Decline in suicidal ideation among patients with MDD is preceded by decline in depression and hopelessness

Journal of Affective Disorders 95 (2006) 95 – 102 www.elsevier.com/locate/jad

Research report

Decline in suicidal ideation among patients with MDD is preceded by decline in depression and hopelessness Petteri Sokero a , Mervi Eerola a,b , Heikki Rytsälä a,c , Tarja Melartin a,d , Ulla Leskelä a,c , Paula Lestelä-Mielonen a,c , Erkki Isometsä a,d,⁎ a

c

Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki, Finland b Department of Mathematics and Statistics, University of Helsinki, Helsinki, Finland Department of Psychiatry, Helsinki University Central Hospital (HUCH), Peijas Hospital, Vantaa, Finland d Department of Psychiatry, Helsinki University Central Hospital (HUCH), Helsinki, Finland Received 28 March 2006; received in revised form 21 April 2006; accepted 24 April 2006 Available online 15 June 2006

Abstract Background: Suicidal ideation is likely to represent a phase preceding suicidal acts among most suicidal patients with major depressive disorder (MDD). Factors predicting reversal of the suicidal process are unknown. Our aim was to test the hypothesis that a decline in suicidal ideation is preceded by a decline in hopelessness among patients with MDD. Method: Of the 269 Vantaa Depression Study patients with DSM-IV MDD, 103 patients scored ≥ 6 points at baseline on the Scale for Suicidal Ideation (SSI). Seventy of these patients were followed-up weekly either until they scored zero points on the SSI, or up to 26 weeks. Results: The median duration for a decline of suicidal ideation to zero was 2.2 months after baseline. The level of baseline suicidal ideation, depressive symptoms, and the presence of any personality disorder predicted duration of suicidal ideation. A decline in both depression (BDI) and hopelessness (HS) independently predicted a decline in suicidal ideation. Limitations: Due to study design, we do not know if suicidal ideation relapsed after the first time the patient reached zero score in the SSI. Conclusions: Among patients with major depressive disorder having suicidal ideation, the decline in suicidal ideation is independently predicted by preceding declines in the levels of both depressive symptoms as well as hopelessness. The findings are consistent with possible causal roles of declines in depression and hopelessness in reversing the suicidal process. © 2006 Elsevier B.V. All rights reserved. Keywords: Depression; Suicidal ideation; Hopelessness

1. Introduction

⁎ Corresponding author. Department of Mental Health and Alcohol Research, National Public Health Institute, Mannerheimintie 166, FIN00300 Helsinki, Finland. E-mail address: [email protected] (E. Isometsä). 0165-0327/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2006.04.028

Suicidal ideation is likely to represent a phase preceding suicidal acts among most patients with major depressive disorder (MDD). In a psychological autopsy study, more than half of the subjects completing suicide during major depression had communicated their intent during the final 3 months (Isometsä et al.,

96

P. Sokero et al. / Journal of Affective Disorders 95 (2006) 95–102

1994), and almost all (95%) attempting suicide had reported suicidal ideation (Sokero et al., 2003). Thus, suicidal ideation appears to precede the decision to act on suicidal impulses among patients with MDD. However, only a minority of those with suicidal ideation actually attempts suicide, and it remains unclear why for most subjects suicidal ideation actually resolves. The prevalence of suicidal ideation in patients with MDD ranges from 47% to 69% (Asnis et al., 1993; Bronisch and Wittchen, 1994; Sokero et al., 2003). Risk factors identified for suicidal ideation in depression include severity of depression (Zisook et al., 1994; Van Gastel et al., 1997; Pages et al., 1997; Alexopoulos et al., 1999), comorbid personality disorders (Van Gastel et al., 1997), comorbid alcohol dependence or abuse (Pages et al., 1997; Cornelius et al., 1995), comorbid anxiety disorder (Schaffer et al., 2000), female gender (Pages et al., 1997; Schaffer et al., 2000), age (Lynch et al., 1999), unemployment (Pages et al., 1997), life events (Monroe et al., 2001), poor social support (Alexopoulos et al., 1999), hopelessness (Van Gastel et al., 1997; Pages et al., 1997; Rudd, 1990), past suicide attempt (Alexopoulos et al., 1999), psychomotor agitation and, perhaps also depressed mixed state (DMX) as an indicator of a possibly unrecognized bipolar disorder (Akiskal et al., 2005). Two meta-analyses verify that emergent suicidal ideation is not more likely on antidepressant than placebo after initiation of antidepressants (Beasley et al., 1991; Montgomery et al., 1995), and two recent studies have documented the actual effectiveness of treatments for depression in reducing suicidal ideation among elderly patients with depression (Szanto et al., 2003; Bruce et al., 2004). Nevertheless, the factors causing a decline in suicidal ideation, and thus, reversal of the suicidal process, are still largely unknown. In this study, we investigated prospectively the short-term course of suicidal ideation among psychiatric patients with MDD. The temporal relationships between suicidal ideation and depressive symptoms, level of hopelessness, and level of anxiety symptoms were examined weekly. First, we investigated the duration of suicidal ideation and its main determinants. Second, we hypothesized that a decline in the level of symptoms, specifically in hopelessness, would be the main determinant for a decline in suicidal ideation. 2. Methods 2.1. Setting The background and methodology of the Vantaa Depression Study (VDS) have been described in detail

