Copyright © 1998 American Academy of Neurology

Volume 50(4)             April 1998             pp 1002-1009

Prevalence of dementia in Chinese elderly in Hong Kong

[Articles]

Chiu, H.F.K. FRC Psych; Lam, L.C.W. MRC Psych; Chi, I. DSW; Leung, T.; Li, S. W. MRC Psych; Law, W. T. MRC Psych; Chung, D.W.S. MRC Psych; Fung, H.H.L. MA; Kan, P. S. FRCR; Lum, C. M.

MRCP; Ng, J. BHSc; Lau, J. PhD

From the Department of Psychiatry (Drs. Chiu, Lam, Law, and Chung, and T. Leung and H.H.L. Fung) and the Centre for Clinical Trials and Epidemiological Research (Dr. Lau), Chinese University of Hong Kong; the Department of Social Work and Social Administration (Dr. Chi), University of Hong Kong; the Castle Peak Hospital (Dr. Li), Hong Kong; the Department of Radiology (Dr. Kan), Prince of Wales Hospital, Shatin, Hong Kong; and the Shatin Hospital (Dr. Lum and J. Ng), Shatin, Hong Kong.Supported in part by a grant from Parke-Davis.Received February 11, 1997. Accepted in final form October 29, 1997.Address correspondence and reprint requests to Dr. Helen F.K. Chiu, Department of Psychiatry, Prince of Wales Hospital, Shatin, Hong Kong.

Outline

Article abstract Acknowledgments References

Graphics

Table Demographic ch...

Figure 1 Figure 2 Figure 3

Article abstract^

This study examines the prevalence of dementia in elderly Chinese aged 70 years and older in Hong Kong using a two-phase design. In phase 1, 1,034 elderly were interviewed with the Cantonese version of the Mini-Mental State Examination. Those who scored below the cutoff points and a subsample of those with scores in the normal range were interviewed in phase 2 to identify those with dementia. The overall weighted prevalence of dementia in our subjects was 6.1 ± 0.7%, which is at the lower end of the range of rates reported in whites. Alzheimer's disease (AD) accounted for 64.6% and vascular dementia, 29.3%. Our results, together with previous studies in Chinese populations, suggest that the rates of AD in Chinese are low compared with those in whites. Substantial differences are possible in the epidemiology of dementia across cultures related to interactions of genetic and environmental factors.

The prevalence of moderate and severe dementia has been examined in a number of studies and is found to be around 2 to 7% for people aged 65 years and over.1 These studies were mainly conducted in Europe, North America, and Australia.1 In a few recent reviews on the epidemiology of dementia, differences in the rates of subtypes of dementia were observed across countries.2,3 In general, Alzheimer's Disease (AD) is more prevalent in countries in Europe and North America, whereas vascular dementia (VD) is more frequent in Russia, Japan, and China.2,3 However, more recent studies in China and Taiwan report conflicting findings on the relative frequency of AD and VD in Chinese populations.4-14

Most studies in China published in the last decade 4-12 reported a very low prevalence of dementia, ranging from 0.5% to 5.4%. Seven of nine studies in China 4-7,9-11,13-14 reported a predominance of VD, whereas AD was found to be more common in two studies.8,12 Three recent studies in Taiwan 13-15 and a study in Singapore 16 on Chinese elderly also found a low rate of dementia, ranging from 1.8% to 2.8%, with AD being more common.

There have been few data on the prevalence of dementia in Hong Kong. Although the population in Hong Kong is ethnic Chinese (as in China and Taiwan), the social and environmental conditions, such as diet, consumption of alcohol and cigarettes, lifestyle, and health services, differ from those of other Chinese communities. Because these social and environmental factors might have an impact on the prevalence of dementia, particularly VD or perhaps even manifestation of AD, it would be of interest to study the prevalence of dementia in Hong Kong. Furthermore, in recent years, there is a growing demand for epidemiologic studies on dementia because of rapid aging of the population, resulting in an

urgent need for planning services and provisions for the elderly. We conducted a study in the community to examine the prevalence of dementia and depression in the elderly in Hong Kong. We report here on the the prevalence of dementia (the results on depression will be reported in a separate article, submitted for publication, 1998). The study has been approved by the local ethics committee.

Methods. Sample. Hong Kong has a population of 6.2 million, 96% of whom are ethnic Chinese. In 1995, the population of elderly aged 65 and over was 10% of the general population, which was one of the highest in Southeast Asia. As in other major cities, there is a constant flow of immigrants. However, because most new immigrants are young people, the population of the elderly is relatively stable. The survey was conducted in the Shatin district of Hong Kong, chosen because it was a geographically defined area within the catchment area of the Department of Psychiatry of the Chinese University of Hong Kong. The general population of Shatin is around 569,000 of which there area about 28,100 elderly aged 70 and older.

