Abstract
Macau is a small territory, with a population of approximately 450 000 people coming mainly from rural areas of the Guangdong Province, Mainland China.[1] The elderly population has increased over the years and 7% of the total population are over 65 years of age; more than one-third of these are over 75 years. The roles of the elderly are dictated by tradition and strong cultural beliefs. They are confronted by Macau's fast-changing society, both on socioeconomic and political levels. After more than 400 years of Portuguese administration, Macau returned in 1999 to the People's Republic of China. This major political change may impact on these senior citizens. Environmental and biological adversity, as well as psychosocial factors, have been shown to be associated with emotional distress, which, in later life, frequently manifests as depression [2].
Community studies in Western countries showed that the prevalence of depression (severe enough to warrant treatment) in people aged 65 and over is about 15% [3], and in primary health care attenders the rates are approximately double those found in a community sample [4, 5]. Blazer [6] reported a prevalence of depression among those over 65 years old in the community ranging from 2% to 5% for major depressive disorders to as high as 50% for depressive symptoms. With regard to the Chinese elderly, previous community-based studies conducted in Hong Kong revealed a prevalence of depression of around 40% [7]. As frequent users of health-care services, elderly patients offer general practitioners (GPs) multiple opportunities to recognize their depressive symptoms. Therefore, the primary health-care setting is ideal for detecting depression in old age. Nevertheless, several studies have identified the lack of detection of depression by GPs [8, 9]. Up to 50% of depressed people seen in primary care are not recognized by the clinician as having this disorder [9]. Regarding the elderly, ‘because of the many physical illnesses and social and economic problems of the elderly, individual health-care providers often conclude that depression is a normal consequence of these problems, an attitude often shared by the patients themselves. All these factors conspire to make the illness underdiagnosed and undertreated’ [3].
It is commonly accepted that Chinese seek help from their social network for problems of a purely psychological nature. In contrast they are likely to perceive physical symptoms as needing medical attention [10]. On the other hand, in traditional Chinese medical thinking [11]
Each of the five major emotions has a corresponding internal organ: happiness is in the heart; anger is in the liver; worry is in the lung; fear is in the kidney and desire is in the spleen. The imbalance of emotions disturbs the functional balance of those organs and vice versa. Excess, incongruence and lack of harmony of emotions are regarded as pathogenic and therefore, high value is placed upon moderation and inhibition of emotions or affective expression.
All those factors should influence the expression of emotional feelings and therefore make the diagnosis of depression difficult.
According to Goldberg and Bridges [12], ‘psychologization’ is to be expected in egocentric cultures, where there is a narcissistic idealization of the self. Somatization is, therefore, a very common form of symptom manifestation in Chinese patients with mental disorders, and is regarded as playing an important role in influencing the underdiagnosis of depression [13].
Morbidity data collected by the Macau primary healthcare department demonstrated that depression had a prevalence of 0.06% [14]. Is depression then, a rare disorder or are the doctors and their patients sharing and practising a restrictive biomedical health model, giving no place for psychosocial complaints?
The simple recognition of illness and the prescription of even a minimal course of antidepressive drugs has been shown to be beneficial compared with the outcomes of unrecognized depression in patients [15]. The total population of Macau is around 450 000 of which 7% are elderly [16]; that means that Macau has a population of around 30 000 senior citizens. The number of senior citizens registered with Primary Health Care Department (PHCD) clinics in 1996 was 17 000, which represents 56% of the Macau elderly. The remainder attend private practitioners. In Macau free health care is provided to the elderly (patients aged 65 and over) by the Government PHCD. Patients attending the eight PHCD clinics are usually from the lower socioeconomic strata of the population. A study to determine depressive symptomatology of elderly Chinese patients (aged < 65), attending the Government Community Health Centre (HC) was set up in order to estimate the prevalence of depression in this population. The objectives of this study were: to identify the prevalence of depressive symptomatology in an elderly population attending HCs, using the Hospital Anxiety and Depression scale (HAD) screening scale; to assess the recognition of depression by the GPs by comparing qualitative GPs’ clinical opinions and the HAD scale results; to assess the reasons for consultation as given by the elderly and compare with the screening scale's results; and to relate depression to sociodemographic data (sex, age, marital status, social class, economic and social support).