elsewhere (Melartin et al., 2002, 2004). In brief, the VDS is a collaborative depression research project between the Department of Mental Health and Alcohol Research of the National Public Health Institute, Helsinki, Finland, and the Department of Psychiatry of the Peijas Medical Care District (PMCD), Vantaa, Finland. Vantaa is the fourth largest city in Finland, with a population of 169 000 in 1997, and the PMCD provides free-of-charge psychiatric services to all of its citizens. The study protocol of the VDS was approved by the ethics committee of the PMCD in December 1996. 2.2. Screening, diagnostic evaluation, and baseline measurements In the first phase, all patients (n = 806) at the Department of Psychiatry of the PMCD were screened for a possible new episode of DSM-IV MDD (APA, 1994) between February 1st, 1997 and May 31st, 1998 (Melartin et al., 2002). Patients with a positive finding were fully informed about the study project, and their participation was requested. Of the 703 eligible patients, 542 (77%) gave their written informed consent. In the second phase, a researcher using the WHO SCAN 2.0 (Wing et al., 1990) interviewed the 542 consenting patients, 269 of whom were subsequently diagnosed with DSM-IV MDD and included in the study. The diagnostic reliability for MDD has been found to be excellent (κ = 0.86 [95% CI = 0.58–1.0]) (Melartin et al., 2002). The Structured Clinical Interview for DSM-III-R personality disorders (SCID-II) (Spitzer et al., 1989) was used to assess Axis II diagnoses. The cohort baseline measurements included the 17-item Hamilton Depression Rating Scale (HAMD) (Hamilton, 1960), the 21-item Beck Depression Inventory (BDI) (Beck et al., 1961), Beck Anxiety Inventory (BAI) (Beck et al., 1988), Beck Hopelessness Scale (HS) (Beck et al., 1974), Scale for Suicidal Ideation (SSI) (Beck et al., 1979), Social and Occupational Functioning Assessment Scale of DSM-IV (SOFAS) (Goldman et al., 1992), Interview for Recent Life Events (IRLE) (Paykel, 1983), Interview Measure of Social Relationships (IMSR) (Brugha et al., 1987), and Perceived Social Support Scale-Revised (PSSS-R) (Blumenthal et al., 1987). Current suicidal ideation was first examined using the SSI. SSI is a 19-item observer scale designed to quantify the intensity of current conscious suicide ideation in various dimensions of self-destructive thoughts or wishes; e.g. the extent of the wish to

P. Sokero et al. / Journal of Affective Disorders 95 (2006) 95–102

die, the desire to make an actual suicide attempt, and details of any plans. Internal deterrents to an active attempt and subjective feelings of control and/or “courage” regarding a proposed attempt are also quantified. Each item consists of three alternative statements graded in intensity from 0 to 2, with the maximum total score being 38. Here, moderate to severe suicidal ideation refers to patients scoring ≥ 6 in the SSI, while no ideation refers to a score of zero. Overall, 103 (38%) of the 269 patients, reported suicidal ideation according to SSI during the current episode. Risk factors for suicidal behaviour, suicidal ideation, and suicide attempt have been reported elsewhere (Sokero et al., 2003). 2.3. Prospective follow-up of depression with a life chart Of the 269 subjects with current MDD initially included in the study, 198 were still alive at the end of the study period, had remained unipolar, and could be followed-up (Melartin et al., 2004), 8 patients (3%) died during the 18 moths after baseline, three (1%) of them by suicide (Sokero et al., 2005). At baseline, the majority (154/198, [78%]) had been taking antidepressants in normal adult doses. Patients whose diagnosis was amended to bipolar disorder during the follow-up (13/269, [5%]) were analysed separately. The outcome of MDD and comorbid disorders was investigated at 6 and 18 months by repeated SCAN 2.0 and SCID-II interviews, observer- and self-report scales, and medical and psychiatric records. A detailed life chart was created, with time after baseline divided into three classes: (a) state of full remission (0/9 criteria symptoms for major depressive episode), (b) partial remission (1– 4/9 symptoms), and (c) major depressive episode (5+/9 symptoms). We used two alternative definitions for duration of the index episode: the uninterrupted duration of the episode in the state of major depressive episode – (1) time with full MDE criteria, and time to the first onset of state of full remission that lasted at least 2 consecutive months – (2) time to full remission (Melartin et al., 2004). 2.4. Weekly follow-up of suicidal ideation and covariates The VDS is a research and development project aimed at promoting educational efforts to enhance scientific knowledge and ensure optimal clinical treatment of depressive disorders. A comprehensive evaluation of patients' suicidality was carried out on a

