In 2010, 16.2% of the world population consisted of people aged 65 or over, a figure that is expected to rise to 26.9% by 2050. 1 Increasing life expectancy highlights the importance of physical and mental health in old age. Reference Karel, Gatz and Smyer2 Previous studies have generated very inconsistent findings about the prevalence of mental illness among older adults, Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha and Bryson3,Reference Snowdon4 although most studies report decreased prevalence rates in advanced age. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha and Bryson3,Reference Scott, Von Korff, Alonso, Angermeyer, Bromet and Bruffaerts5,Reference Kessler, Birnbaum, Bromet, Hwang, Sampson and Shahly6 Studies have tended to focus on selective disorders such as dementia Reference Launer, Andersen, Dewey, Letenneur, Ott and Amaducci7 or depression, Reference Braam, Prince, Beekman, Delespaul, Dewey and Geerlings8–Reference Riedel-Heller, Busse and Angermeyer10 implying that the entire range of mental disorders has been insufficiently addressed. Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9 Previous studies using different study designs have found lifetime and current prevalence rates of mental disorders in elderly people ranging from 1 to 18%. Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9,Reference Reynolds, Pietrzak, El-Gabalawy, Mackenzie and Sareen11 Studies of bipolar disorder, anxiety disorders and alcohol disorders based on structured and standardised assessment instruments such as the Composite International Diagnostic Interview (CIDI) 12 are scarce. Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9 Currently prevalence estimates for depression – the only disorder that is examined consistently – are approximately 3%. Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9 A few studies report lifetime prevalence rates of substance-related (in particular alcohol-related) disorders in people 65 years and over ranging from 1 to 12%; for schizophrenia, schizotypal disorders and other psychotic disorders, the lifetime and current rate is estimated at 0.5–1.0%, respectively. Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9 Rates for anxiety disorders vary between 0.9 and 6.7%. Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9 Only one study used the CIDI to evaluate somatoform disorders in elderly people Reference Leiknes, Finset, Moum and Sandanger13 and found a current prevalence rate of 18.4% (participants were 66 and older from Norway's general population). Reference Leiknes, Finset, Moum and Sandanger13
There is debate over the source and the causes of the heterogeneity of these empirical results: some authors have argued that older people may have developed coping strategies over the course of their lives that enable them to manage their mental health better than younger people, Reference Yang14,Reference Villamil, Huppert and Melzer15 whereas others have attributed the heterogeneity of the findings to a lack of feasible and age-sensitive standardised and structured instruments for diagnosing mental disorders in elderly people. Reference Knäuper and Wittchen16–Reference Bryant18 Older adults with health problems may also deny symptoms when asked to complete lengthy assessments. Reference Knäuper and Wittchen16,Reference O'Connor and Parslow17 Additionally, important information on the planning of intervention-based approaches must consider severity, impairment, quality of life and coping mechanisms. Taken together, this information indicates an urgent need to administer diagnostic instruments that have been adapted to the needs of elderly people.
The aim of the study is to determine lifetime, 12-month and current prevalence estimates for a wide range of mental disorders for people aged 65–84 years based on DSM-IV 19 in different European and associated countries using a standardised and structured interview that was specifically adapted for elderly people.
Method
Participants
The MentDis_ICF65+ study is a cross-sectional multicentre survey Reference Andreas, Harter, Volkert, Hausberg, Sehner and Wegscheider20 and the protocol has been previously reported. Reference Andreas, Harter, Volkert, Hausberg, Sehner and Wegscheider20 The selection of different catchment areas and countries was balanced according to geographical and socioeconomic population distribution in Europe. Southern European regions of Ferrara (Italy) and Madrid (Spain) were selected as well as London and Canterbury (England) for northern Europe and Hamburg (Germany) for central Europe. The sample further consisted of European Union (EU)-associated regions including Jerusalem (Israel) and Geneva (Switzerland). A random sample of n = 3142 older men and women (65–84 years) living in selected catchment community areas of each participating country (at least 500 participants from each country) stratified by age and gender was drawn from the population registries in Hamburg and Ferrara and from postal addresses of market research units in Madrid, Geneva, London/Canterbury and Jerusalem. Inclusion criteria for participating in the study included the ability to provide informed consent, having residence in the predefined catchment area at the beginning of the study, and being at least 65 and less than 85 years old. Potential participants were excluded on the basis of moderate cognitive impairment as assessed by the Mini-Mental State Examination (MMSE; cut-off score >18) Reference Folstein, Folstein and McHugh21 or an insufficient level of corresponding language. A harmonised procedure in contacting each participant and conducting the survey was realised, including initial contact by phone and mail, standardised interviewer training, implementation of a standardised study protocol for all test centres, and using stringent, high-quality data-control procedures.