Method
A cross-sectional study was carried out on a systematic sample of elderly Chinese patients currently resident in Macau, who attended an HC for consultation. Data were collected over 2 weeks in December 1997.
Population and samples
Four health centres were chosen by random selection, from a total of eight, which cover the entire peninsula of Macau. Every fourth elderly (65 years of age or over) Chinese patient registered in an HC who came for medical consultation was selected to participate in the study. All cases of non-response and non-participation were registered and analysed to control possible bias. Applying the principle of proportionality, the number of people interviewed from each of the four HCs was calculated according to the population of elderly registered at each HC. The sample size was calculated using the following formula n = pq/(E/1.96) [1], where n was the minimum sample size required; p was the maximum expected prevalence rate (%); q = 100p and E was the margin of sampling error tolerated (%). Data from previous studies have shown a wide range of prevalence of depression, from less than 10% to more than 40% among the elderly population in primary care settings. For this study, the prevalence of depressive disorders was estimated to be around 10%. The margin of sampling error tolerated was 5%. Using this formula a minimum required sample size of 352 was calculated. Because of the considerable variation of the prevalence values observed in the literature, 10% of the minimum sample size was added, in order to diminish any possible sampling error. This made 338 as the final sample size. This sample was a representative sample of the elderly attending the HC, and therefore, the results of this study could be generalized to the elderly population attending HCs in Macau.
A pilot study was performed initially to test the performance of all participants, focusing on the role played by the interviewers (interobserver reliability and the test–retest reliability). Details are available from authors on request.
All clients of the HC that fulfilled the following criteria were considered for entry into the study: client of the HC, age < 65 years, currently resident in Macau, Chinese (both parents are of Chinese origin) and without evidence of severe cognitive impairment detected by the Short Portable Mental Status Questionnaire (SPMSQ) [17]. All individuals that did not fulfil the inclusion criteria were excluded.
Data collection instruments, concepts and variables
Participants completed a demographic questionnaire to identify individual characteristics such as sex, age, marital status, social class. In addition, a questionnaire with a scale for detecting depressive status was used, the Hospital Anxiety and Depression scale (HAD). This scale has been a screening test used in many research studies at the primary health-care level [18–21]. Developed by Zigmond and Snaith [22], it has been validated on Chinese elderly population by Lam et al. [18]. Lam's validation was done by Clinical Interview Schedule (CIS). Relative operating characteristic analyses revealed that the optimal cutoff points of the HAD scale for depression were a score of 6 or more, and for anxiety a score of 3 or more [18]. Depression has been defined as having a HAD subscale score 6 or higher. For purposes of comparison, a different cut-off point [22], was also considered. A score of 7 or less was considered a ‘non-case’; between 8 and 10, ‘doubtful cases’; and scores of 11 or more, ‘definitive cases’ There are seven out of 14 items specifically designed to measure depression measured, on a scale of 0–3 for each item.
The SPMSQ [17] was used to detect severe cognitive impairment, a condition for which patients were excluded from this study.
A checklist of 28 complaints/symptoms to indicate their recent (last month) complaints was presented to the elderly as well as a question to evaluate the reason for the consultation. The elderly chose from the list of symptoms any of those appropriate to their situation.
Clinical records (charts) were analysed in order to collect the following clinical data: (i) the morbidity that was registered by all the GPs according to the International Classification of Health Problems in Primary Care (ICHPPC-2-defined) [23] relating to that day of consultation; and (ii) recent relevant treatment (e.g. antidepressive drugs) provided by the general practitioner or referral to a specialist (psychiatrist or psychologist). These management approaches were considered as recognition of a psychological or psychiatric disorder.
Indication of economic and financial support was determined by asking the patients whether others significantly helped the respondent to pay for their housing expenses, food or clothing. Support could be either from family/friends or institutional.
Social support was gauged by asking patients who they lived with and whether anyone was willing to listen to them talk about their problems, difficulties and worries.
According to the reason for consultation and HAD score, the patients were classified as: psychological presenters, patients that have psychological morbidity (positive HAD test) and present only psychological symptoms/complaints; somatic presenters, patients that probably have psychological morbidity (positive HAD test), but present only physical symptoms/complaints; mixed presenters, cases that do not fulfil the criteria for inclusion as either psychological presenters or somatic presenters; non-cases, patients who have only physical symptoms/complaints and are physically ill.