97

weekly basis until suicidal ideation resolved. Seventy of the 103 patients with current suicidal ideation at baseline were followed-up weekly; however, due to poor adherence 33 patients could not be followed-up. The SSI, HS, BAI, and BDI scores were measured weekly. In order to avoid circularity, we omitted the suicidality items of the BDI. We initially planned that the weekly observation could be discontinued upon a patient receiving two consecutive scores of zero in the SSI. However, this goal was perhaps overly optimistic, and thus, we subsequently decided to analyze the weekly observations after the first score of zero in the SSI. All 70 patients were followed-up from baseline to at least two observations, with the maximum follow-up time being 26 weeks. For 47 patients suicidal ideation resolved, 8 patients dropped out, and 15 patients were followed up for the maximum period. The 70 patients who were successfully followed up, as compared with those 33 who did not participate in the weekly followup, had higher level of psychopathology, more anxiety disorders (46[66%] vs. 13[39%], χ2 = 6.349, df = 1, p = .018), more cluster B personality disorders (19 [27%] vs. 2[6%], χ2 = 6.141, df = 1, p = .017), including borderline personality disorder (16[23%] vs. 2[6%], χ2 = 4.387, df = 1, p = .05), higher level of hopelessness (12.6 ± 4.7 vs. 10.5 ± 4.8, F = 4.083, df = 1, p = .046). The 8 patients, who dropped out from the weekly follow-up, were more often in-patients at the baseline (5[63%] vs. 16[26%], χ2 = 4.543, df = 1, p = .05), had higher level of depression (HAM-D) (25.8 ± 7.4 vs. 20.9 ± 6.1, df = 1, F = 4.414, p = .04) and anxiety (BAI) (33.5 ± 7.2 vs. 23.5 ± 10.2, df = 1, F = 7.219, p = .009). The overall demographic characteristics are presented in Table 1. 2.5. Statistical analysis The decline of suicidal ideation during the followup was studied with survival methods by defining the outcome as the first time when two consecutive zero measurements of SSI were found. The overall decline is displayed with the Kaplan–Meier survival curve. Cox's proportional hazard models with time-varying covariates were used to study the influence of reaching threshold levels in hopelessness, depression or anxiety scores prior to the decline of suicidal ideation while adjusting for the initial scores of suicidal ideation. For each measure, the appropriate threshold level was defined separately. The timevarying covariates representing decline in hopelessness, depression or anxiety scores, were given the value ‘one’ if the corresponding threshold level

98

P. Sokero et al. / Journal of Affective Disorders 95 (2006) 95–102

Table 1 Characteristics of the suicidal patients participating in weekly followup (n = 70) Characteristic Sex Males Females Age (years), mean ± S.D. Marital status Married or co-habiting Treatment setting In-patients Psychiatric comorbidity Alcohol dependence/abuse Anxiety disorder (any) Personality disorder (any) Cluster A Cluster B Cluster C Comorbid disorder (any) Suicidality SI in the history SA in the history SA during the index episode Status variables at baseline HAM-D score, mean ± S.D. BDI score, mean ± S.D. BAI score, mean ± S.D. HS score, mean ± S.D. PSSS-R score, mean ± S.D. SOFAS score, mean ± S.D. SSI score, mean ± S.D.

n (%) 26 (37) 44 (63) 38.9 ± 10.1 31 (44) 21 (30) 24 (34) 46 (66) 36 (51) 20 (29) 19 (27) 26 (37) 57 (81) 39 (56) 23 (33) 18 (26) 21.4 ± 6.4 27.2 ± 6.4 24.7 ± 10.3 12.6 ± 4.7 34.5 ± 12.7 48.8 ± 10.9 15.6 ± 5.5

BAI = Beck Anxiety Inventory, BDI = Beck Depression Inventory, HAM-D = Hamilton rating scale for Depression, HS = Beck Hopelessness Scale, MDD = major depressive disorder, PSSS-R = Perceived Social Support Scale-Revised, SA = suicide attempt, SI = suicidal ideation, SOFAS = Social and Occupational Functioning Assessment Scale, SSI = Scale for Suicidal Ideation. Item 3 from HAM-D and item 9 from BDI were excluded from the analyses in order to avoid circularity.

