J Geriatr Psychiatry 12(9): Catalan J, Gath DH, Bond A, et al General practice patients on long-term psychotropic drugs: a controlled investigation. Br J Psychiatry 152: Chong MY, Chen CC, Tsang HY, et al Community study of depression in old age in Taiwan. Br J Psychiatry 178: Cohen J Weighted kappa: Nominal scale agreement with provision for scaled disagreement or partial credit. Psychol Bull 70(4): Copeland JRM, Dewey ME, Griffiths-Jones HM Computerized psychiatric diagnostic system and case nomenclature for elderly subjects: GMS and AGECAT. Psychol Med 16: Copeland JRM, Dewey ME, Wood N, Searle R, Davidson IA, McWilliam C. 1987a. Range of mental illness among the elderly in the community: prevalence in Liverpool using the GMS- AGECAT package. Br J Psychiatry 150: Copeland JRM, Gurland BJ, Dewey ME. 1987b. Is there more dementia, depression and neurosis in New York? A comparative study of the elderly in New York and London using the computer diagnosis AGECAT. Br J Psychiatry 151: Copeland JRM, Kelleher MJ, Kellett JM, et al A semi-structured clinical interview for the assessment of diagnosis and mental state in the elderly: The Geriatric Mental State Schedule. I. Development and reliability. Psychol Med 6: Fichter MM, Meller I, Schroppel H, Steinkirchner R Dementia and cognitive impairment in the oldest old in the community: prevalence and comorbidity. Br J Psychiatry 166: Gwet K Handbook of Inter-Rater Reliability: How to Estimate the Level of Agreement Between Two or Multiple Raters. Stataxis publishing: Gaithersburg, MD, USA. Graham JE, Rockwood K, Beattie BL, et al Prevalence and severity of cognitive impairment with and without dementia in an elderly population. Lancet 349(9068): Henderson AS, Kay DWK The epidemiology of mental disorders in the aged. In Handbook of Studies in Psychiatry and Old Age, Kay DWK, Burrows G (eds). Elsevier: Amsterdam. Jorm AF Does old age reduce the risk of anxiety and depression? A review of epidemiological studies across the adult life span. Psychol Med 30(1): Kay DWK, Bergman K Epidemiology of mental disorder among the aged in the community. In Handbook of Mental Health and Aging, Birren JE, Sloan B (eds). Prentice Hall: Englewood Cliffs, NJ. Lindesay J, Briggs K, Murphy E The Guy s/age Concern Survey: prevalence rates of cognitive impairment, depression and anxiety in an urban elderly community. Br J Psychiatry 155: McCracken CFM, Boneham MA, Copeland JRM, et al Prevalence of dementia and depression among elderly people in Black and ethnic minorities. Br J Psychiatry 171: Newman SC, Sheldon CT, Bland RC Prevalence of depression in an elderly community sample: a comparison of GMS- AGECAT and DSM-IV diagnostic criteria. Psychol Med 28(6): Schaub RT, Linden M Anxiety and anxiety disorders in the old and very old-results from the Berlin Aging Study (BASE). Comp Psychiatry 41(2 suppl 1): Skegg DG, Doll R, Perry J Use of medicines in general practice. Br Med J 1(6076): Yohannes AM, Baldwin RC, Connolly MJ Depression and anxiety in elderly outpatients with chronic obstructive pulmonary disease: prevalence, and validation of the BASDEC screening questionnaire. Int J Geriatr Psychiatry 15(12):