Social class classification (Register-General's Social Class [24]) was made according to the present profession or the profession they had before retirement. Those with home duties, mostly women, were classified according to the present profession of their spouse. In the event that the spouse was deceased, they were classified by the spouse's profession prior to death.
Standard procedures for data collection
Two administrative clerks from each HC were trained to select patients. An explanatory statement and a written informed consent statement was provided by the nurse to the participants and witnesses if present. The HAD questionnaire was administrated by trained nurses. One nurse, at each of the four clinics, was trained to administer the HAD questionnaire. This training was done using a prerecorded tape of a prototype interview, giving all prospective interviewers an example of the manner in which to correctly perform the interview thereby maintaining acceptable uniformity and reliability of procedures. Because of the very high rate of illiteracy among elderly, each question and its possible answers were read to the patient.
Data analysis
Data were coded and entered into the dBaseIIIplus program [25]. For statistical analyses SPSS version 7.0 [26] was used. Epi Info Version 6.04 [27] was also used for odds ratio (OR) analyses. To measure the stability and consistency of the HAD questionnaire the Split Half Correlation Test and Cronbach's alpha were calculated. For numerical variables, mean, mode, median and standard deviation were calculated. Chi-squared tests were performed in order to correlate the different variables with depression, using a 95% confidence interval. Odds ratios were performed to estimate the risk of depression considering the independent variables sex, age, social class and social support. Multiple logistic regression calculations were performed to correlate depression (dependent variable) with some of the most significant independent variables (sex, age, social class and social support).
Results
General practitioner characteristics
The interviewed elderly came from a pool of 31 GPs who had the following characteristics: seven (22.6%) were GPs with vocational training; 11 (35.5%) were GPs without vocational training but with more than 5 years of clinical experience; and 13 (41.9%) were GPs with less than 5 years of clinical experience.
Patient characteristics
For patient characteristics, see Table 1. Four hundred patients aged 65 or over were invited to participate in the study. Twelve patients refused to be interviewed (3%), all giving shortage of time as the reason for refusal. Of the remaining 388, two (0.5%) were excluded because of cognitive impairment, recording a score of 9–10 detected by the SPMSQ. The total number of participants was therefore 386, that is 97% of the initial sample.
Patients' characteristics
The majority of elderly patients showed considerable interest and some of them reported that ‘it was the first time that someone was interested in them’, particularly concerning those ‘sensitive matters’. Some reacted to the interview by crying. There were no reports of other emotional reactions or physical complaints. The average time spent conducting each interview was 15 min.
The ‘old’ (i.e. 75 years and over) comprised 36.8% of the study participants. The great majority of the participants were from social class IV and V. The lower social classes accounted for 84.9% of the participants (n = 328). The 15 commonest ‘diagnoses’ were, with one exception (insomnia), all physical in nature despite six of the 15 presenting complaints with a psychological basis (Table 2). Nine clinical charts mentioned ‘insomnia’ (2.3%), only three mentioned ‘anxiety’ (0.7%), and only one mentioned ‘problems within the family’ (0.2%).
Checklist of 15 most common symptoms/complaints and medical diagnosis/health problems (ICHPPC −2D) of all the participants
The minority of people were single (1.9%) or divorced (1.0%). These figures reflect the importance of marriage in this predominantly Chinese Christian community.
The clinical diagnosis of depression was not made in any cases nor were referrals to specialists (psychiatrists/psychologists) mentioned in the charts. The upper level of the depression subscale is 21 so a score of < 11 is certainly high. A total of 101 patients or 26.2% scored < 11.
Testing the reliability of the questionnaire, using split-half test, the difference between the divided groups was not significant (0.57). Cronbach's alpha was also performed for the evaluation of the internal consistency of the questionnaire, showing a reliable internal consistency (0.56).
The final score for each patient on the HAD was therefore obtained by the sum of all the partial scores of HAD scale questionnaire. Depressive status was considered by using the more stringent cut-off point of Zigmond and Snaith [22]. A total of 48% were considered cases, although only 26.2% were classified as definitely depressed (Table 3).