(BDI < 10, BAI < 10, HS < 9) was reached for the first time, and ‘zero’ before that. Sensitivity of the results was investigated by varying the chosen threshold values. The plausibility of the proportional hazards assumption was checked by plotting the logarithms of the cumulative baseline hazards against the followup time in appropriate comparison groups, as well as with residual analyses of the models. Since the data were collected mostly from outpatients, there were some missing appointments and therefore, missing weekly measurements. The proportions of missing values were on average 32% for SSI, 21% for HS, 20% for BAI and 30% for BDI. Therefore, a more robust measure than the weekly scores of the covariates was needed in the analyses of decline. SPSS software, version 11.0 (SPSS Inc., 1989–2001),

Fig. 1. Decline in proportion of cases with suicidal ideation by the Kaplan–Meier survival curve.

and the software Stata (StataCorp LP) were used for the estimations. 3. Results 3.1. Course of suicidal ideation The overall level of psychopathology of the patients varied from moderate to high. Suicidal ideation resolved in 47 (67%) patients and in 15 (21%) it persisted for the whole follow-up period. 8 patients (11%) dropped out. The decline in the proportion of cases with suicidal ideation is presented with the Kaplan–Meier survival curve in Fig. 1 and in Table 2, in this analysis 50% of the population reached zero in 2.2 months (9.6 weeks). Among those reaching zero level of the suicidal ideation, the median time for this was 1.6 months (6.8 weeks). For those patients having both the weekly follow-up (duration of suicidal ideation), and the life chart (time with full MDE criteria and time to full remission) measures available (N = 53), the median time for decline to zero of suicidal ideation was 2.7 months, the median time spent with full MDD criteria was 2.6 months and the median time to reach full remission was 4.2 months, Table 2 Weekly distribution of SSI scores during follow-up SSI median score

Baseline Week 1 Week 2 Week 3 Week 4 Week 5

N

25%

50%

75%

70 70 70 70 66 57

11.0 4.0 0.0 0.0 0.0 0.0

15.0 11.0 9.0 6.5 7.0 7.0

20.3 18.3 17.3 16.0 16.0 14.5

P. Sokero et al. / Journal of Affective Disorders 95 (2006) 95–102 Table 3 Cox regression model for the duration of suicidal ideation adjusted for baseline variables Variable

HR

95% CI

p

Age, years HS baseline score BAI baseline score BDI baseline score SSI baseline score Personality disorder (any)

0.99 1.02 1.04 0.93 0.88 0.28

0.95–1.03 0.93–1.11 0.99–1.08 0.87–1.00 0.81–0.97 0.11–0.72

.75 .70 .11 .04 .009 .008

99

having or not having an antidepressant as a predictor variable, but this did not significantly influence the findings. 3.3. Predictors for decline in suicidal ideation In separate analyses, decline in hopelessness, depressive symptoms and anxiety were each significant predictors for the decline of suicidal ideation. In all analyses, adjusting for the initial level of suicidal ideation showed that the decline depends significantly on the severity of the symptoms; the higher the initial level, the longer the duration. The importance of the initial level of suicidal ideation was stable in all separate analyses. When analysing the influence of decline in hopelessness, depressive symptoms and anxiety symptoms jointly on the decline of suicidal ideation, the apparent separate effect of anxiety turned out to be nonsignificant (Table 4), whereas declines both in depression and hopelessness had an independent effect on the decline of suicidal ideation. This may be due to the fact that correlation between subsequent scores in hopelessness and anxiety was as high as 0.8 (for other measures approximately 0.4). Similar analyses of decline as for SSI were conducted by treating HS, BAI and BDI as outcomes (analyses not shown here). They revealed that the decline of hopelessness was more rapid than that of anxiety. This is likely to be the reason why it took over the effect of a later decline in anxiety.

