Suicidal ideation and reported suicide attempts in Greece during the economic crisis

Abstract

The financial crisis in Greece is largely impinging on the health and mental health of the population, raising concerns about a potential rise in suicide rates. The aim of this study was to explore changes in suicidal ideation and reported suicide attempts between 2009 and 2011 in a representative sample of the population and in several population subgroups. The socio-economic predictors of suicidal ideation and suicide attempts in 2011 were also investigated. Two nationwide cross-sectional telephone surveys were conducted in 2009 and 2011 using the same methodology. A random and representative sample of 2192 and 2256 people, respectively, took part in the surveys. Between 2009 and 2011, there was a substantial increase in the prevalence of suicidal ideation and reported suicide attempts. People suffering from depression, men, married individuals, people experiencing financial strain, people with low interpersonal trust, and individuals with a history of suicide attempts were particularly vulnerable.

Suicide is a huge public health issue worldwide. Each year roughly one million people die by suicide around the world, making it one of the leading causes of death internationally 1,2. Congruent with this, the World Health Organization has encouraged collection of data on the prevalence and risk factors for suicide to inform the design of public health strategies and health care policies 3. Previous research has indicated that suicidal thoughts and attempts, hereafter referred to as “suicidality”, should be conceptualized as predictors of more serious suicidal acts 4,5.

Suicide/suicidality has been found to bear a strong association with both psychiatric and socio-economic factors 6. The presence of a psychiatric disorder is among the most consistently identified risk factors for suicide/suicidality 7,8, with psychological autopsy studies revealing that more than 90% of suicide victims suffered from mental illness at the time of their death 9, most commonly from mood disorders 10. Conversely, growing evidence has documented a link between suicide/suicidality and socio-economic factors, such as unemployment, income, and family status 6,11,12. In line with all this, during periods of economic recession, with its ramifications including financial hardship, job loss, and elevated prevalence of depression 13–16, suicide becomes a nagging concern 17.

In 2008, as a consequence of the global financial crisis and the local unrelenting spending, the Greek population started facing a serious socio-economic turmoil. In 2010, a memorandum of economic and financial policies was signed in order to avert Greece's default. In 2011, the financial and socio-political climate became even worse: the unemployment rate reached 16.6% in May, the gross domestic product (GDP) further declined by 6.1%, and the debt grew from 105.4% of GDP in 2007 to 160.9% 18,19. The health and mental health implications of the economic crisis in Greece have attracted increasing attention 13,20,21, with a strong emphasis on a potential rise in suicides/suicidality 21–24. This is of primary importance, given the country's low suicide rates recorded in previous studies 25–27.

In an endeavor to gauge the impact of the financial crisis on the mental health of the Greek population, two cross-sectional nationwide studies were conducted in 2009 and 2011 (before and after the introduction of the memorandum of economic and financial policies in the daily life of the population). The present study concentrates on suicidality findings. It reports on the prevalence of suicidal ideation and suicidal attempts in a representative sample of the Greek population and in various population subgroups between 2009 and 2011 and on predictors of suicidal ideation and suicidal attempts in 2011.

METHODS

The sampling frame of the study was the national phone-number databank, providing coverage of the vast majority of households in the country. A random sample of telephone numbers belonging to individuals were selected from the directory. Within each household, the person who had his or her birthday last was selected for an interview.

Telephonic interviews were conducted with adults aged between 18 and 69 years during the same time period (February to April) in 2009 and 2011. In 2009, of the 2667 calls made, there were 2192 successfully completed interviews (response rate=82.2%). One hundred eighty four people (6.9%) hung up immediately and 290 (10.9%) refused to be interviewed or did not complete the interview. In 2011, of the 2820 calls made, there were 2256 successfully completed interviews (response rate=80.5%). Two hundred and three people (7.2%) hung up immediately and 347 (12.3%) refused to be interviewed or did not complete the interview. No statistically significant differences were found between participants who could be interviewed and those who could not in terms of gender, age, and place of residence.

All interviews were carried out by well-trained graduates in social sciences, including psychology. The training included lectures, role playing, and pilot phone interviews which were taped and reexamined, for a total of 60 h.

The method of computer-assisted telephone interviewing was used 28, as it enables automatic control of questionnaire branching, on-line verification checks, and automatic scheduling of future call backs (e.g., if the call is not answered or the interview is not completed).