HAD subscale score for depression
According to their ‘reason for consultation’ as well as the score they obtained by the HAD scale, patients were further classified as psychological, somatic or mixed presenters, and non-cases. The majority were considered as somatic presenters (54.7%) and 45% were considered non-cases.
Women patients were in the majority. This reflects the normally higher frequency of women attending for medical consultation in the HC. Depression was detected in 59.3% of women and 46.1% in men (χ2 = 6.029, p = 0.014). To be female was probably a risk factor for depression (OR = 1.70, 95% CI = 1.09–2.67). The association of female gender with depressive status was also confirmed through multiple logistic regression analyses (p = 0.010). Multiple logistic regression was performed using the variables sex, age, social class and social support (some one to talk to).
Dividing the studied population into two subgroups according to their age (65–74; 75 or older), it was shown that depression was more frequent in those 75 years or older (χ2 = 6.782, p = 0.009). It seems there is an elevated risk of depression in these ‘old’ patients (OR = 1.84, 95% CI = 1.13–3.00). Being from a low social class (IV and V) showed a higher frequency of depression, compared with other classes, although this frequency was only just statistically significant (χ2 = 3.851, p = 0.050).
All types of living arrangements, including living with family or friends, living with someone unrelated or living alone showed no association with depression. Being entirely economically dependent upon others was also not associated with depression.
Not having someone willing to listen to them talk about their problems, difficulties and worries was significantly related to depression (χ2 = 4.17, p = 0.041). There is a risk of depression in those elderly who have no one to talk to (OR = 1.58, 95% CI = 0.99–2.53). This tendency was confirmed through multiple logistic regression analyses (p = 0.037).
Except for constipation, depressed elderly expressed all other symptoms/complaints from the checklist more frequently than the non-depressed. Table 4 represents the difference in frequency of symptoms/ complaints expressed, between depressed and non-depressed elderly.
The elderly with a higher score for depression (HAD ≥ 6) generally had more symptoms/complaints than those with low scores (Table 4). Patients with depression also showed a high score for anxiety (χ2 = 20.68, p = 0.0001).
Symptoms/complaints. Frequency in patients with depression and without depression (n = 258)
Discussion
This study had a high participation rate with the structured nature and brevity of the interviews contributing positively to the outcomes of the interview. The nurses conducted the interview before or just after the doctor's consultation. The small time difference between the interview and the consultation probably did not influence significantly the psychological condition of the elderly compared with the emotional response to being interviewed; however, neither should have affected responses to the questionnaire which refer to events in the past few weeks. The interview being performed by a trained nurse and not by a doctor reduced the possible perception of being ‘disloyal’ to their own GP.
In Chinese culture, the doctor is still highly regarded and perceived disloyalty could be a significant issue. The nurse performing the interview contributed to a relaxed environment, facilitating the expression of emotional feelings and hence, disclosure of psychological problems. Also, the use of a checklist of symptoms/ complaints for the patients facilitated the expression of complaints.
By using a validated screening instrument for the detection of depression the measurement bias for this target population was reduced. The reliability of the questionnaire was tested in the present study by performing the split-half test and Cronbach's alpha correlation test. The results showed correlation that is acceptable for an attitude survey questionnaire [28].
The very low diagnostic coding for psychological problems found in the medical records contrasted with the high frequency of the psychological symptomatology of the studied patients and the results of the HAD scores. With a cut-off < 8, proposed by Zigmond and Snaith [22], which includes the ‘doubtful’ and the ‘definitive’ cases, the figure was 48%. Those results were higher than another recent community-based study done in Hong Kong [7], which has a very similar population to Macau. In the current study, the target population was elderly patients who attended either for acute situations or follow up for chronic disorders, and who therefore were probably more susceptible to depression than elderly from the community at large. The authors attribute this increase to the uncertainty and instability in the community due to the return of the colony to China.
Katona [4] found depression to be more common in women than men; more common in the very old, those with chronic disability, those who had experienced recent major life events, and those who were socially isolated or with lack of intimacy and elderly, divorced or widowed. In the current study, being divorced or widowed did not represent a risk factor for depression. Living alone and being economically dependent did not show a relationship with depression. However, being female, very old, from a low social class and without a confidant were circumstances that contributed to the high level of depressive symptomatology, and therefore to increasing the frequency of depression. Those conditions were significantly related to depression.