HR = Hazard ratio, HS = Beck Hopelessness Scale, BAI = Beck Anxiety Inventory, BDI = Beck Depression Inventory, SSI = The Scale for Suicide Ideation.

respectively. Specifically, the difference in the median time for decline of suicidal ideation was not statistically significant between patients with cluster B personality disorders and those without it (1.7 months [7.4 weeks] vs. 1.6 months [6.7 weeks], p = .59, log rank test). 3.2. Baseline factors predicting the duration of suicidal ideation We used the Cox regression model to study the effect of risk factors predicting the duration of suicidal ideation, and used sex, age, HS baseline score, BAI baseline score, BDI baseline score, SSI baseline score, PSSS-R baseline score, SOFAS baseline score, marital status, presence of alcohol dependency or abuse, presence of any anxiety disorder, presence of suicidal ideation or suicide attempts in the history and presence of any personality disorder as predictor variables. Of these baseline factors, the level of suicidal ideation, depressive symptoms and presence of any personality disorder each predicted longer duration of suicidal ideation (Table 3). If variable “any personality disorder” was replaced with some other personality disorder diagnosis, the significance was lost. We also created an alternative model including

4. Discussion Suicidal ideation resolved in the majority of the suicidal MDD patients during the first 2 to 3 months. Our hypothesis was that a decline in hopelessness would be the main determinant for the decline of suicidal ideation. However, in the joint analyses the declines both in depression and hopelessness independently predicted the following decline in suicidal

Table 4 Cox proportional hazard models for the decline of suicidal ideation adjusted separately for the normalization of depressive symptoms (BDI), anxiety (BAI), and hopelessness (HS) and jointly for all Variable

Separate models

Joint model

Depression

SSI baseline BDI BAI HS

Anxiety

Hopelessness

HR

95% CI

p

HR

95% CI

0.91 7.68

0.84–0.97 3.73–15.85

.008 <.001

0.92

0.86–0.99

4.70

1.99–11.09

p .023

HR

95% CI

p

0.92

0.86–0.97

.02

5.90

2.24–15.52

<.001

<.001

HR

95% CI

p

0.90 5.74 1.62 3.51

0.84–0.97 2.69–12.25 0.59–4.48 1.15–10.73

.004 <.001 .34 .03

SSI = Scale for Suicide Ideation, BDI = Beck Depression Inventory, BAI = Beck Anxiety Inventory, HS = Beck Hopelessness Scale, HR = Hazard ratio.

100

P. Sokero et al. / Journal of Affective Disorders 95 (2006) 95–102

ideation. In our view, this finding is consistent with the interpretation that they both could have a causal role in reversal of the suicidal process. The duration of suicidal ideation was longer for patients with higher level of psychopathology, such as initially high level of suicidal ideation or depressive symptoms at baseline, or some personality disorder. 4.1. Strengths, limitations and generalizability The present study has some major methodological strengths. To our knowledge, this is the first study to have employed a specific scale to measure suicidal ideation (SSI) prospectively on weekly basis following up in a sample of adult patients with MDD and the first study to investigate factors preceding decline in suicidal ideation. The VDS involves a relatively large (N = 269) cohort of both out- and inpatients with MDD, from which the group of suicidal patients (N = 103) was screened out and followed-up. The patients were carefully diagnosed using structured interviews with excellent reliability (κ = 0.86) for the diagnosis of MDD (Melartin et al., 2002). In addition, information on all comorbid Axes I and II disorders at baseline and later interviews were used. Methodological details are discussed in earlier reports (Sokero et al., 2003; Melartin et al., 2002, 2004). We used a pre-determined cut-off point (SSI ≥ 6) to define moderate to severe current suicidal ideation. In retrospect, this may have been somewhat high (Beck et al., 1999). In addition, although the internal consistency of SSI was high (Cronbach's α = 0.85–0.90), its inter-rater reliability remains unknown. 70 suicidal patients (68%) could be followed-up on weekly basis. These patients had an overall higher level of psychopathology than the VDS cohort overall, or those suicidal cases who did not participate in the weekly follow-up. It is unlikely that the findings with those 33 suicidal patients included would have been markedly different. If anything, we assume that the median time for duration of suicidal ideation could have been shorter. Suicidal depressive patients commonly differ more from the non-suicidal in their subjective than objective measures on depression. (Malone et al., 1995; Van Praag and Plutchik, 1984; Cornelius et al., 1995; Oquendo et al., 1999). In our sample there appeared to be a similar trend. Although our findings were highly significant, type II error needs to be considered, regarding some predictors of suicidal ideation. In order to reduce complexity, we deliberately focused on the first time the patient reached zero score in the SSI. We do not know if suicidal ideation relapsed after that. It is to be noted, that our findings are also dependent on the

