Journal of Psychiatric Research 47 (2013) 1925e1931
Contents lists available at ScienceDirect
Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/psychires
Electrodermal hyporeactivity as a trait marker for suicidal propensity in uni- and bipolar depression L.H. Thorell a, *, M. Wolfersdorf b, R. Straub c, J. Steyer c, S. Hodgkinson c, W.P. Kaschka c, M. Jandl c a b c
Department of clinical and experimental medicine, Linköping University and Emotra AB, Gothenburg, Östgötagatan 60B, 582 32 Linköping, Sweden Department of Psychiatry and Psychotherapy, State Hospital, Bayreuth, Germany Department of Psychiatry and Psychotherapy I, University Hospital Ulm, ZfP (Zentrum für Psychiatrie)-Südwürttemberg, Ravensburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 April 2013 Received in revised form 30 July 2013 Accepted 26 August 2013
Background: A meta-analysis of studies investigating electrodermal activity in depressed patients, suggested that electrodermal hyporeactivity is sensitive and specific for suicide. Aims: To confirm this finding and to study electrodermal hyporeactivity relative to type and severity of depression, trait anxiety, its stability and independence of depressive state. Method: Depressed inpatients (n ¼ 783) were tested for habituation of electrodermal responses and clinically assessed using the Beck Depression Inventory (BDI) and the STAI-Trait scale for trait anxiety. Results: The high sensitivity and raw specificity of electrodermal hyporeactivity for suicide were confirmed. Its prevalence was highest in bipolar disorders and was independent of severity of depression, trait anxiety, gender and age. Hyporeactivity was stable, while reactivity changed into hyporeactivity in a later depressive episode. Conclusions: The findings support the hypothesis that electrodermal hyporeactivity is a trait marker for suicidal propensity in depression. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Suicidal behaviour Depression Electrodermal hyporeactivity Sensitivity Specificity Stability
Low electrodermal activity in depressed patients was first described more than 100 years ago (Vigouroux, 1890). Since then, more than fifty studies have confirmed this original observation. In particular, electrodermal hyporeactivity has repeatedly been found in psychotic depressed patients. Edman and co-workers found that patients with violent suicide attempts exhibited faster habituation of the electrodermal response to repeated neutral tones than patients with non-violent attempts (Edman et al., 1986). Thorell showed that a history of suicide attempts was significantly related to electrodermal hyporeactivity (Thorell, 1987). In their Weissenau Psychiatric Hospital study, Wolfersdorf and co-workers tested a large sample of depressed inpatients using a similar approach (Straub, 1988; Straub et al., 2003; Keller et al., 1991; Wolfersdorf et al., 1993, 1999). Recently, Jandl and co-workers published the first study on the habituation of the auditory event related potential Novelty P3 and electrodermal response to the same acoustic stimuli in relation to suicidal behaviour in depressed patients (Jandl et al., 2010). Novelty P3 displayed faster habituation in patients with major depression
* Corresponding author. Tel.: þ46 703 26 38 53. E-mail address:
[email protected] (L.H. Thorell). 0022-3956/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jpsychires.2013.08.017
who had a history of severe (i.e. intent to die and intensive care treatment) suicide attempt than in those who had not. Using data drawn from a number of studies, a meta-analysis displayed that the sensitivity and specificity of electrodermal hyporeactivity for suicide in depressed patients was 97% and 93%, respectively (Thorell, 2009). It was also found that depressed patients with very high suicidal propensity, i.e. those who later committed suicide, had higher prevalence of hyporeactivity than those with low suicidal propensity. The present study was aimed to analyse the whole Weissenau sample of depressed inpatients for prevalence of suicidal behaviour and its association with electrodermal hyporeactivity, to study the stability of electrodermal hyporeactivity over episodes of depression and its relationships to severity of depression, trait anxiety, gender, and age. 1. Method Data from a total of 892 patients who were treated on the Depression Ward of the Centre for Psychiatry, Weissenau in southern Germany between 1985 and 2002 was analysed. Duplicate data (n ¼ 40) and data on patients not tested for electrodermal reactivity (n ¼ 51) was removed from this dataset. Where patients
1926
L.H. Thorell et al. / Journal of Psychiatric Research 47 (2013) 1925e1931
were assessed during more than one depressive episode, information from the latest episode was used. These patients (n ¼ 26) were selected for an analysis of the stability of electrodermal reactivity over episodes. The diagnostic categories used in the study are those outlined in ICD-9 (World Health Organization, 1978) or ICD-10 (World Health Organisation, 1992). Patients with a primary diagnosis other than depression (n ¼ 18) were excluded. Thus, 783 patients were available for further analysis. Three age groups were defined: (a) 18e35 years, (b) 36e49 years and (c) 50e66 years. 1.1. Ethics All patients gave their informed consent to participate in the study. The research was approved by the local ethical committee for “Sonderforschungsbereich 129” at the University of Ulm. Over the years, as data accumulated indicating the usefulness of the research, the habituation method was implemented as a routine assessment on the Depression Ward. The research was conducted in accordance with the Helsinki Declaration as revised in 1989. 1.2. Suicidal behaviour Reported suicidal behaviour was assigned to one of five categories, “Violent suicide”, “Nonviolent suicide”, “Violent attempt”, “Nonviolent attempt” and “Other or no suicidal behaviours”. It was assumed that the presence of death intent was more often associated with violent suicide attempts than with attempts utilising non-violent methods (Astruc et al., 2004). Since the electrodermal hyporeactivity was most strongly related to suicidal acts in which intent of death was presumed to be most plausible (Thorell, 2009), the present study is focused more on suicide and violent attempts than on nonviolent attempts. 1.3. Severity of depression The Beck Depression Inventory (BDI) is a multiple-choice selfrating scale for depressive symptomatology, listing 21 groups of symptoms and attitudes graded in four levels, 0e3 (Beck et al., 1961). A score between 0 and 9 represents minimal depressive symptoms, scores of 10e16, mild depression, 17e29, moderate depression, and 30e63, severe depression. The total BDI score was used to estimate the severity of a patient’s depression. BDI scores were available for 750 patients.