The response rate was defined as the total percentage of participants who completed interviews in the study compared with who were contacted with a written invitation letter. Reference Kviz22 In most of the study centres, a written invitation letter was followed by a phone call to ask potential participants if they were willing to take part in the study. As a result of ethical regulations in some countries, potential participants had to write back to indicate their interest in participating; phone calls were not acceptable. The response rates varied by country, age and gender. Responder analyses showed significant differences in the response rate between the centres (P<0.001) and age groups (P<0.001) but not between genders (P = 0.738). The age effect indicates that the response rate was significantly higher for younger participants than for older participants. The overall response rate of our study was 20%, which is comparable with that of previous studies with similar recruitment procedures. Reference Keeter, Kennedy, Dimock, Best and Craighill23 Furthermore, representativeness analysis showed that the differences were small between the catchment areas in our study compared with catchment areas of the overall population of the participating countries with regard to sociodemographic characteristics (such as work status, marital status and education) according to the effect sizes by Somers' d Reference Goodman and Kruskal24 (all d<0.01); however, these differences were significant because of the large size of the databases. Furthermore, the minor differences that were identified are not clinically relevant.
Measures
CIDI adaptation process, structure and training
Mental disorders were diagnosed with an adapted, age-sensitive version of the CIDI, the CIDI65+. Reference Wittchen, Strehle, Gerschler, Volkert, Dehoust and Sehner25 This fully structured lay interview generates diagnoses according to DSM-IV criteria. 19 The process of adapting the test to the unique conditions of elderly people included several facets, including adding words, alternative questions and detailed section introductions, breaking down long questions into less complicated questions, sensitising scales upfront and embedding a fuller spectrum of syndromes. The English paper and pencil version was translated into German, Spanish, Hebrew, Italian and French with a back-translation and then computerised. A pre-testing phase was conducted in Germany and the UK to evaluate feasibility and verify the acceptability of the CIDI65+ to respondents. To evaluate the usability and reliability of the CIDI65+, a pilot phase was conducted. Reference Wittchen, Strehle, Gerschler, Volkert, Dehoust and Sehner25 Overall, the results on the reliability of the CIDI65+ were good for most diagnoses (such as depression (k = 0.79) and anxiety disorders (k = 0.69)). Lower k-scores were found for less frequent disorders such as panic disorders (k = 0.37), similar to other diagnostic instruments that also report lower reliability scores Reference Wittchen26 or that have classification systems that are less precise. Reference Wittchen, Strehle, Gerschler, Volkert, Dehoust and Sehner25
The interview covers a wide range of mental health problems such as anxiety disorders, affective disorders, psychotic symptoms, obsessive–compulsive disorder, substance misuse, somatoform disorders and acute and post-traumatic stress disorders. Cognitive impairment, somatic morbidity and the use of healthcare services were also assessed. The instrument also provides differential diagnoses for mental disorders because of general medical conditions.
Statistical analyses
Survey analyses were weighted according to the number of inhabitants in each country and stratified by gender and two age groups: 65–74 years old and 74 years or older. The adjusted lifetime, 12-month and current prevalence rates and 95% confidence limits were estimated as marginal means from a weighted logistic regression adjusting for age in 5-year intervals, gender and test centre. Reference Byers, Yaffe, Covinsky, Friedman and Bruce27 Group differences were tested using the main effect P-value of the model. Odds ratios (OR) and 95% confidence limits were also reported. All analyses were computed using Stata 12.1.
Results
Sample characteristics
The mean age of the n = 3142 MentDis_ICF65+ participants was 73.7 years after stratification (s.d. = 5.6), and half of the sample was female (50.7%). Participants had attended school for a mean of 10.3 years (s.d. = 3.2). The majority of participants were married (61%), 35% were separated, divorced or widowed and 5% had never been married (Table 1). Approximately 85% of participants were retired. About half of the participants rated their financial situation as good or very good (55%), with 8% rating it as poor or very poor.