Our aim was to examine a representative sample of Chinese elderly aged 70 years and older living in Shatin. In Hong Kong, there is no such satisfactory register of elderly in the population like the general practitioner's list or the Electoral Roll, and the best available register is the old age allowance list. Old age allowance is available to all elderly aged 70 years and older who have lived in Hong Kong for at least 5 years and is independent of the person's income. Nevertheless, people with physical or mental disability who are eligible for disability allowance would have to forfeit old age allowance. Thus, the list of old age allowance recipients would undersample elderly with dementia or mental illness who are on disability allowance. For this study, we added the list of recipients of disability allowance. These two lists supplement each other and constituted the sampling frame of our study and cover approximately 94% of all elderly aged 70 years and older in Hong Kong. Subjects were recruited by a stratified random sampling from a register of all Chinese elderly receiving old age and disability allowances residing in Shatin. Subjects were stratified by age and sex to comprise approximately 150 subjects in the 70- to 74-year-old and 75- to 79-year-old age groups for each sex and 100 in the 80- to 84-year-old, 85- to 89-year-old, and 90-year-old+ age groups for each sex. These numbers were chosen to fit into the scope of our resources.

Procedures. A two-stage design was used to examine the prevalence of dementia.

Phase 1. The study was conducted from July to October 1995. To maximize response and to comply with laws of confidentiality in Hong Kong, subjects were informed of the study and their consent to participate was obtained by the Social Welfare Department.

Subjects who had agreed to be interviewed were then visited at home by trained interviewers consisting of psychiatrists, psychiatric nurses, and lay interviewers. Just before phase 1, these interviewers had undergone a 1-week intensive training that included practice sessions of live interview with elderly living in the community and demented subjects. The interview lasted for about 40 minutes and consisted of the Cantonese version of the Mini-Mental State Examination (CMMSE),17 the Cantonese version of Geriatric Depression Scale (CGDS),18 and a questionnaire eliciting information on demographic data,

physical illness, the Chinese version of Lubben's Social Support Scale,19 Barthel's Index 20 and Instrumental Activities of Daily Living,21 and a modified life event scale.22

Phase 2. Subjects scoring below the cutoff point on the CMMSE(see Screening instruments) and those with scores above the cutoff point of the CGDS were invited to attend a clinical interview conducted by experienced psychiatrists in a psychogeriatric outpatient clinic in Shatin. Subjects who could not complete the screening instruments were also interviewed. Those who declined to go to the clinic were examined at home by a psychiatrist. In addition, a random sample of 6% of the elderly who had scores in the normal range on the CMMSE and CGDS were also interviewed. The psychiatrists were blind to the findings of the initial screening score. Relatives were also interviewed whenever possible. For subjects with suspected dementia, a physical examination was done. These diagnostic interviews were conducted within 3 months of the first interview. Where possible, appropriate investigations, including a full blood count; renal, liver, and thyroid function; VDRL; glucose; B12 and folate levels; chest radiograph; ECG; and brain CT, were performed for demented patients. Clinical diagnosis of dementia was made according to DSM-IV 23 criteria. The diagnosis of AD was based on DSM-IV criteria, whereas the diagnosis of VD was based on the criteria of CAMDEX.24 Thus, the diagnosis of VD was supported by any three of the following 24-relatively sudden onset, step-like deteriorating course, history suggestive of strokes or TIAs, or focal neurologic signs and symptoms or focal disturbances of higher cortical function such as aphasia or apraxia in the presence of relatively good general intellectual preservation-or any two of the following-patchiness of psychological deficits, emotional lability, preservation of insight, fluctuating depression and/or anxiety, epileptic fits, hypertension, frequent headache and/or dizziness, and unsteady broad-based gait.

Screening instruments. CMMSE was the screening instrument used in phase 1.17 This was modified and translated from the original version of the Mini-Mental State Examination(MMSE)25 and has been validated on the local population of elderly with satisfactory reliability and validity.17 Several items of the original MMSE were modified for use in our locality. For the writing test, the strategy used in the Chinese version of the MMSE developed by Yu et al.26 was adopted. Thus, the subject was asked to "Say a sentence" instead of "Write a sentence" to avoid failure due to lack of education. The serial sevens test was retained, but there was no equivalent to the item of spelling "W-O-R-L-D" backward. As an alternative, reversal of five digits was used.

In the original validation study of the CMMSE, a cutoff point of 19 or below was adopted.17 However, scores on the MMSE were found to be affected by educational background in other studies on Chinese elderly, and different cutoff points according to educational levels were used in studies in Shanghai and Beijing.8,9 For this project, we conducted a pilot study on 298 Chinese elderly (134 demented subjects and 164 normal elderly) in our locality to examine the effect of education on the cutoff points. In general, Chinese elderly in Hong Kong have a low educational level. Based on the results of this pilot study, the cutoff points for detecting cognitive impairment were established for three groups. The optimal cutoff points were 18 or below for illiterate subjects, 20 or below for those with 1 to 2 years of schooling, and 22 or below for those with more than 2 years of schooling. These scores were adopted for use in the present study.

The Chinese version and Cantonese version of MMSE have been used in surveys in Chinese populations with good sensitivity and specificity in detecting cognitive impairment.17,26 In particular, the psychometric properties of the Chinese version of MMSE have been widely studied, and it was reported that modifications in the Chinese MMSE did not compromise its comparability with the original MMSE.26 In another cross-cultural study of dementia, the mean Chinese MMSE scores of Chinese elderly are similar to the MMSE scores of Finnish elderly after taking into consideration the effect of education.27

The CGDS 18 is a 15-item self-rating scale that assesses the subject's current depressive symptoms. It was specifically developed for use in the elderly population and was designed in a yes-no format.28 The scale's brevity and simple response format are particularly favorable for use among the elderly. The CGDS was modified and translated to Cantonese from the original version of the Geriatric Depression Scale 28 and has been validated on the local population of elderly with satisfactory reliability and validity. A cutoff point of eight or above was recommended for use in detecting depression in Hong Kong.