Information about the occurrence of a major depressive episode, suicidal ideation, and suicide attempts during the past month was collected using the relevant module of the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) 29, which has been standardized in the Greek population and extensively used in clinical and epidemiologic studies (e.g., 30,31.

To confirm the reliability of the diagnosis, a random sample of the calls (10%) was rated by ten interviewers. The Fleiss’ kappa was found to be 0.89 in 2009 and 0.90 in 2011. Furthermore, a random sample of participants (10%) agreed to be reinterviewed by a psychiatrist. The rate of diagnostic agreement was 87% in 2009 and 89% in 2011.

To assess participants’ degree of financial strain, the Index of Personal Economic Distress (IPED) was used 13. This self-reported measure encompasses eight questions describing participants’ difficulty in fulfilling daily financial demands of a household during the last six months. Responses are made on a three-point scale (never, 1; sometimes, 2; often, 3), while the composite score of the scale can range from 8 (no economic problems) to 24 (serious economic problems). A previous exploration of the psychometric properties of the scale identified 15 as the cut-off point producing the best results in terms of sensitivity and specificity 13. Congruent with this, participants scoring above this cut-off point were regarded as being in high financial distress. The scale displayed good internal consistency (Cronbach α=0.93).

Interpersonal trust was assessed by the relevant questions from the European Social Survey 32. Questions were rated on a scale from 0 to 10, with values close to 0 indicating low interpersonal trust. As the internal consistency of the scale was good (Cronbach α=0.77), a composite scale score was computed.

The socio-demographic characteristics of the respondents (gender, age, family status, place of residence, education, employment status) and their previous contacts with mental health professionals, use of psychiatric medication, and admissions to an inpatient service were recorded.

The study received approval from the University Mental Health Research Institute Ethics Committee and was performed in accordance with the ethical standards delineated in the 1964 Declaration of Helsinki. Informed consent was obtained from all participants.

To explore differences between categorical variables, χ² test with Yates’ correction was used. To identify predictors of suicidal ideation and recent suicide attempt, two logistic regression models were computed, with presence/absence of suicidal ideation/recent suicidal attempt as the dependent variable. All analyses were performed by SPSS v.17.

RESULTS

In Table 1, the socio-demographic characteristics of the 2009 and 2011 samples are presented, along with those of the Greek population according to the 2001 census. The composition of the samples was congruent with the population census.

Table 1.

Socio-demographic characteristics of the 2009 and 2011 samples compared to the 2001 population census

	2009		2011		2001
	n	%	n	%	%
Gender
Male	1080	49.3	1090	48.3	46.8
Female	1112	50.7	1166	51.7	54.2
Age groups
<24	210	9.6	223	9.9	9.2
25–34	411	18.7	426	18.9	19.8
35–44	325	14.8	367	16.3	16.2
45–54	452	20.6	425	18.8	19.4
55–64	384	17.5	368	16.3	16.0
>65	409	18.6	447	19.8	18.4
Marital status
Single	486	22.2	455	20.2	21.0
Married	1487	67.8	1585	70.3	70.0
Widowed	123	5.9	138	6.1	7.1
Divorced	96	4.3	78	3.5	3.0
Education (years)
<11	1077	49.1	1351	59.9	61.0
12	813	37.1	654	29.0	28.0
>13	302	13.8	251	11.1	11.0
Place of residence
Athens greater area	808	36.9	847	37.5	39.3
Thessalonica and central Macedonia	275	12.5	388	17.2	15.0
Rest of country	1109	50.6	1021	45.3	45.7
Occupation
Employed	1184	84.4	1214	85.5	94.0
Unemployed	218	15.6	206	14.5	6.0

The proportion of respondents who reported suicidal ideation was 6.7% in 2011 versus 5.2% in 2009 (χ²=3.92, df=1, p=0.04). As shown in Table 2, a significant increase in the prevalence of suicidal ideation was observed in men (7.1% vs. 4.4%, χ²=6.41, df=1, p=0.011), but not in women (6.3% vs. 5.9%, χ²=0.05, df=1, p=0.817). The prevalence of suicidal ideation also increased among respondents aged 55–64 years (7.2% vs. 1.9%, χ²=14.41, df=1, p<0.001), while it decreased in those younger than 24 years (4.9% vs. 13.9%, χ²= 15.83, df=1, p<0.001). An increase was also observed among married individuals (7.3% vs. 2.3%, χ²=38.08, df=1, p<0.001), while the prevalence decreased in unmarried respondents (5.6% vs. 11.6%, χ²=17.24, p<0.001). Suicidal ideation was also increased between 2009 and 2011 among respondents who used psychotropic medications (22.7% vs. 4.5%, χ²=11.10, df=1, p<0.001) and those who had sought help from a mental health professional (17.3% vs. 8.3%, χ²=13.36, df=1, p<0.001).