Demographic characteristic | Total sample (n = 3142) |
---|---|
Age, n (%) | |
65–74 years | 1715 (54.6) |
75–84 years | 1427 (45.4) |
Gender, n (%) | |
Women | 1592 (50.7) |
Men | 1550 (49.3) |
Education, years of schooling (cut-off 13 years): mean (s.d.) | 10.3 (3.2) |
Born in country of interview, n (%) | 2519 (80.2) |
Marital status, n (%) | |
Married | 1915 (61.0) |
Separated/divorced/widowed | 1082 (34.5) |
Never been married/other | 142 (4.5) |
Work status, retired: n (%) a | 2640 (84.6) |
Financial situation, n (%) a | |
Very good | 356 (11.4) |
Good | 1372 (43.8) |
Just enough | 1145 (36.6) |
Poor | 219 (7.0) |
Very poor | 37 (1.2) |
a. n = 3128
12-month and lifetime prevalence
One in two individuals aged 65–84 years had experienced a mental disorder in their lifetime (Table 2). About one-third of the sample had a mental disorder within the past year (35.2%, 95% CI 31.0–39.5) (Table 3). There were significant differences between centres for all mental disorders in the past year except for panic disorder and major depressive episodes, with the highest prevalence rates found in Geneva (Switzerland: 47.1%), London/Canterbury (England: 44.4%), Jerusalem (Israel: 36.7%) and Hamburg (Germany: 35.4%). The lowest 1-year prevalence rates were in Madrid (Spain: 32.2%) and Ferrara (Italy: 27.7%).
Hamburg (Germany) | Ferrara (Italy) | London (England) | Madrid (Spain) | Geneva (Switzerland) | Jerusalem (Israel) | Over all | |
---|---|---|---|---|---|---|---|
Affective disorder | |||||||
Major depressive episode | 11.6 (9.5–13.7) | 9.2 (6.5–12.0) | 14.6 (11.4–17.8) | 10.5 (5.3–15.7) | 23.3 (19.3–27.4) | 14.8 (10.7–18.9) | 11.8 (9.7–14.0)*** |
Dysthymia | 2.5 (1.4–3.6) | 3.1 (2.1–4.1) | 3.4 (2.6–4.2) | 2.5 (1.6–3.5) | 4.0 (2.6–5.4) | 5.2 (3.5–6.8) | 2.9 (2.3–3.5)* |
Any bipolar disorder | 4.7 (3.2–6.2) | 2.4 (1.4–3.5) | 4.2 (2.9–5.5) | 6.9 (4.0–9.7) | 3.0 (2.2–3.7) | 6.1 (4.8–7.4) | 4.4 (3.3–5.5)* |
Any affective disorder | 13.2 (10.9–15.4) | 11.9 (9.5–14.4) | 18.0 (14.9–21.1) | 13.0 (7.8–18.3) | 26.5 (22.5–30.5) | 18.5 (12.3–24.6) | 14.3 (12.0–16.6)*** |
Anxiety | |||||||
Agoraphobia | 9.3 (7.3–11.4) | 4.7 (2.8–6.5) | 15.1 (11.8–18.4) | 10.0 (8.1–12.0) | 6.5 (4.3–8.7) | 4.1 (3.4–4.7) | 9.3 (6.7–12.0) |
Panic disorder | 3.9 (2.8–5.1) | 5.0 (2.3–7.8) | 7.5 (4.2–10.9) | 1.3 (0.8–1.8) | 3.5 (1.6–5.4) | 3.0 (2.2–3.8) | 4.5 (2.6–6.4)** |
Post-traumatic stress disorder | 2.0 (0.8–3.1) | 1.3 (0.8–1.7) | 6.3 (4.9–7.7) | 0.3 (0.0–0.8) | 2.0 (1.3–2.6) | 8.4 (5.6–11.2) | 2.5 (1.1–4.0)*** |