chosen threshold levels for depressive and anxiety symptoms and hopelessness. However, the findings are robust and according to the sensitivity analyses conducted, would not be markedly different with other threshold levels. Finally, because the study was conducted in a secondary level psychiatric setting, it was inevitable that there were some missed appointments. It is impossible to exclude the possibility that during some of these missed appointments, level of SSI could have already reached zero but then relapsed later. This could have led us to somewhat overestimate the uninterrupted duration of suicidal ideation, but is unlikely to cause other biases. 4.2. Decline of suicidal ideation The duration of decline in suicidal ideation is strongly associated with the initial level of symptoms; the higher the initial level, the longer the duration. Personality disorders overall had also a significant impact on the duration of suicidal ideation. Contrary to our expectations, this was more related to overall, rather than specifically to cluster B or borderline personality disorder. Suicidal ideation appears to resolve gradually after depressive symptoms and hopelessness have started to alleviate. The duration of ideation approximately corresponds the time the patients fulfil the criteria for a major depressive episode. Thus, even some decrease in the level of depression seems to be enough to initiate the decline in the intensity of suicidal ideation. This is consistent with earlier findings (Szanto et al., 2003; Bruce et al., 2004) among elderly depressives about the impact of treatment interventions to suicidal ideation. There has been an ongoing debate on antidepressants and suicidal behaviour (Healy, 2003; Casey, 2004; Painuly and Basu, 2004); most of the patients in our sample were on antidepressants, and having or not having them had no significant influence on the duration of suicidal ideation. To our knowledge, this is the first study investigating factors explaining the reversal of suicidal ideation. We found that the decline of suicidal ideation is strongly associated with the preceding decline of depressive symptoms, level of hopelessness and anxiety. Hopelessness, severity of depression and anxiety are all identified risk factors for suicidal behaviour. We hypothesized that a decline in the level of hopelessness would be the main determinant for a decline in suicidal ideation. Hopelessness as it occurs in depressed patients may be viewed as having both state and trait characteristics. During depression, hopelessness escalates and then subsides along the course of illness (Beck

P. Sokero et al. / Journal of Affective Disorders 95 (2006) 95–102

et al., 1990). Cognitive research on suicide and risk prediction has developed a model of suicidal behaviour in which hopelessness is a key psychological variable (Beck and Weishaar, 1990). In our joint analysis, the decline of depression and hopelessness were the significant independent risk factors. We examined which of these three factors appear to have a plausible causal relationship to decline in suicidal ideation. The criteria for causality, as reviewed by Rothman and Greenland (1998) are strength of association, consistency, specificity, temporality (the most important one), biologic gradient, plausibility, coherence, experimental evidence and analogy. Both depressive symptoms and hopelessness have a strong and consistent association with suicidal ideation (Sokero et al., 2003; Van Gastel et al., 1997; Pages et al., 1997; Malone et al., 2000), and they are plausible and theoretically coherent risk factors for suicidal behaviour. Experimental evidence from the two reports (Szanto et al., 2003; Bruce et al., 2004) on effectiveness of treatments for depression to alleviate also suicidal ideation at least among elderly depressives exists, although these studies cannot inform whether this alleviation is related to depressive symptoms per se, or the role of underlying hopelessness. To our knowledge, our findings provide first information on the temporal course, which is crucial when estimating potential causal role. Overall, our findings are consistent with the interpretation that declines in both depression and hopelessness could have a causal role in reversing the suicidal process. However, we can never exclude the possibility of existence of other possible, perhaps even more primary causal factors we were not able to measure. Future studies may confirm and further clarify (or falsify) our working hypothesis by prospectively following suicidal ideation plus both depressive symptoms as well as hopelessness, and investigating the effectiveness of treatments for depression, or perhaps psychotherapies specifically targeted at hopelessness, in reversing suicidal ideation among adult patients with MDD. References Akiskal, H.S., Benazzi, F., Perugi, G., Rihmer, Z., 2005. Agitated “unipolar” depression re-conceptualized as a depressive mixed state: implications fort he antidepressant-suicide controversy. JAD 85, 245–258. Alexopoulos, G.S., Bruce, M.L., Hull, J., Sirey, J.A., Kakuma, T., 1999. Clinical determinants of suicidal ideation and behavior in geriatric depression. Archives of General Psychiatry 56, 1048–1053. American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders, 4th edn. APA, Washington, DC. DSM-IV.