The importance of a confidant needs further reflection. A significant number of those living with family or friends, can remain alone during all day. The research data showed that, even when living with family or friends, 28.5% of the elderly reported that they did not have anyone to talk to about their problems. According to the Macau Social Welfare Department [29], the progressive increase of the percentage of employed women, traditionally the support of the elderly, could aggravate the problem of social isolation of the elderly, with all the associated negative consequences on their emotional wellbeing. To have a confidant can act as a buffer for depression even if other situations remain less favourable for the elderly.
The very high level of hypertension (50.5%), ischaemic heart disease (36.7%) and diabetes (24.0%) registered in the medical records reflects the fact that chronic illness can be associated with depressive symptomatology and depression is frequently found in patients with cardiovascular disease [30].
The high percentage of depressive symptomatology expressed by the elderly, when asked directly (using a checklist of symptoms), confirms previous authors’ opinions. It has been stated that missing the diagnosis of depression or any other psychological distress is because the doctor does not monitor psychological aspects of their patients’ lives [8]. The high to moderate percentage of psychological complaints such as poor memory; inability to sleep; feeling weak, anxious, sad, tense, angry or frightened; poor concentration; loss of interest and even the wish to die contrasts with the lack of psychological diseases diagnosed in the clinical charts, where, almost invariably, only physical diagnoses have been registered. This pattern was more evident among those considered depressed.
Kantona [4] in a review article on depression in old age in Western communities, formulated that old patients had greater initial insomnia, agitation and hypochondriasis but less depersonalization, suicidal intent and loss of libido. The same article, citing a study done by Fredman in a community sample, showed that sleep disturbance was present in 14.8% and thoughts of death in 10%. Depressed mood was present in only 5% and loss of libido in 1.4%. In the present study, difficulty sleeping (36.7%), feeling sad (33.4%) and thoughts of death (12.2%) had a higher frequency than in the general studies done in depressive elderly. Compared with another study done in an elderly Chinese population [31] symptoms/complaints such as sleep disturbance, headache, loss of appetite, palpitations, weakness, feeling frightened, epigastric pain, chest discomfort, constipation, and feeling anxious were generally more frequent in the current study. This difference can be related to the difference in target populations. The current study is of the elderly attending for medical consultations, who frequently have multiple pathology and correspondingly multiple complaints.
If we apply a more selective cut-off point in the use of the HAD scale, as has been proposed by Zigmond and Snaith [22], 26.2% of the participants were definitely considered depressed. Although a large number of participants have depressive symptoms, the great majority of these were considered to have a mild to moderate form of depression. According to Blazer, the frequency of depressive symptoms compatible with mild to moderate depression is considerably higher than the frequency of major depression [32]. Even if the high level of depressive symptomatology did not have a correspondingly high level of clinical depression, it is the obligation of the clinician to be more sensitive to the psychological problems of their elderly patients. General practitioners deal with a broad range of depressive illness that includes not only major depression, but also many less severe but chronic or intermittent types of depression that are frequently intertwined with chronic physical illness. Depression that does not meet psychiatric criteria for a major depressive episode may still have a considerable impact on the quality of patient's life.
Emotional and psychological distress were not expressed spontaneously by the patients nor were they elicited by the doctors. The physical expression of psychological distress as presented by the elderly in this study has a cultural and social dimension. As mentioned by Kleinmen, a ‘culturally shaped psychological process lead the Chinese to suppress distressing emotions’ [33]. Among Chinese people ‘pain, communicated as a symptom but experienced as a feeling, is an opportunity to reintegrate the sick person into the social support group and to reaffirm the norms of solidarity and social control’ [34]. For all these reasons ‘somatization flourishes’.