Statistical analysis. The age-specific prevalence rates of dementia subtypes adjusted for sampling design were estimated from the clinical evaluation of those interviewed subjects. The gender-specific prevalence and overall prevalence were computed from the age-specific prevalence rates and were weighted with respect to the population distribution. Sensitivity and specificity rates were used to evaluate efficiency of the screening instruments. Intraclass correlation and kappa were used to measure interrater reliability. Pearson's [chi]2 was used to study the effects of the demographic variables on the prevalence rates. When confounding variables were considered in analyzing the effects of demographic variables, Mantel-Haenszel chi-squares (MH) were used. When cell frequencies were too small in some of the categories, the exact method 29 of the above tests was used to evaluate the effects. Results were considered significant if p < 0.05. Using the age-specific rates, the prevalence for age groups not covered in this study were estimated by a logistic model described by Ritchie and Kildea.30

Results. Sample characteristics and response rate. The effective target sample was 1,503 after excluding subjects who had moved or had wrong or incomplete address, 31 subjects who were already dead, and 21 subjects who had been hospitalized. Of these, 1,034 (68.8%) were successfully interviewed. Among the rest, 428 (28.5%) refused to be interviewed, 29 (1.9%) could not be contacted, and 12 (0.8%) were too deaf or physically ill to be interviewed. In phase 1, the respondents did not differ significantly from the nonrespondents by sex ([chi]2 = 0.00092; df = 1; p = 0.98) or age ([chi]2 = 2.114;df = 4; p = 0.715). Among the interviewed subjects, 176(17.0%) were living in old age homes or care and attention homes (similar to nursing homes), whereas 858 (83.0%) were living at home in the community. Some 12.6% of the elderly were living alone. The mean educational level of the subjects was 3.3 ± 3.8 years, with 36% of the sample being illiterate. The table gives the demographic structure of the interviewed sample.

Table Demographic characteristics of the subjects in phase I and phase II

Of the 1,034 subjects interviewed in phase 1, 307 (29.7%, including 18 who did not complete the CMMSE) were identified as probably demented or probably depressed by their scores on the CMMSE or CGDS and were selected for the phase 2 interview. Forty-four subjects were randomly selected out of the remaining subjects who had scores in the normal range on both CMMSE and CGDS for further examination, giving a total of 351 subjects for interview in phase 2. Of these elderly, 332 (94.6%) were successfully interviewed, 14 (4%) refused, 2 (0.6%) were dead, 1 (0.3%) was too physically ill to be interviewed, and 2 (0.6%) could not be contacted. For the 99 subjects found to be demented, blood investigations were done for about 36% and CT of the brain was performed for 15% only. Thus, only 15% of subjects with dementia received complete medical evaluations including CT of the brain.

Evaluation of the screening instruments and interrater reliability. Sensitivity and specificity rates were used to evaluate the efficiency of the screening instruments. Clinical diagnosis by psychiatrists was used as the gold standard. Among the 332 subjects interviewed in phase 2, CMMSE had very good sensitivity (94.9%) and specificity (85.6%) rates in detecting dementia. Interrater reliability of the screening instruments was measured by the intraclass correlation. The correlations were excellent, CMMSE being 0.9 and CGDS, 0.91. The interrater agreement of clinical diagnosis for the five psychiatrists involved in the study was measured by kappa. Thirty subjects were selected for this part of the reliability study. The kappa for AD was 0.93 and for VD was 0.87; the overall kappa for dementia was 0.94.

Prevalence of dementia. Figure 1 shows the weighted prevalence rates of dementia by age groups and by sex. The overall weighted prevalence of dementia in those aged 70 years and older was 6.1± 0.7%. The overall age-specific rates of dementia were 1.7 ± 0.8% for the group aged 70 to 74, 4.1 ± 1.2% for the group aged 75 to 79, 10.7 ± 2.2% for the group aged 80 to 84, 18.8 ± 3.3% for the group aged 85 to 89, and 25.8 ± 3.8% for the group aged 90+. The rate of dementia increased with age (MH = 66.83; df = 1;p < 0.0001) and approximately doubled for every 5 years until around 90 years. Even when sex and educational level were controlled, the age effect was still significant (MH = 49.21; df = 1;p < 0.0001). Females had a higher rate of dementia (7.1%) compared with males (4.7%) (OR = 1.93; [chi]2 = 9.09;df = 1; p = 0.003), even when age was controlled(MH = 4.29; df = 1; p < 0.04). However, when educational level was controlled, the sex effect became insignificant(MH = 0.35; df = 1; p = 0.55). Educational level had a significant effect on rates of dementia. The prevalence of dementia increased for those with low educational level (MH = 12.77;df = 1; p < 0.001). Even when sex or age were controlled, this effect remained significant (MH >= 4.39;df = 1; p < 0.04).