101

Asnis, G.M., Friedman, T.A., Sanderson, W.C., Kaplan, M.L., van Praag, H., Harkavy-Freidman, J.M., 1993. Suicidal behaviors in adult psychiatric outpatients: I. Description and prevalence. American Journal of Psychiatry 150, 108–112. Beasley, C.M., Dornseif, B.E., Bosomworth, J.C., Sayler, M.E., Rampey Jr., A.H., Heiligenstein, J.H., Thompson, V.L., Murphy, D.J., Masica, D.N., 1991. Fluoxetine and suicide: a metaanalysis of controlled trials of treatment for depression. BMJ 303, 685–692. Beck, A.T., Weishaar, M.E., 1990. Suicide risk assessment and prediction. Crises 22, 22–30. Beck, A.T., Ward, C.H., Mendelson, M., Moch, J.E., Erbauch, J.K., 1961. An inventory for measuring depression. Archives of General Psychiatry 4, 561–571. Beck, A.T., Weismann, A., Lester, D., Trexler, L., 1974. The measure of pessimism: the hopelessness scale. Journal of Consulting and Clinical Psychology 42, 861–865. Beck, A.T., Kovacs, M., Weissman, A., 1979. Assessment of suicidal intention: a scale for suicidal ideation. Journal of Consulting and Clinical Psychology 47, 343–352. Beck, A.T., Brown, G., Epstein, N., Steer, R.A., 1988. An inventory for measuring clinical anxiety: psychometric properties. Journal of Consulting and Clinical Psychology 56, 893–897. Beck, A.T., Brown, G., Berchick, R.J., Stewart, B.L., Steer, R.A., 1990. Relationship between hopelessness and ultimate suicide: a replication with psychiatric outpatients. American Journal of Psychiatry 147, 190–195. Beck, A.T., Brown, G.K., Steer, R.A., Dahlsgaard, K.K., Grisham, J.K., 1999. Suicide ideation at its worst point: a predictor of eventual suicide in psychiatric outpatients. Suicide & Life-Threatening Behavior 29, 1–9. Blumenthal, J.A., Burg, M.M., Barefoot, J., Williams, R.B., Haney, T., Zimet, G., 1987. Social support, type A behaviour, and coronary artery disease. Psychosomatic Medicine 49, 331–340. Bronisch, T., Wittchen, H.-U., 1994. Suicidal ideation and suicide attempts: comorbidity with depression, anxiety disorders, and substance abuse disorder. European Archives of Psychiatry and Clinical Neuroscience 244, 93–98. Bruce, M.L., Ten Have, T.R., Reynolds, C.H., Katz, I.I., Schulberg, H. C., Mulsant, B.H., Brown, G.K., McAvay, G.J., Pearson, J.L., Alexopoulos, G.S., 2004. Reducing suicidal ideation and depressive symptoms in depressed older primary care patients. A randomized controlled trial. Journal of the American Medical Association 291, 1081–1091. Brugha, T.S., Sturt, E., MacCarthy, B., Potter, J., Wykes, T., Bebbington, P.E., 1987. The interview measure of social relationships: the description and evaluation of a survey instrument for assessing personal social resources. Social Psychiatry 22, 123–128. Casey, P., 2004. SSRI and suicide. Psychotherapy and Psychosomatics 73 (4), 259–260. Cornelius, J.R., Salloum, I.M., Mezzich, J., Cornelius, M.D., Fabrega, H., Ehler, J.G., Ulrich, R.F., Thase, M.S., Mann, J.J., 1995. Disproportionate suicidality in patients with comorbid major depression and alcoholism. American Journal of Psychiatry 152, 358–364. Goldman, H.H., Skodol, A.E., Lave, T.R., 1992. Revising Axis V for DSM-IV: a review of measures of social functioning. American Journal of Psychiatry 149, 1148–1156. Hamilton, M., 1960. A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry 23, 56–62.