Following a rest period of 3 min, a series of 10 sinus tones (85 dB, 1 kHz, 1 s duration) and varying interstimulus intervals (15, 20 and 25 s) were administered in a sequence that appeared random to the test subject. Electrodermal activity was measured as skin conductance (mS) using a standard quasi-constant voltage (0.5 V) method (Boucsein, 1992) applied through two AgeAgCl electrodes (0.5 cm2 each) with Hellige Electrodencreme Nr. 21708307 on the volar side of the middle phalanges of the index and middle fingers of the nondominant hand. Conductance measurements were normalized to an area of 1 cm2. Skin conductance responses were separately amplified with a time constant of 10 s. Responses were defined as amplitudes of at least 0.02 mS starting in a time interval of 0.5e4 s after the start of the stimulus. Responses outside that window were considered as non-stimulus related. The criterion for habituation was non-responsiveness to three successive stimuli. The habituation score was taken as the sequence position of the first in a sequence of at least three stimuli not evoking an electrodermal response. The criterion for electrodermal hyporeactivity was a habituation score 5 (first stimulus indexed as zero ). The methodological approach adhered to international guidelines for the conduct of such procedures (Boucsein, 1992; Lykken and Venables, 1971; Boucsein et al., 2012). Independent laboratories have reported no significant relationship between skin conductance level and suicide or violent suicide attempts (Edman et al. 1986; Thorell, 1987; Keller et al. 1991). The focus of the present study was, for that reason, centred on electrodermal hyporeactivity. Skin conductance level is only considered in the discussion in terms of issues relating to the technical detection of electrodermal responses. 1.6. Statistics Descriptive statistics (frequency, means (M), medians (Md), standard deviation (SD) and semi-interquartile distances (SIQD)) were used to summarise data from various assessments. The c2-test or the Fisher’s exact test, depending on expected cell frequencies, was used to assess the significance of differences in frequencies between categories. The ManneWhitney U-test was applied to between-group differences in ratings. The Spearman rank correlation coefficient, rho (r), with correction for ties was used to estimate associations between rating variables.
1.4. Trait anxiety The self-rating scale of trait anxiety, STAI-Trait, is based on 20 statements which describe the way the subject generally feels (Laux et al., 1981). Thirteen of the statements assess the general level of perceived anxiety and seven further statements assess the perceived absence of anxiety and are reverse scored. Patient answers are categorised using a four point Likert scale: (a) almost never, (b) sometimes, (c) often and (d) almost always. The total STAI-Trait score was used. STAI-Trait scores were available for 680 patients. 1.5. Electrodermal measurements Electrodermal activity measurements were performed during a habituation experiment using procedures that have remained essentially the same over the period 1985e2002 (Straub, 1988; Keller et al., 1991). The equipment used was a Hellige polygraph, Hugo Sachs Instruments, Freiburg, Germany.
Fig. 1. Distribution of BDI scores (n ¼ 750). The median and semi-inter-quartile distance of the BDI scores were 25.0 and 8.0, respectively.