Although there was a high percentage of depressed elderly, none of them was classified as a psychological presenter, that is, none presented to the doctor with a psychological symptom/complaint as a reason for encounter. The sick role of a depressive Chinese person ‘is socially and culturally unsanctioned and therefore suppressed. Somatization is sanctioned and expressed and it carries both cultural cachet and social efficacy. Depressive affect is unacceptable because it means illness and the breakdown of social harmony’ [34]. Macau's Chinese elderly's personality was moulded in a politically turbulent period and the difficult time of World War II, when they were educated to resist adverse situations, both physical and psychological and to suppress emotional manifestations arising from such adverse conditions. The family orientation of the Chinese remains powerful in Macau society and, although the family is a strong buffer for psychological distress, it can also limit the intervention of health professionals. Because psychological problems are usually shared only among family and close friends, the health professional generally has only the opportunity to explore one side of the equation, namely the physical side. The Chinese reluctance to express or discuss one's feelings, especially with anyone outside of one's family, may also play a part in the inhibitions of the manifestation of a depressive condition [10]. Mental illness is still considered a social stigma. This shame of having a psychological or psychiatric illnesses affects not only the patient but also the patient's family and therefore there is also a tendency to suppress the discussion of these topics even within the family. According to Bridges et al. [28], somatization is associated with an unwillingness to discuss emotional distress with a doctor. The patient believing that his family doctor would only be interested in physical illnesses, would behave in a way which would inhibit the discussion of psychological symptoms during the consultation. The doctor ‘colludes’ with this belief by not asking for psychological symptoms.
Our elderly seem to ‘have not developed the language to label and report their emotional states’ [35]. They utilize defences or coping styles that minimize outward psychological manifestations and, finally, they believe their problem is a physical one. The way patients spontaneously present their problems in the study reflects the above mentioned cultural situation. Although the doctor is a professional, both doctors and patients share the same cultural background, and consequently the way the patient communicates (somatizing) is not recognized by the doctor as a way of seeking (psychological) help. Nevertheless, in this study, when asked directly, the elderly patients did express their emotional symptoms of sadness, anger, fear, tension, loss of interest and others. As mentioned by Cheung [36], ‘the low rate of spontaneous reporting on emotional features may thus be a reflection of the insensivity of the health professionals, especially those in general practice who may not be aware of concomitant psychological aspects’. The manner in which the patients express their needs and how they are subsequently interpreted by the doctor, shows that the ‘difference between the patient's concept of illness and the doctor's concept of disease hinders the consultative communication’. This leads to the possibility of ‘misinterpreting the meaning of somatic complaints’. This scenario occurs very frequently in our clinical practice and is demonstrated in this study where physical complaints (symptoms) were treated and not recognized as symptoms of depression by the doctors, postponing indefinitely the identification and resolution of the real problem.
The Cantonese translation of the HAD, a validated instrument, was used to screen for psychological problems in the elderly. The data show that GPs do not recognize the emotional distress detected by the HAD subscales. This finding is not unique to Macau GPs using the HAD questionnaire. The empirical results of this study indicate that the elderly were not used to being asked about their emotional wellbeing. Regardless of the instrument used, this study reveals the poverty of clinical questioning of the elderly about their emotional or mental health. General practitioners need to explore psychological issues of the elderly patients in this population.
Conclusion
Some of the elderly's characteristics, such as age over 75 years, being female, from a low social class and socially isolated, were found to be related to depression. The frequency of depression in elderly attending the HCs was high (26.2%) and probably serves to compound the discomfort of other coexisting illnesses. Even if the high level of depressive symptomatology did not have a correspondingly high level of clinical depression, it is the obligation of the GP to be more sensitive to the psychological problems of their elderly patients. Depression that does not meet psychiatric criteria for a major depressive episode may still have a considerable impact on the quality of patients’ lives in this study; medical consultations were generally concerned only with physical symptoms, and psychological disorders were not recognized.
Using a rating scale that has been shown to be feasible, practical and useful on a regular basis, one can ‘improve awareness and interview skills, helping focus the consultation towards psychological problems especially when the patient preferentially presents somatic symptoms’ [33]. General practitioners need to explore psychological issues of the elderly patients in this population regardless of whatever questionnaire is used. The detection of depression is of particular importance as suicide, which become increasingly frequent with increasing age [37], is almost invariably associated with recent primary care contact [38]. It is therefore important for GPs to implement a management plan for all levels of depressive illness, which should include a suicide risk prevention perspective. General practitioners/family physicians can play an important role in the early detection and management of psychological disorders. They can provide for the elderly a better quality of life, contributing positively to a longer and happier living condition.