Table 2.

Prevalence (%) of suicidal ideation in population subgroups in 2009 and 2011

	2009 (n=2192)	2011 (n=2256)	χ2 (df=1)	p
Gender
Male	4.4	7.1	6.41	0.011
Female	5.9	6.3	0.05	0.817
Age
<24	13.9	4.9	15.83	<0.001
25–34	7.0	9.9	1.19	0.275
35–44	3.4	6.6	2.69	0.100
45–54	7.4	7.8	0.00	0.923
55–64	1.9	7.2	14.41	<0.001
>65	3.2	5.0	1.08	0.299
Marital status
Unmarried	11.6	5.6	17.24	<0.001
Married	2.3	7.3	38.08	<0.001
Education
<11	4.3	6.8	3.39	0.065
12	6.3	6.8	0.01	0.921
>13	6.0	6.6	0.17	0.680
Place of residence
Rest of country	4.9	5.8	0.95	0.328
Athens	5.7	8.0	3.16	0.075
Employment status
Unemployed	4.0	10.2	3.31	0.069
Economically active	5.2	7.2	3.05	0.080
Not economically active	5.3	4.9	0.06	0.796
Sought professional help
No	4.5	4.7	0.05	0.810
Yes	8.3	17.3	13.36	<0.001
Used psychotropic medications
No	5.2	5.9	0.86	0.353
Yes	4.5	22.7	11.10	<0.001
Major depression diagnosis
No	2.5	2.3	0.15	0.69
Yes	54.9	56.0	0.00	0.94

The proportion of respondents who reported a recent suicide attempt was 1.5% in 2011 versus 1.1% in 2009 (χ²=1.16, df=1, p=0.28) 23. As shown in Table 3, a significant increase in the prevalence of a recent suicide attempt was observed in men (2.0% vs. 0.4%, χ²=11.12, df=1, p<0.001), but not in women (1.0% vs. 1.8%, χ²=1.93; df=1, p=0.165). The prevalence of a recent suicide attempt also increased among people aged 25–34 years (1.7% vs. 0.0%, χ²=6.73, df=1, p=0.009) and 35–44 years (2.3% vs. 0.0%, χ²=4.56, df=1, p=0.033), whereas it decreased in respondents aged 45–54 years (0.7% vs. 3.2%, χ²=5.33, df=1, p=0.021). An increase was also observed among married people (1.6% vs. 0.3%, χ²=10.81, df=1, p=0.001) and in those with the highest educational attainment (1.2% vs. 0.1%, χ²=5.96, df=1, p=0.015). No unemployed respondent reported a recent suicide attempt in 2009, while the proportion was 4.4% in 2011 (χ²=4.12, df=1, p=0.042).

Table 3.

Prevalence (%) of reported suicide attempts in population subgroups in 2009 and 2011

	2009 (n=2192)	2011 (n=2256)	χ2 (df=1)	p
Gender
Male	0.4	2.0	11.12	<0.001
Female	1.8	1.0	1.93	0.165
Age
<24	0.0	1.3	1.53	0.215
25–34	0.0	1.7	6.73	0.009
35–44	0.0	2.3	4.56	0.033
45–54	3.2	0.7	5.33	0.021
55–64	1.2	2.8	1.93	0.164
>65	1.4	0.6	0.55	0.457
Marital status
Unmarried	2.8	1.4	3.14	0.076
Married	0.3	1.6	10.81	0.001
Education
<11	2.1	2.9	0.39	0.530
12	0.3	1.4	1.25	0.262
>13	0.1	1.2	5.96	0.015
Place of residence
Rest of country	1.3	1.8	0.74	0.388
Athens	0.9	1.1	0.02	0.874
Employment status
Unemployed	0.0	4.4	4.12	0.042
Economically active	0.4	1.1	1.95	0.162
Not economically active	1.9	1.6	0.03	0.861
Sought professional help
No	0.7	1.3	2.49	0.115
Yes	2.7	2.6	0.00	1.000
Used psychotropic medications
No	1.1	1.3	0.36	0.546
Yes	1.1	5.2	1.26	0.260
Major depression diagnosis
No	0.8	0.5	0.97	0.324
Yes	6.6	13.0	2.56	0.109

As shown in Table 4, the significant predictors of suicidal ideation in 2011 were the presence of major depression during the previous month, financial hardship experienced, a previous history of suicide attempt, and a low interpersonal trust. People who fulfilled DSM-IV criteria for major depression were 48 times more likely to present suicidal ideation than people without that diagnosis (OR=47.72, 95% CI=22.18–62.66, p<0.001). Respondents with a history of suicide attempt were seven times more likely to report suicidal ideation than people without such a history (OR=6.89, 95% CI=1.66–8.60, p<0.01).