Any simple phobia | 15.7 (12.7–18.6) | 11.6 (9.3–14.0) | 21.2 (17.3–25.2) | 21.5 (16.4–26.7) | 15.3 (13.0–17.5) | 14.9 (11.8–17.9) | 16.7 (13.4–20.1)*** |
Any anxiety disorder | 24.1 (21.1–27.0) | 20.1 (16.5–23.7) | 32.6 (27.1–38.1) | 29.3 (23.9–34.8) | 20.7 (16.6–24.8) | 21.5 (16.3–26.6) | 25.6 (21.4–29.7)*** |
Substance misuse | |||||||
Alcohol dependence or misuse | 12.9 (9.8–16.1) | 1.7 (0.9–2.4) | 13.8 (10.5–17.1) | 3.8 (1.5–6.1) | 14.1 (9.5–18.8) | 3.2 (2.4–3.9) | 8.8 (4.5–13.2)*** |
Any substance-related disorder | 21.3 (18.1–24.6) | 12.9 (10.6–15.2) | 20.6 (17.1–24.0) | 16.3 (12.3–20.3) | 21.8 (16.6–26.9) | 12.8 (10.9–14.6) | 18.2 (14.6–21.8)*** |
Any somatoform disorder | 9.2 (7.4–11.0) | 5.9 (3.6–8.2) | 9.4 (7.2–11.5) | 3.5 (2.1–4.8) | 7.7 (6.6–8.8) | 10.7 (9.3–12.0) | 7.5 (5.7–9.3)*** |
Any mental disorder | 47.0 (44.1–49.9) | 38.8 (34.3–43.3) | 56.3 (49.7–63.2) | 46.3 (41.6–51.0) | 55.7 (53.5–58.0) | 46.0 (42.0–50.1) | 47.0 (42.8–51.3)*** |
a. Rates are adjusted for age and gender and accompanied by 95% confidence intervals.
* P<0.05,
** P<0.01,
*** P<0.001.
Hamburg (Germany) | Ferrara (Italy) | London (England) | Madrid (Spain) | Geneva (Switzerland) | Jerusalem (Israel) | Over all | |
---|---|---|---|---|---|---|---|
Affective disorder | |||||||
Major depressive episode | 11.1 (9.1–13.2) | 9.1 (6.4–11.8) | 14.5 (11.5–17.4) | 10.3 (5.3–15.3) | 23.1 (18.9–27.4) | 14.6 (10.3–18.9) | 11.6 (9.5–13.6) |
Dysthymia | 2.5 (1.4–3.6) | 3.1 (2.1–4.1) | 3.4 (2.6–4.2) | 2.5 (1.6–3.5) | 4.0 (2.6–5.4) | 5.2 (3.5–6.8) | 2.9 (2.3–3.5)* |
Any bipolar disorder | 2.1 (0.3–4.0) | 1.2 (0.3–2.2) | 2.9 (0.8–5.0) | 4.9 (2.1–7.6) | 1.2 (0.3–2.2) | 4.7 (3.0–6.5) | 2.5 (1.3–3.7)* |
Any affective disorder | 12.7 (10.6–14.8) | 11.1 (8.5–13.6) | 17.6 (14.7–20.4) | 12.4 (7.7–17.0) | 25.7 (21.2–30.3) | 17.8 (12.2–23.3) | 13.7 (11.4–15.9)*** |
Anxiety | |||||||
Agoraphobia | 6.0 (4.6–7.4) | 1.9 (0.2–3.5) | 7.4 (5.5–9.2) | 4.5 (2.7–6.2) | 3.7 (2.3–5.1) | 2.0 (1.5–2.5) | 4.9 (3.3–6.6)*** |
Panic disorder | 3.9 (2.9–5.0) | 4.2 (2.1–6.4) | 5.4 (3.3–7.4) | 1.1 (0.7–1.6) | 2.7 (1.6–3.9) | 2.3 (1.8–2.8) | 3.8 (2.6–5.0) |
Post-traumatic stress disorder | 1.0 (0.0–2.6) | 0.7 (0.0–1.5) | 2.4 (1.2–3.6) | 0 | 1.1 (0.1–2.2) | 5.9 (4.2–7.5) | 1.4 (0.4–2.4)*** |
Any simple phobia | 9.1 (7.3–10.9) | 6.8 (5.5–8.1) | 9.4 (6.1–12.7) | 12.7 (9.4–16.1) | 9.2 (7.4–11.1) | 8.9 (7.1–10.6) | 9.2 (7.2–11.1)* |
Any anxiety disorder | 16.8 (14.4–19.2) | 14.4 (11.6–17.3) | 20.8 (15.6–26.0) | 18.3 (14.4–22.3) | 14.1 (10.4–17.8) | 14.7 (10.7–18.7) | 17.2 (14.0–20.4)* |
Substance misuse | |||||||