Figure 1. Prevalence of dementia by age and sex.

About 45% (n = 45) of the 99 demented subjects were living in institutions. If subjects living in institutions were excluded, the prevalence of dementia among those living at home in the community was 4.5 ± 0.7%. The prevalence of those living in institutions was four times that of subjects living at home, being 17.4 ± 2.8% (OR = 5.11; [chi]2 = 62.67;df = 1; p < 0.0001).

Based on our age-specific rates of dementia, the logistic model described by Ritchie and Kildea 30 was fitted to predict the overall prevalence of dementia if our sample had been aged 65 years and older. Using the predicted rates from this logistic model, the weighted overall prevalence of dementia estimated for those aged 65 and older was 4%.

Prevalence of dementia subtypes. Clinically diagnosed AD accounted for 64.6% of cases of dementia. VD accounted for 29.3%, whereas 6.1% were due to other causes: two with Parkinson's disease, one with alcohol-induced persisting dementia, one with meningioma, and two with uncertain etiology. The overall weighted prevalence rates in those aged 70 years and older of AD were 3.6 ± 0.5% and of VD, 2.2 ± 0.5%.Figure 2 shows the prevalence rates of AD by age and sex and figure 3, that of VD. Prevalence of AD increased with age (MH = 62.60; df = 1; p < 0.0001). Even when sex and education were controlled, age was still a significant factor (MH = 44.11; df = 1; p < 0.0001). Females had a higher prevalence rate of AD (OR = 2.79; [chi]2 = 13.43; df = 1; p < 0.001) even when age was controlled (MH = 7.56;df = 1; p < 0.005). However, when education was controlled, sex became nonsignificant (MH= 1.16; df = 1; p = 0.283). Educational level had a significant effect on rates of AD (MH = 15.53; df = 1; p < 0.0001), which remained significant even after controlling for sex or age (MH > 6.27; df = 1;p < 0.02). In VD, although there was a trend of increasing prevalence with age (MH = 4.87; df = 1; p < 0.03), the increase was not so marked as in AD. Sex had no significant effect on rates of VD independently ([chi]2 = 0.003;df = 1; p = 0.96). Similarly, educational level had no significant effect on rates of VD (MH = 0.90; df = 1; p = 0.34). Because of the small number of cases of VD, only simple association was conducted.

Figure 2. Prevalence of Alzheimer's disease by age and sex.

Figure 3. Prevalence of vascular dementia by age and sex.

Discussion. There are a few methodologic limitations to this study. First, the sample size is relatively small. A larger sample would have permitted a more accurate estimate of prevalence. Nevertheless, a sample size of 1,000 elderly is still comparable with many previous epidemiologic studies on the prevalence of dementia. Second, a response rate of 68.8%, although similar to the response rates in several recent studies on prevalence of dementia in Europe,31 North America,32 and Australia,33 is less than ideal. A disproportionate number of those who refused to be interviewed may be cognitively impaired and thus methodologic problems may have contributed to our result of a low rate of dementia. Unfortunately, because of laws of confidentiality in Hong Kong, we do not have further information on the subjects who refused interview and the reasons for refusal nor is it possible to ask the family members about the functional status of these subjects. However, similarly high rates of refusal are also reported in some recent epidemiologic studies on dementia in Australia 33 and Canada.34 Thus, this may be a realistic rate expected from surveys of elderly in big cities. In addition, other Chinese studies with a low refusal rate 7,10,15 also showed a low prevalence of dementia, lending support to our finding of a low prevalence of dementia. Third, a structured clinical interview was not used in phase 2 for clinical diagnosis because of the lack of any satisfactory instrument for use in Chinese elderly. Fourth, it is still uncertain how best to diagnose VD, and a major difficulty in the study of VD concerns disease definition. The development of several sets of diagnostic criteria of VD, including the DSM-IV, ICD-10, the criteria of the State of California Alzheimer's Disease Diagnostic and Treatment Centres, and the criteria of NINCDS-AIREN, has not really solved this problem. Indeed, these criteria are all found to be deficient in various ways.35 The use of the Hachinski Ischemic Score has also been heavily criticized.36 In our study, the CAMDEX criteria for VD was chosen because it provides a clear and pragmatic guideline on diagnosis. Neuro-imaging may have improved the diagnostic accuracy, but brain CT was only obtained in less than one-fifth of our demented subjects. This problem is common in epidemiologic studies because many elderly are frail and may decline such investigations. Fifth, because the diagnoses of AD and VD were not based on brain biopsy, the clinical diagnoses of these cases would be more correctly labeled as probable AD and probable VD. Some of these cases might be found to be wrongly classified on postmortem studies.

There are several merits to our study. First, our sample includes elderly living in the community and in institutions. Populations differ in the percentage of demented elderly living in the community depending on the likelihood of these elderly entering institutions. Thus, a study involving both groups of elderly would yield a clearer picture of the true prevalence of dementia in that population. Second, the relatively short time lag between the two phases has minimized attrition rates due to deaths or loss of contact with subjects and thus has contributed to a high response rate of the second phase.