102

P. Sokero et al. / Journal of Affective Disorders 95 (2006) 95–102

Healy, D., 2003. Lines of evidence on the risks of suicide with selective serotonin reuptake inhibitors. Psychotherapy and Psychosomatics 72 (2), 71–79. Isometsä, E.T., Henriksson, M.M., Aro, H.M., Heikkinen, M.E., Kuoppasalmi, K.I., Lönnqvist, J.K., 1994. Suicide in major depression. American Journal of Psychiatry 151, 530–536. Lynch, T.R., Johnson, C.S., Mendelson, T., Robins, C.J., Krishnan, K. R.K., Blazer, D.G., 1999. New onset and remission of suicidal ideation among a depressed adult sample. Journal of Affective Disorders 56, 49–54. Malone, K.M., Haas, G.L., Sweeney, J.A., Mann, J.J., 1995. Major depression and the risk of attempted suicide. Journal of Affective Disorders 34, 174–185. Malone, K.M., Oquendo, M.A., Haas, G.L., Ellis, S.P., Li, S., Mann, J.J., 2000. Protective factors against suicidal acts in major depression: reasons for living. American Journal of Psychiatry 157, 1084–1088. Melartin, T.K., Rytsälä, H.J., Leskelä, U.S., Lestelä-Mielonen, P.S., Sokero, T.P., Isometsä, E.T., 2002. Current comorbidity of DSM IV major depressive disorder in psychiatric care – Vantaa Depression Study. Journal of Clinical Psychiatry 63, 126–135. Melartin, T.K., Rytsälä, H.J., Leskelä, U.S., Lestelä-Mielonen, P.S., Sokero, T.P., Isometsä, E.T., 2004. Severity and comorbidity predict episode duration and recurrence of DSM-IV major depressive disorder. Journal of Clinical Psychiatry 65, 810–819. Monroe, S.C., Harkness, K., Simons, A.D., Thase, M.E., 2001. Life stress and the symptoms of major depression. Journal of Nervous and Mental Diseases 180, 168–175. Montgomery, S.A., Dunner, D.L., Dunbar, G.C., 1995. Reduction of suicidal thoughts with paroxetine in comparison with reference antidepressants and placebo. European Neuropsychopharmacology 5, 5–13. Oquendo, M.A., Malone, K.M., Ellis, S.P., Sackheim, H.A., Mann, J.J., 1999. Inadequacy of antidepressant treatment for patients with major depression who are at risk for suicidal behavior. American Journal of Psychiatry 156, 190–194. Pages, K.P., Russo, J.E., Roy-Byrne, P.P., Ries, R.K., Cowley, D.S., 1997. Determinants of suicidal ideation: the role of substance use disorders. Journal of Clinical Psychiatry 58, 510–515. Painuly, N., Basu, D., 2004. SSRI and suicide? Psychotherapy and Psychosomatics 73 (4), 261–262. Paykel, E.S., 1983. Methodological aspects of life event research. Journal of Psychosomatic Research 27, 341–352.

Rothman, K.J., Greenland, S., 1998. Causation and causal inference. In: Rothman, K.J., Greenland, S. (Eds.), Modern Epidemiology. Lippincott-Raven Publishers, Philadelphia, pp. 7–28. Rudd, M.D., 1990. An integrative model of suicidal ideation. Suicide & Life-Threatening Behavior 20 (1), 16–30. Schaffer, A., Levitt, A.J., Bagby, R.M., Kennedy, S.H., Levitan, R.D., Joffe, R.T., 2000. Suicidal ideation in major depression: sex differences and impact of comorbid anxiety. Canadian Journal of Psychiatry 45, 822–826. Sokero, T.P., Melartin, T.K., Rytsälä, H.J., Leskelä, U.S., LesteläMielonen, P.S., Isometsä, E.T., 2003. Suicidal ideation and attempts among psychiatric patients with major depressive disorder. Journal of Clinical Psychiatry 64, 1094–1100. Sokero, T.P., Melartin, T.K., Rytsälä, H.J., Leskelä, U.S., LesteläMielonen, P.S., Isometsä, E.T., 2005. Prospective study of risk factors for attempted suicide among patients with DSM-IV major depressive disorder. British Journal of Psychiatry 186, 314–318. Spitzer, R., Williams, J.B.W., Gibbon, M., First, M.B., 1989. Structured Clinical Interview for DSM-III-R Personality Disorder (SCID-II). Biometrics Research Department, New York State Psychiatric Institute, New York. SPSS Inc., 1989–2001. Release 11.0.1. Statistical Package for the Social Sciences for Windows. Chicago, IL, USA. StataCorp LP. Release 9. Statistical Software for Professionals. College Station, Texas, USA. Szanto, K., Mulsant, B.H., Houck, P., Dew, M.A., Reynolds, C.H., 2003. Occurrence and course of suicidality during sort-term treatment of late-life depression. Archives of General Psychiatry 60, 610–617. Van Gastel, A., Schotte, C., Maes, M., 1997. The prediction of suicidal intent on depressed patients. Acta Psychiatrica Scandinavica 96, 254–259. Van Praag, H.M., Plutchik, R., 1984. Depression type and depression severity in relation to risk of violent suicide attempt. Psychiatry Research 12, 333–338. Wing, J.K., Babor, T., Brugha, T., Burke, J., Cooper, J.E., Giel, R., Jablenski, A., Reiger, D., Sartorius, N., 1990. SCAN. Schedules for clinical assessment in neuropsychiatry. Archives of General Psychiatry 47, 589–593. Zisook, S., Arleen, G., Sledge, P., Schuchter, S.R., 1994. Reported suicidal behavior and current suicidal ideation in a psychiatric outpatient clinic. Annals of Clinical Psychiatry 6 (1), 27–31.