Alcohol dependence or misuse | 7.4 (5.5–9.2) | 1.0 (0.5–1.5) | 9.5 (6.4–12.6) | 1.3 (0.4–2.2) | 9.1 (5.7–12.5) | 1.2 (0.2–2.2) | 5.3 (2.3–8.2)*** |
Any substance-related disorder | 11.2 (9.7–12.7) | 5.8 (1.5–10.2) | 11.2 (6.2–16.1) | 5.6 (3.5–7.6) | 12.7 (9.4–15.9) | 3.7 (2.6–4.7) | 8.9 (6.1–11.7)*** |
Any somatoform disorder | 4.8 (4.0–5.7) | 2.7 (1.3–4.2) | 5.3 (3.0–7.6) | 2.5 (1.3–3.8) | 3.7 (3.0–4.5) | 8.4 (5.4–11.5) | 4.1 (3.1–5.1)* |
Any mental disorder | 35.4 (33.2–37.6) | 27.7 (21.1–34.4) | 44.4 (39.5–49.3) | 32.2 (29.1–35.3) | 47.1 (44.4–49.9) | 36.7 (33.0–40.4) | 35.2 (31.0–39.5)*** |
a. Rates are adjusted for age and gender and accompanied by 95% confidence intervals.
** P<0.05,
** P<0.01,
*** P<0.001.
The most prevalent mental disorders were anxiety disorders (17.2%, 95% CI 14.0–20.4), affective disorders (13.7%, 95% CI 11.4–15.9) and substance-related disorders (8.9%, 95% CI 6.1–11.7). Elderly people living in London/Canterbury and in Madrid showed the highest prevalence rates for anxiety disorder in the past year (England: 20.8%, 95% CI 15.6–26.0; Spain: 18.3, 95% CI 14.4–22.3), whereas participants living in Ferrara and Geneva reported the lowest prevalence rates (Italy: 14.4, 95% CI 11.6–17.3; Switzerland: 14.1, 95% CI 10.4–17.8). Participants living in Jerusalem reported the highest lifetime and 12-month rates of post-traumatic stress disorder (Israel: 5.9, 95% CI 4.2–7.5) (Tables 2 and 3). Elderly people in Geneva reported the highest prevalence rate for affective disorder within the past year (Switzerland: 25.7%, 95% CI 21.2–30.3), followed by Jerusalem (Israel: 17.8%, 95% CI 12.2–23.3) and London/Canterbury (England: 17.6%, 95% CI 14.7–20.4). Lower prevalence rates were found in Hamburg (Germany: 12.7, 95% CI 10.6–14.8), Madrid (Spain: 12.4, 95% CI 7.7–17.0) and Ferrara (Italy: 11.1%, 95% CI 8.5–13.6). The highest prevalence rates for substance–related disorders were found in Geneva (Switzerland: 12.7%, 95% CI 9.4–15.9), Hamburg (Germany: 11.2%, 95% CI 9.7–12.7) and London/Canterbury (England: 11.2%, 95% CI 6.2–16.1). The lowest prevalence rate for substance-related disorders was found in Jerusalem with 3.7% (Israel: 95% CI 2.6–4.7) (Table 3).
Current prevalence
Current prevalence rates are shown in Table 4. Nearly a quarter of the sample were diagnosed with any current mental disorder (23.3%, 95% CI 19.9–26.7) with the highest prevalence rates found in Geneva (30.4%), London/Canterbury (28.4%) and Jerusalem (27.0%). Lower prevalence rates for any current mental disorder were found in Hamburg (23.3%), Madrid (21.0) and Ferrara (19.5%).