The main findings of this study are that prevalence rates of dementia in our sample are similar to those reported in whites, with AD being more common. Nevertheless, our rates of AD and VD are on the low

side of rates in European and American studies. The EURODEM studies found that rates of dementia in European studies ranged from 1.0% at ages 60 to 64 to 32.2% at ages 90 to 94, whereas those for AD ranged from 0.3% at ages 60 to 69 to 10.8% at ages 80 to 89.37,38 The rates of VD ranged from 2.2% in those aged 70 to 79 to 9.2% in those aged 80 to 89.39 Studies in the United States and Canada reported higher rates of dementia. For instance, Evans et al.40 found that rates of dementia ranged from 3.0% at ages 65 to 74 to 47.2% at age 85 and older. In Canada, the prevalence for those aged 65 and older was 8.0%, ranging from 2.4% among those aged 65 to 74 to 34.5% in those aged 85 and older.34 The corresponding figures for AD in Canada was 5.1% overall, ranging from 1.0% to 26.0%; for VD it was 1.5% overall, ranging from 0.6% to 4.8%.34

Several reviews found regional differences in the prevalence of dementia.2,3 In particular, Jorm 2 concluded that AD was more common in Europe and in countries with populations largely derived from European emigration, whereas VD was more frequent in Japan and China. However, it is still unclear whether the differences are due to methodologic issues like diagnostic criteria or whether there are true differences across regions.

Emerging evidence from recent studies show substantial differences in the prevalence of dementia between ethnic groups, and interactions between genetic and environmental factors are pertinent. For Africans, two separate studies by Schoenberg et al.41 and Heyman et al.42 showed that blacks in America had higher rates of dementia compared with whites. In contrast, Osuntokun et al.43 and Ogunniyi et al.44 found that AD was rare among Africans. In an important study, Hendrie et al.45 compared the prevalence of dementia in sub-Saharan Africans with African-Americans using the same methodogy and diagnostic criteria. Sub-Saharan Africans had low rates of dementia, in particular AD, whereas African-Americans had rates in the high range of that reported in whites, with an overall prevalence of 8%. The authors suggested that this might be due to an interaction between genetic and environmental factors. For instance, the frequency of apolipoprotein E (apoE) [epsilon]4 allele was high(30%) in Nigerians, yet the rates of AD and atherosclerosis were low in these people. This was thought to be related to the influence of environmental factors such as diet.45

In Chinese, there is a wide range in the rates of dementia, although most studies showed lower prevalence of dementia and AD compared with rates in Western countries. Recently, apoE [epsilon]4 is shown to be a major susceptibility factor for AD. This association has also been replicated in Chinese in Hong Kong and Taiwan.46,47 Because the frequency of apoE [epsilon]4 allele in the Chinese population is lower than that in the white population,46-48 this may be one factor to account for the lower rate of AD observed in Chinese. Nevertheless, the variability in the rates of AD may be due to methodologic differences across studies and/or real differences related to interactions between genetic and environmental factors in various regions.

A recent study on pathologic changes in the brains of elderly Chinese in Hong Kong has provided further insights into this problem.49 It was found that whereas the pathologic features of AD were similar to those observed in the West, the prevalence of AD-type changes (i.e., neuritic plaques and

neurofibrillary tangles) in normal elderly seemed to be much lower than that in the Western population. This is an intriguing finding that awaits further study.

It is noteworthy that studies in China and Hong Kong showed that intracerebral hemorrhages accounted for one-third to one-half of all cases of stroke in Chinese, which is much higher than the frequency in North America stroke registries.50,51 Further, lacunar infarcts were also more common in Hong Kong Chinese compared with whites.51 These clinical findings, which suggest a selective involvement of small vessels in cerebrovascular disease in Chinese, are supported by a postmortem study that showed a different pattern of cerebral atherosclerosis in Hong Kong Chinese compared with that in white populations.51 The extent of intracranial atherosclerosis was found to be much more severe in Hong Kong Chinese, with marked involvement of medium and small vessels, whereas atherosclerotic narrowing of extracranial carotid artery was less prominent.52 We speculate that these factors might lead to a different pattern or frequency of VD in Chinese compared with whites.

There are marked differences in stroke incidence among cities in China, with north China having higher rates compared with the south.53 This variability may result from differences in diet, alcohol and cigarette consumption, or prevalence and treatment of hypertension.53,54 Thus, this possibly may account for the variation in the ratio of AD to VD reported in different regions of China, like Shanghai and Beijing,10 and in places like Taiwan and Hong Kong whose population has mostly originated from the southern part of China. Further cross-national studies, especially on incidence and neuropathologic diagnosis, are necessary to examine substantial differences in the prevalence and types of dementia in Chinese compared with other ethnic groups. In particular, a greater knowledge of the exact distribution and risk factors for VD has important health implications because some factors might be preventable or modifiable.