Table 4.

Logistic regression results with suicidal ideation (presence/absence) as dependent variable

	OR	95% CI	p
Major depression diagnosis
Noa
Yes	47.72	22.18–62.66	<0.001
Gender
Femalea
Male	1.96	0.93–4.13	NS
Age (years)	0.96	0.72–1.29	NS
Education (years)	1.11	0.58–2.13	NS
Place of residence
Athensa
Rest of country	1.57	0.70–3.54	NS
Marital status
Not marrieda
Married	1.51	0.59–3.82	NS
Employment status
Unemployeda
Economically active	0.92	0.26–3.23	NS
Not economically active	0.71	0.15–3.25	NS
IPED (score)	1.07	1.07–1.40	<0.05
Previous suicide attempt
Noa
Yes	6.89	1.66–8.60	<0.01
Interpersonal trust scale (score)	0.93	0.87–0.98	<0.001

^aReference category

IPED – Index of Personal Economic Distress

As reported in Table 5, the significant predictors of a recent suicide attempt in 2011 were the presence of major depression during the previous month, financial hardship, a previous history of suicide attempt, being male, and being married. People who fulfilled DSM-IV criteria for major depression were 97 times more likely to report a suicide attempt during the previous month than people without that diagnosis (OR=97.39, 95% CI=71.06–134.45, p<0.01). Men were 12 times more likely to report a recent suicide attempt than women (OR=12.26, 95% CI=5.10–16.78, p<0.05), and married people were 53 times more likely to report it than unmarried ones (OR=53.29, 95% CI=33.29–64.40, p<0.05). People with a previous history of suicide attempt were 14 times more likely to report a recent suicide attempt than those without such a history (OR=14.41, 95% CI=1.79–15.95, p<0.01).

Table 5.

Logistic regression results with reported suicide attempt (presence/absence) as dependent variable

	OR	95% CI	p
Major depression diagnosis
Noa
Yes	97.39	71.06–134.45	<0.01
Gender
Femalea
Male	12.26	5.10–16.78	<0.05
Age (years)	0.43	0.16–1.16	NS
Education (years)	2.88	0.53–5.69	NS
Place of residence
Athensa
Rest of country	1.89	0.21–6.67	NS
Marital status
Unmarrieda
Married	53.29	33.29–64.40	<0.05
Employment status
Unemployeda
Economically active	0.40	0.02–6.74	NS
Not economically active	14.90	0.35–24.76	NS
IPED (score)	1.23	1.02–1.49	<0.05
Previous suicide attempt
Noa
Yes	14.41	1.79–15.95	<0.01
Interpersonal trust scale (score)	1.03	0.87–1.22	NS

^aReference category

IPED – Index of Personal Economic Distress

DISCUSSION

This study provides evidence of a substantial increase in the prevalence of suicidal ideation and reported suicide attempts between 2009 and 2011 in a representative sample from the Greek population. People suffering from depression, men, married individuals, people experiencing financial strain, people with low interpersonal trust, and those with a previous history of suicide attempts were particularly vulnerable.

Previous research in the country has displayed important fluctuations in reported suicide ideation and suicide attempt rates across the years as well as a strong link between these rates and socio-economic factors 13,26. Specifically, in an epidemiologic study conducted in 1978, the 1-month prevalence of suicidal ideation was found to be 4.8%, whereas in 1984 it reached 10.9% 26. This substantial rise was attributed to the economic recession striking the country at the time, which peaked in 1984 and necessitated the application of economic measures during the period 1985–1986 33. A similar trend was recorded for reported suicide attempts during the month prior to the interview, with a rate of 0.7% in 1978 and 2.6% in 1984. In the year 2008, prior to the outset of the economic crisis in Greece, the 1-month prevalence of suicidal ideation was 2.4% and that of suicide attempts was 0.6 13. Both rates are substantially lower than those in 2009 and 2011.