Hamburg (Germany) | Ferrara (Italy) | London (England) | Madrid (Spain) | Geneva (Switzerland) | Jerusalem (Israel) | Overall | |
---|---|---|---|---|---|---|---|
Affective disorder | |||||||
Major depressive episode | 5.0 (4.2–5.8) | 5.5 (3.5–7.4) | 8.0 (6.3–9.7) | 4.9 (2.2–7.5) | 13.0 (10.1–16.0) | 9.4 (3.8–15.0) | 6.0 (4.7–7.3)*** |
Dysthymia | 2.5 (1.3–3.7) | 2.9 (1.8–4.1) | 2.9 (2.0–3.8) | 2.6 (1.8–3.4) | 3.8 (2.6–5.1) | 4.3 (3.2–5.4) | 2.7 (2.1–3.4)* |
Any bipolar disorder | 0 b | 0 b | 0 b | 0 b | 0 b | 0 b | 0 b |
Any affective disorder | 6.9 (5.6–8.3) | 7.5 (5.2–9.8) | 10.3 (7.9–12.6) | 6.5 (4.7–8.3) | 15.6 (11.5–19.7) | 11.4 (5.2–17.7) | 8.0 (6.3–9.6)*** |
Anxiety | |||||||
Agoraphobia | 4.9 (3.9–5.9) | 1.5 (0.0–3.6) | 3.8 (2.1–5.6) | 2.9 (0.9–3.9) | 2.7 (1.9–3.5) | 1.6 (1.1–2.0) | 3.4 (0.1–4.8)*** |
Panic disorder | 1.4 (0.8–2.1) | 0.7 (0.1–1.4) | 1.2 (0.5–1.8) | 0.4 (0.1–0.6) | 0.6 (0.0–1.1) | 0.8 (0.6–1.0) | 1.0 (0.6–1.4) |
Any simple phobia | 7.0 (5.8–8.1) | 5.0 (3.9–6.0) | 7.6 (4.9–10.3) | 7.1 (5.1–9.2) | 5.3 (3.9–6.7) | 5.6 (4.0–7.1) | 6.6 (5.1–8.0)* |
Post-traumatic stress disorder | 1.0 (0.0–2.6) | 0.7 (0.0–1.5) | 2.4 (1.2–3.6) | 0 | 1.1 (0.1–2.2) | 5.6 (3.8–7.4) | 1.4 (0.4–2.4)** |
Any anxiety disorder | 12.0 (10.1–13.8) | 9.4 (7.5–11.4) | 13.4 (10.2–16.5) | 11.3 (8.5–14.0) | 7.8 (5.5–10.1) | 10.2 (7.3–13.1) | 11.4 (9.1–13.6)* |
Substance misuse | |||||||
Alcohol dependence or misuse | 1.8 (1.5–2.1) | 0.2 (0.0–0.4) | 1.4 (1.0–1.8) | 0.2 (0.0–0.5) | 4.9 (3.5–6.3) | 0.5 (0.1–0.9) | 1.1 (0.5–1.7)*** |
Any substance-related disorder | 5.4 (5.0–5.8) | 4.8 (2.5–7.1) | 3.4 (2.4–4.3) | 4.1 (2.0–6.2) | 7.7 (7.0–8.4) | 2.1 (1.2–3.0) | 4.6 (3.7–5.6)*** |
Any somatoform disorder | 4.0 (3.1–4.9) | 2.0 (1.0–3.1) | 4.7 (2.8–6.5) | 2.3 (1.3–3.4) | 2.7 (1.7–3.7) | 7.4 (5.0–9.7) | 3.4 (2.5–4.4) |
Any mental disorder | 23.3 (21.2–25.6) | 19.5 (16.0–23.1) | 28.4 (23.8–33.0) | 21.0 (17.0–25.3) | 30.4 (27.1–33.6) | 27.0 (20.0–33.9) | 23.3 (19.9–26.7)*** |
a. Rates are adjusted for age and gender and accompanied by 95% confidence intervals.
b. Prevalence was below 0.5.