The low prevalence of secondary dementia may be related to the relatively old age of our sample because potentially reversible causes of dementia are more frequent in younger age groups.55 It is also noteworthy that there is only one case of alcoholic-induced persisting dementia in our sample. This is compatible with the very low rates of alcohol dependence in Chinese as reported by many previous studies in Hong Kong, Taiwan, and Singapore.56-58

Another finding of our study is that the increase in rates of dementia with advancing age slows down after ages 85 to 89. Indeed, in our sample of subjects aged 90 and over, the prevalence of dementia was only 25.8%. This is consistent with the findings of a recent meta-analysis that showed that the prevalence of senile dementia increases exponentially with age but the rate of increase falls from ages 80 to 84 and flattens out to zero at age 95, when prevalence is 40%.30 The authors suggested that senile dementia may be age-related rather than ageing-related.30

Low educational level is reported to be a risk factor for AD in some studies,8,59 but its exact role is still controversial. Two recent studies 60,61 on the incidence of dementia failed to find an association of education with risk of AD, and it has been suggested that survival and selection bias might explain the apparent association of low educational level and AD in some prevalence studies.60 In our subjects, low educational level was associated with a higher rate of dementia.

In contrast to findings in Beijing,7 Shanghai,8 and Taiwan,14 where no subject with dementia was found to be living in institutions, 45% of our demented subjects were living in institutions. This figure is very similar to the pattern in Europe and North America 3,63 and reflects the growing westernization of Hong Kong, with a gradual loss of the traditional family support for the elderly. The problem of lack of family support for the elderly is exacerbated by a recent wave of emigration of many young people because of the change of sovereignty in 1997.

The results of our study highlight the importance of dementia as a significant health problem in Hong Kong. Further epidemiologic studies on the incidence, subtypes, and risk factors of dementia are necessary to elucidate the complex problem of dementia in Chinese.

Acknowledgments^We thank Mrs. Anna Mak, Mr. Wong Shun, and Miss Ann Hon of the Social Welfare Department for their help.

References^1. Cooper B. The epidemiology of the dementias of late life. In: Jacoby R, Oppenheimer C, eds. Psychiatry in the elderly. 2nd ed. Oxford: Oxford University Press, 1997:439-453. [Context Link]

2. Jorm AF. Cross-national comparisons of the occurrence of Alzheimer's and vascular dementias. Eur Arch Psych Clin Neurosci 1991;240:218-222. [Context Link]

3. Henderson AS. Dementia. Geneva: World Health Organization, 1994:9-26. [Context Link]

4. Kuang PG, Zhao ZD. A survey of senile disturbance in Wuhan City in 1981. Chin J Epidemiol 1984;5:95-99. [Context Link]

5. Zhao YZ. The epidemiological study of psychoses of different kinds, drug and alcohol dependence and personality disorders in 12 different regions of China. Chin J Neurol Psychiatry 1986;19:70-71. [Context Link]

6. Chen XH. An epidemiological investigation of mental disorders of the aged in an urban district of Beijing. Chin J Neurol Psychiatry 1987;20:145-150. [Context Link]

7. Li G, Shen YC, Chen CH, Zhao YW, Li SR, Lu M. An epidemiological survey of age-related dementia in an urban area of Beijing. Acta Psychiatr Scand 1989;79:557-563. [Medline Link] [PsycINFO Link] [Context Link]

8. Zhang MY, Katzman R, Salmon D, et al. The prevalence of dementia and Alzheimer's disease in Shanghai, China: impact of age, gender and education. Ann Neurol 1990;27:428-437. [Medline Link] [PsycINFO Link] [Context Link]

9. Li G, Shen YC, Chen CH, Zhao YW, Li SR, Lu M. A three-year follow-up study of age-related dementia in an urban area of Beijing. Acta Psychiatr Scand 1991;83:99-104. [Medline Link] [PsycINFO Link] [Context Link]

10. Shen YC, Chen CH, Li SR, Li G, Zhao YW, Zhang WX. Epidemiology of age-related dementia in China. HK J Psych 1994;4(suppl):17-19. [Context Link]

11. Zhang HR, Lin SX, Liu XP, et al. A study of the epidemiology of senile dementia in Hainam province. Proceedings of the Dementia International Symposium, Hainam province, 1994:7-11. [Context Link]

12. Yu SX, Zhang CF, Guo MX, et al. An epidemiological survey of dementia in the elderly in Sheaxi. Proceedings of the Dementia International Symposium, Hainam province, 1994:1-6. [Context Link]

13. Lin HN, Tsai MT, Rin H. Psychiatric disorders among old people in a rural area: Hung-Ch'un study. Bull Chin Soc Neurol Psychiatry 1984;1(special issue):64-79. [Context Link]

14. Liu HC, Chou P, Lin KN, et al. Assessing cognitive abilities and dementia in a predominantly illiterate population of older individuals in Kinmen. Psychol Med 1994;24:763-770. [Medline Link] [PsycINFO Link] [Context Link]

15. Liu HC, Lin KN, Teng EL, et al. Prevalence and subtypes of dementia in Taiwan: a community survey of 5297 individuals. J Am Geriatr Soc 1995;43:144-149. [Medline Link] [PsycINFO Link] [CINAHL Link] [BIOSIS Previews Link] [Context Link]

16. Kua EH. The prevalence of dementia in elderly Chinese. Acta Psychiatr Scand 1991;83:350-352. [Medline Link] [PsycINFO Link] [Context Link]

17. Chiu HFK, Lee HCB, Chung D, Kwong PK. Reliability and validity of the Cantonese version of the Mini-Mental State Examination: a preliminary study. HK J Psych 1994;4(suppl 2):25-28. [Context Link]