In a similar vein, Zacharakis et al 25 explored suicide rates in Greece in the period 1980–1995 and found they were among the lowest in the world, possibly due to an array of social and cultural factors: the strong family ties, the support and help offered by social networks, and religious practices. Suicide rates remained low until 2009. According to the official data from the police registry, there were 507 completed suicides and serious suicide attempts in 2009, 622 in 2010, and 598 in 2011 34.

These findings run in parallel with the gradual increase in the prevalence of major depression in Greece during the past few years 16, confirming that the economic crisis in Greece has impinged on the mental health of the population, similarly to what observed in other countries 14,35.

Our findings about the population subgroups most vulnerable to suicidality can be explained by the socio-economic variables that play a prominent role during recession, in particular unemployment, job insecurity, and loss of income 36–38. For example, people with high educational attainment might face difficulty in finding a job or might be experiencing a mismatch between their educational level and their salary. Men and married people are the breadwinners in the Greek family and may feel incapable of providing their families with the essentials due to unemployment or job insecurity. The age groups displaying the highest increase in suicidality in our study were those experiencing the largest changes in unemployment rates in recent years 39. Furthermore, people aged 55–64 years had to deal with early compulsory retirement in 2011. Research has shown that people who approach retirement at the time of an economic crisis are particularly susceptible to stress and uncertainty due to disruptions to their financial preparations 40 and reliance on fixed income resources (e.g., pensions), which are subject to market fluctuations 41.

The significant increase in the prevalence of suicidality among those who have sought professional help and those who take psychotropic medications is congruent with the increase in the use of mental health services and in the incidence of help-seeking for psychological problems during times of economic adversity 42.

In our study, the only significant protective factor with respect to suicidal ideation was interpersonal trust. This has been considered for long as a measure of social capital 43, which in turn has been linked to lower mortality rates 44, including deaths from suicide 45, lending in this way support to Durkheim's seminal work on social integration and suicide 46. Social capital in general and interpersonal trust in particular can protect citizens from developing suicidal ideation, even during periods of economic crisis. This supports recommendations to enhance social capital to mitigate the health and mental health effects of the recession 17,47.

Our data suggest that being male increases substantially the odds of attempting suicide. This can be explained by a variety of factors, such as job insecurity, loss of income, men's social roles being closely bound to employment conditions, and low availability of social support. For example, Moller-Leimkuhler 48 associated male vulnerability for suicide with the observation that men tend to be less socially integrated, have less social support, and are emotionally more isolated.

The increased risk for suicide attempts among married individuals can be interpreted in the light of the family stress model 49,50, which postulates that couples experiencing financial strain become angry, sad, and pessimistic about the future. The emotional distress leads the spouses to interact in a nonsupportive and irritable manner, resulting in marital conflict. Marital conflict in turn is associated with depression 49,50 and suicide 51. This model has received support in studies conducted in the United States 52, Czech Republic 53, Finland 54, Romania 55, and Korea 56.

Clinicians should be aware of the above risk factors. They should explore in depth whether men, people with major depression, those experiencing financial strain, individuals with previous suicide attempts, married individuals, and people with low levels of interpersonal trust display suicidal ideation or have attempted suicide. Clinicians should not ignore that about 90% of unplanned and 60% of planned first suicide attempts occur within 1 year of the onset of suicidal ideation 8. Within this context, early detection of mental health problems and depression in particular is of primary importance 57. Programs specifically designed to train primary care physicians in the recognition and treatment of depression have reported reductions in suicide rates 58. The development of social protection programs such as labor force expansion programs, social support systems, and access to health care and health insurance might mitigate the negative mental health effects of economic crisis and unemployment 17,47.

The present study was not without its shortcomings. Due to the cross-sectional nature of the design, causal inferences cannot be drawn. In addition, reported suicidal ideation and suicide attempts can be susceptible to recall bias or to reluctance on the part of respondents to disclose such a sensitive information. Although the anonymity of telephone interviewing has been shown to facilitate the disclosure of this type of information 59,60, it cannot be excluded that the prevalence rates found in this study represent an underestimation of true rates.

In conclusion, our findings confirm that the influence of the recession on mental health in the Greek population is being pervasive and that systematic efforts should be undertaken to mitigate these effects. A strategic plan for treating major depression and preventing suicide should be implemented, if Greece is to retain its low position in suicide rates worldwide.