* P<0.05,
** P<0.01,
*** P<0.001.
The most prevalent category was anxiety disorders (11.4%, 95% CI 9.1–13.6), followed by affective disorders (8.0%, 95% CI 6.3–9.6) and substance-related disorders (4.6%, 95% CI 3.7–5.6). Elderly participants living in London/Canterbury or Hamburg had the highest prevalence rates for anxiety disorders in the past month (England: 13.4%, 95% CI 10.2–16.5; Germany: 12.0%, 95% CI 10.1–13.8). In contrast, elderly people living in Ferrara and Geneva had the lowest current prevalence rates for anxiety disorders (Italy: 9.4%, 95% CI 7.5–11.4; Switzerland: 7.8%, 95% CI 5.5–10.1) (Table 4). The current prevalence rate for affective disorders was markedly higher in Geneva (Switzerland: 15.6%, 95% CI 11.5–19.7) than in Ferrara (Italy: 7.5%, 95% CI 5.2–9.8) and Madrid (Spain: 6.5%, 95% CI 4.7–8.3). The highest prevalence rates for current substance-related disorders were 7.7% in Geneva (Switzerland: 95% CI 7.0–8.4) and 5.4% in Hamburg (Germany: 95% CI 5.0–5.8); the lowest rate was 2.1% in Jerusalem (Israel: 95% CI 1.2–3.0) (Table 4).
Discussion
Main findings
Studies of mental disorders that do not focus on dementia or depression among elderly Europeans are very scarce. This is the first study to use a standardised and structured clinical interview for mental disorders adapted to the needs of elderly people to report lifetime, 12–month and current prevalence rates for a range of mental disorders in elderly people residing in different European and associated countries. One in two individuals aged 65 to 84 years had experienced at least one mental disorder in their lifetime, one in three had done so within the past year, and nearly one in four currently had a mental disorder. The most prevalent disorders were anxiety disorders, followed by affective and substance-related disorders.
Comparison with findings from other studies
In comparison with other epidemiological studies of old age Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9,Reference Byers, Yaffe, Covinsky, Friedman and Bruce27 and adulthood Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha and Bryson3,Reference Kessler, Birnbaum, Shahly, Bromet, Hwang and McLaughlin28 that used standardised interviews such as the CIDI, our findings show higher prevalence rates, whereby the proportion of those affected is in accordance with the prevalence rates in adulthood. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha and Bryson3,Reference Kessler, Birnbaum, Shahly, Bromet, Hwang and McLaughlin28 Compared with other studies on old age, we found higher current prevalence rates of major depression (6% in our study v. 3.3%), Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9,Reference Byers, Yaffe, Covinsky, Friedman and Bruce27 agoraphobia (3.4% v. 0.5%) Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9,Reference Byers, Yaffe, Covinsky, Friedman and Bruce27 and alcohol disorders (1.4% v. 0.96%). Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9,Reference Byers, Yaffe, Covinsky, Friedman and Bruce27 There was only one European study from the 1990s that found comparable rates for current affective disorders, which also used an age-sensitive measure for depression. Reference Copeland, Beekman, Braam, Dewey, Delespaul and Fuhrer29 It is plausible that previous epidemiological studies underestimated the prevalence rates of mental disorders in elderly people because they did not use an interview adapted to meet concerns specific to elderly people. The sentences in the CIDI65+ were changed to make them easier for elderly people to understand and respond to, and this may have contributed to more valid estimates of mental disorders. Reference Wittchen, Strehle, Gerschler, Volkert, Dehoust and Sehner25 Another reason for the higher prevalence rates in our study could be that all countries used the same methodological approach, whereas previous studies may have underestimated prevalence through use of different instruments and possible measurement errors. Another important point to consider is the use of categorical v. dimensional instruments. There is an explicit difference in prevalence rates obtained with dimensional and categorical instruments in affective disorders. Our study builds on established categorical criteria for mental disorders as defined by the DSM-IV. 19 However, due to the multidimensional nature of psychopathology, the criteria and thresholds of the DSM-IV are not without major problems. Reviews that compare findings for dimensional measures of current psychopathology with categorical current prevalence typically reveal higher rates for dimensional measures that might vary depending on the choice of cut-offs that are used. Reference Volkert, Schulz, Härter, Wlodarczyk and Andreas9 There is also a possibility that the presented rates are still under-estimating the true prevalence of mental disorders in elderly people because we excluded people with severe cognitive impairment from the study. Reference O'Connor30
Prevalence rates in different countries
The prevalence rates found in our study are comparable among the participating countries except for fluctuations in the absolute size. Furthermore, there were several important differences between the six catchment areas. Interpreting those differences is complex because of the large differences between the countries with regard to the availability of mental healthcare, the specific economic situation, immigration status, living circumstances, attitudes towards mental disorders in elderly people, experience of traumatic events and lifestyles. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha and Bryson3,Reference Lahelma, Laaksonen, Martikainen, Rahkonen and Sarlio-Lahteenkorva31 Such factors may be associated with greater or lesser willingness and ability to express psychological symptoms during an interview. Another interesting finding is the marked difference between the rates of substance misuse in the southern areas of Ferrara (Italy), and Madrid (Spain) and the more northern European areas of London/Canterbury (England), Hamburg (Germany), Geneva (Switzerland); these results are in line with previous European studies on substance misuse in adulthood. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha and Bryson32,Reference de Girolamo, Polidori, Morosini, Scarpino, Reda and Serra33
Strengths and limitations
An advantage of the current study was the use of a reliable, structured and standardised instrument that was adapted to the needs of elderly people. Trained interviewers assessed participants in catchment areas in Hamburg (Germany), in London/Canterbury (England), Geneva (Switzerland), Madrid (Spain), Ferrara (Italy) and Jerusalem (Israel) face to face according to DSM-IV criteria, and the reliability of the instrument was evaluated beforehand in a pilot phase of the study.