18. Lee HCB, Chiu HFK, Wing YK, Kwong PK, Leung CM, Chung DWS. The GDS short form as a screening test for the Chinese elderly. A preliminary study. Clin Gerontol 1993;14:37-41. [PsycINFO Link] [BIOSIS Previews Link] [Context Link]

19. Lubben JE. Assessing social network among elderly populations. Fam Comm Health 1988;11:42-52. [Context Link]

20. Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index: a reliability study. MD State Med J 1965;14:61-65. [Context Link]

21. Lawton MP, Brody EM. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist 1969;9:179-186. [Medline Link] [Context Link]

22. Paykel ES, Prusoff BA, Uhlenhuth EH. Scaling of life events. Arch Gen Psychiatry 1971;25:338-347. [Medline Link] [Context Link]

23. Diagnostic and statistical manual of mental disorders. 4th edition. Washington, DC: American Psychiatric Association, 1994. [Context Link]

24. Roth M, Tym E, Mountjoy CQ, et al. CAMDEX: a standardized instrument for the diagnosis of mental disorders in the elderly with special reference to the early detection of dementia. Br J Psych 1986;149:698-709. [Context Link]

25. Folstein MF, Folstein SE, McHugh PR. Mini-Mental State: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-198. [Medline Link] [PsycINFO Link] [Context Link]

26. Yu ESH, Liu WT, Levy P, et al. Cognitive impairment among elderly adults in Shanghai, China. J Gerontol 1989;44:S97-S106. [Medline Link] [PsycINFO Link] [Context Link]

27. Salmon DP, Riekkinen PJ, Katzman MD, Zhang MY, Jin H, Yu E. Cross-cultural studies of dementia. Arch Neurol 1989;46:769-772. [Medline Link] [PsycINFO Link] [Context Link]

28. Yeasavage JA, Brink TL. Development and validation of a geriatric depression scale. A preliminary report. J Psychiatr Res 1983;17:37-49. [Context Link]

29. Mehta C, Patel N. StatXact user manual, version 2.11. Cambridge, MA: CYTEL Software Corporation, 1992. [Context Link]

30. Ritchie K, Kildea D. Is senile dementia "age-related" or "ageing-related"?-evidence from meta-analysis of dementia prevalence in the oldest-old. Lancet 1995;346:931-934. [Medline Link] [BIOSIS Previews Link] [Context Link]

31. Dartigues JF, Gagnon M, Michel P, et al. Le programme de recherche Paquid sur l' epidemiolog de la demence: methodes et resultats initiaux. Rev Neurol 1991;147:225-230. [Medline Link] [Context Link]

32. Folstein MF, Bassett SS, Anthony JC, Romanovski AJ, Nestadt GR. Dementia: case ascertainment in a community survey. J Gerontol 1991;46:M132-M138. [PsycINFO Link] [Context Link]

33. Henderson AS, Jorm AF, Mackinnon A, et al. A survey of dementia in the Canberra population: experience with ICD-10 and DSM-III-R criteria. Psychol Med 1994;24:473-482. [Medline Link] [PsycINFO Link] [Context Link]

34. Canadian Study of Health and Ageing Working Group. Canadian study of health and ageing: study methods and prevalence of dementia. Can Med Assoc J 1994;150:899-913. [Fulltext Link] [BIOSIS Previews Link] [Context Link]

35. Bowler J, Hachinski V. History of the concept of vascular dementia: two opposing views on current definitions and criteria for vascular dementia. In: Prohovnik I, Wade J, Knezevics, Tatemichi T, Erkinjuntti T, eds. Vascular dementia. England: John Wiley & Sons, 1996:1-28. [Context Link]

36. Dening TR, Berrios GE. The Hachinski Ischemic Score: a re-evaluation. Int J Geriatr Psych 1992;7:585-590. [Context Link]

37. Hofman A, Rocca WA, Brayne C, et al. The prevalence of dementia in Europe: a collaborative study of 1980-1990 findings. Int J Epidemiol 1991;20:736-748. [Medline Link] [Context Link]

38. Rocca WA, Hofman A, Brayne C, et al. Frequency and distribution of Alzheimer's disease in Europe: a collaborative study of 1980-1990 prevalence findings. Ann Neurol 1991;30:381-390. [Medline Link] [PsycINFO Link] [Context Link]

39. Rocca WA, Hofman A, Brayne C, et al. The prevalence of vascular dementia in Europe: facts and fragments from 1980-1990 studies. Ann Neurol 1991;30:817-824. [Medline Link] [PsycINFO Link] [Context Link]

40. Evans DA, Funkenstein HH, Albert MS, et al. Prevalence of Alzheimer's disease in a community population of older persons: higher than previously reported. JAMA 1989;262:2551-2556. [Medline Link] [Context Link]

41. Schoenberg BS, Anderson DW, Haerer AF. Severe dementia: prevalence and clinical features in a biracial US population. Arch Neurol 1985;42:740-743. [Medline Link] [PsycINFO Link] [Context Link]

42. Heyman A, Fillenbaum G, Prosnitz B, Raiford K, Burchett B, Clark C. Estimated prevalence of dementia among elderly black and white community residents. Arch Neurol 1991;48:594-598. [Medline Link] [PsycINFO Link] [Context Link]