Nevertheless, the study has some limitations. First, the size of the sample was limited per country and per catchment area. Second, the representativeness of our study may be limited because we found small but significant differences for some sociodemographic data between our sample and the total population of the catchment area or country. Third, this study found higher prevalence rates of mental disorders in older people than did previous studies; thus, the question arises of whether the response rate in this study is associated with an overestimation or underestimation of prevalence compared with previous studies. However, previous studies have found both higher Reference Eaton, Kramer, Anthony, Dryman, Shapiro and Locke34,Reference de Graaf, Bijl, Smit, Ravelli and Vollebergh35 and lower prevalence rates of mental disorders in non-responders. Reference Alonso, Angermeyer, Bernert, Bruffaerts, Brugha and Bryson3 Kessler et al Reference Kessler, Chiu, Demler, Merikangas and Walters36 found no evidence for a selection bias related to mental illness in the US National Comorbidity Survey Replication (NCS-R). The authors concluded that to the extent the bias exists, prevalence estimates may be regarded as more conservative. In addition, many authors have stated that non-response can, but need not, automatically mean there is a non-response bias in survey estimates. Reference Groves37–Reference Keeter, Miller, Kohut, Groves and Presser39 Therefore, it may also be possible that the prevalence of mental disorders in older people is still under-estimated in our study. Furthermore, we were unable to include other important population variables such as educational level or financial situation. Another limiting factor regarding representativeness was our set of inclusion criteria: we did not include people with severe cognitive impairment, homeless people or people who did not have sufficient knowledge of the language in which the interview was conducted. In addition, due to the nature of epidemiological studies with elderly people, we were unable to control for a possible recall bias, especially regarding lifetime symptoms. Reference Byers, Yaffe, Covinsky, Friedman and Bruce27
Future directions for research
Our study showed a high prevalence rate of lifetime, 12-month and current mental disorders in people aged 65 to 84 years in different European and associated countries. The newly adapted CIDI65+ instrument shows the need for further research in the diagnostics of mental disorders in elderly people, which is a crucial step towards more comprehensive mental health approaches for these age groups. Future studies could investigate the prevalence of mental disorders in even older people (of 85 years and above, as this age group is growing rapidly). However, this group may require additional modifications in diagnostic assessment, as additional challenges are associated with very old age (for example, cognitive impairment). Future European studies could also include more countries, such as those from the Scandinavian or Eastern European regions, and consider including nursing home residents or elderly people with cognitive impairments. Translations into further languages and extensions to surveys in other continents would be a further milestone. In addition, data about somatic diseases and their relationship with quality of life are needed. Additional studies could also integrate primary care perspectives into the diagnostics as this is where the majority of older adults with mental health problems are treated; thus, this will enable the general practitioner to give advice on specialised mental healthcare. Finally, further studies should examine whether the use of services corresponds to the high burden of mental illness in elderly people.
Funding
This study is funded by a grant from the European Commission (Grant No: ) within the 7th Framework Research Program of the European Union.
Acknowledgements
We gratefully thank all participants and all interviewers in our study.
eLetters
No eLetters have been published for this article.