43. Osuntokun BO, Adeuja AOG, Schoenberg BS, et al. Neurological disorders in Nigerian Africans: a community-based study. Acta Neurol Scand 1987;75:13-21. [Medline Link] [Context Link]

44. Ogunniyi AO, Osuntokun BO, Lekwauwa UG, Falope ZF. Rarity of dementia (DSM IIIR) in an urban community in Nigeria. East Afr Med J 1992;69:10-14. [Context Link]

45. Hendrie HC, Osuntokun BO, Hall KS, et al. Prevalence of Alzheimer's disease and dementia in two communities: Nigerian Africans and African Americans. Am J Psych 1995;152:1485-1492. [Fulltext Link] [Medline Link] [PsycINFO Link] [BIOSIS Previews Link] [Context Link]

46. Mak YT, Chiu H, Woo J, et al. Apolipoprotein E genotype and Alzheimer's disease in Hong Kong elderly Chinese. Neurology 1996;46:146-149. [Fulltext Link] [Medline Link] [Context Link]

47. Hong CJ, Liu TY, Liu HC, et al. E4 allele of apolipoprotein E increases risk of Alzheimer's disease in a Chinese population. Neurology 1996;46:1749-1751. [Fulltext Link] [Medline Link] [PsycINFO Link] [BIOSIS Previews Link] [Context Link]

48. Hallman DM, Boerwinkle E, Saha N, et al. The apolipoprotein E polymorphism: a comparison of allele frequencies and effects in nine populations. Am J Hum Genet 1991;49:338-349. [Medline Link] [Context Link]

49. Ng HK, Cheng Y, Poon WS. Alzheimer-type of pathological changes in Chinese. J Neurol Sci 1997;145:97-103. [Medline Link] [BIOSIS Previews Link] [Context Link]

50. Shi FL, Hart RG, Sherman DG, Tegeler CH. Stroke in the People's Republic of China. Stroke 1989;20:1581-1585. [Medline Link] [Context Link]

51. Huang CY, Chan FL, Yu YL, Woo E, Chin D. Cerebrovascular disease in Hong Kong Chinese. Stroke 1990;21:230-235. [Medline Link] [Context Link]

52. Leung SY, Ng THK, Yuen ST, Lauder IJ, Ho FCS. Pattern of cerebral atherosclerosis in Hong Kong Chinese. Stroke 1993; 24:779-786. [Medline Link] [BIOSIS Previews Link] [Context Link]

53. Cheng XM, Ziegler DK, Lai YH, et al. Stroke in China, 1986 through 1990. Stroke 1995;26:1990-1994. [Fulltext Link] [Medline Link] [BIOSIS Previews Link] [Context Link]

54. He J, Klag MJ, Wu ZL, Whelton PK. Stroke in the People's Republic of China II. Meta-analysis of hypertension and risk of stroke. Hypertension 1995;26:2228-2232. [Medline Link] [BIOSIS Previews Link] [Context Link]

55. Smith JS, Kiloh LG. The investigation of dementia results in 200 consecutive admissions. Lancet 1981;1:824-827. [Medline Link] [Context Link]

56. Chen CN, Wong J, Lee N, et al. The Shatin community mental health survey in Hong Kong. II. Major findings. Arch Gen Psychiatry 1993;50:125-133. [Medline Link] [PsycINFO Link] [BIOSIS Previews Link] [Context Link]

57. Hwu HG, Yeh EK, Chang LY. Prevalence of psychiatric disorders in Taiwan defined by the Chinese Diagnostic Interview Schedule. Acta Psychiatr Scand 1989;79:136-147. [Medline Link] [PsycINFO Link] [Context Link]

58. Kua EH. Drinking habits of elderly Chinese. Br J Addict 1990; 85:571-573. [Medline Link] [PsycINFO Link] [Context Link]

59. Katzman R. Education and the prevalence of dementia and Alzheimer's disease. Neurology 1993;43:13-20. [Medline Link] [PsycINFO Link] [BIOSIS Previews Link] [Context Link]

60. Beard CM, Kokmen E, Offord KP, Kurland TL. Lack of association between Alzheimer's disease and education, occupation, marital status, or living arrangement. Neurology 1992; 42:2063-2068. [Medline Link] [PsycINFO Link] [BIOSIS Previews Link] [Context Link]

61. Cobb JL, Wolf PA, Au R, et al. The effect of education on the incidence of dementia and Alzheimer's disease in the Framingham study. Neurology 1995;45:1707-1712. [Fulltext Link] [Medline Link] [PsycINFO Link] [BIOSIS Previews Link] [Context Link]

62. van Dujin CM. Epidemiology of the dementias: recent developments and new approaches. Neuroepidemiology 1996;60:478-488.

63. O'Connor DW, Pollitt PA, Hyde JB, et al. The prevalence of dementia as measured by the Cambridge Mental Disorders of the Elderly Examination. Acta Psychiatr Scand 1989;79:190-198. [Medline Link] [Context Link]

Accession Number: 00006114-199804000-00034

Copyright (c) 2000-2001 Ovid Technologies, Inc.Version: Version rel4.4.0, SourceID 1.5516.1.131