L.H. Thorell et al. / Journal of Psychiatric Research 47 (2013) 1925e1931
1927
Table 1 Distributions of patients in categories of suicidal behaviour according to diagnostic category, gender and age. Categories
Violent suicide
Diagnostic groups: Bipolar Unipolar Other diagnoses Gender: Women Men Age groups: Age e 35 years Age 36e49 years Age 50 e Grand totals
Nonviolent suicide
Violent attempt
Nonviolent attempt n
n
%
n
%
n
%
4 23 5
3.2 4.3 4.2
0 4 0
0.0 0.7 0.0
18 60 6
14.3 11.1 5.1
17 15
3.7 4.6
1 3
0.2 0.9
42 42
6 11 15 32
2.8 3.6 5.8 4.1
1 0 3 4
0.5 0.0 1.2 0.5
31 35 18 84
The rejection level for the null hypothesis was set to a ¼ 0.05. All tests were two-tailed. 1.7. Definitions The prevalence of a certain suicidal behaviour or of electrodermal hyporeactivity is defined as the number of patients showing this feature as a percentage of the total patient number in a defined group. The sensitivity of hyporeactivity for a particular category is the percentage among patients with a suicidal behaviour that are electrodermally hyporeactive. The raw specificity of electrodermal hyporeactivity for a certain suicidal behaviour is here the percentage among all electrodermally reactive patients in a defined group who did not manifest that behaviour. The term suicidal propensity is used strictly for a condition assumed to exist behind the acts of suicide and of suicide attempt with death intent. Trait marker is used for an objective measure that shows high sensitivity and specificity for a time extended disease, here the suicidal propensity. 2. Results The patient sample comprised 454 women (age M ¼ 42.3; SD ¼ 11.8) and 329 men (age M ¼ 43.7; SD ¼ 11.0). The age range was between 18 and 66 years, with a mean of M ¼ 42.9; SD ¼ 11.5 years. Five patients lacking information on age were given the mean age, 43 years.
Table 2 Distributions of groups of electrodermal reactivity according to diagnostic category, age and gender. Categories
Diagnostic groups: Bipolar Unipolar Other Gender: Women Men Age groups: Age e 35 years Age 36e49 years Age 50 e Grand totals
Hyporeactives
Reactives
Totals
n
%
n
%
n
%
101 363 69
Prevalence 80.2 67.3 58.5
25 176 49
19.8 32.7 41.5
126 539 118
100.0 100.0 100.0
316 217
69.6 66.0
138 112
30.4 34.0
454 329
100.0 100.0
142 201 190 533
66.0 65.0 73.4 68.1
73 108 69 250
34.0 35.0 26.6 31.9
215 309 259 783
100.0 100.0 100.0 100.0
Other or no suicidal behaviour
Totals
%
n
%
n
%
9 73 21
7.1 13.5 17.8
95 379 86
75.4 70.3 72.9
126 539 118
100.0 100.0 100.0
9.3 12.8
65 38
14.3 11.6
329 231
72.5 70.2
454 329
100.0 100.0
14.4 11.3 6.9 10.7
34 45 24 103
15.8 14.6 9.3 13.2
143 218 199 560
66.5 70.6 76.8 71.5
215 309 259 783
100.0 100.0 100.0 100.0
The patient sample comprised; bipolar (n ¼ 126), unipolar (n ¼ 539) and “other diagnoses” (n ¼ 118), comprising dysthymia (n ¼ 87) or depression with personality or adjustment disorder (n ¼ 31). The distribution of BDI scores is shown in Fig. 1. The patient sample included 36 patients (4.6% of 783) that subsequently committed suicide, 187 patients who attempted to do so before assessment or later (23.9% of 783) of which 84 had employed violent methods (10.7% of 783) and 103 non-violent methods (13.2% of 783) (Table 1). These percentages represent the prevalence of the different suicidal behaviours in the different groups (Table 1). Electrodermal reactivity according to diagnostic category, gender and age is shown in Table 2. The percentages in the column “Hyporeactives” indicate the prevalence of electrodermal hyporeactivity in the different groups. Differences in the distribution of electrodermal reactivity was statistically significantly between the three diagnostic categories (df ¼ 2; c2 ¼ 13.6; p ¼ 0.0011) and between bipolar and unipolar together with other diagnoses (df ¼ 1; c2 ¼ 10.09; p ¼ 0.0015) (Table 2). The prevalence of hyporeactivity was significantly higher in the bipolar group (Table 2). Electrodermal reactivity was similar in women and men (df ¼ 1; c2 ¼ 1.17; p ¼ 0.28, N.S.) (Table 2). There was a higher proportion of hyporeactives in the oldest age group, but the difference was not statistically significant (df ¼ 2; c2 ¼ 5.04; p ¼ 0.08, N.S.) (Table 2). The sensitivity and raw specificity of electrodermal hyporeactivity for suicide in the total sample and in the diagnostic groups are shown in Table 3 and for suicide or violent suicide attempts in Table 4. There was no significant association between electrodermally reactive and hyporeactive patients in terms of their total BDI scores (ManneWhitney U-test: z ¼ 0.86; p ¼ 0.39, N.S.), (Fig. 2, Table 5). The proportions of hyporeactive patients in the four classes of depression severity according to the BDI scores are shown Table 5.
Table 3 Sensitivity and raw specificity of electrodermal hyporeactivity for suicide in diagnostic groups. Categories
Diagnostic groups: Bipolar Unipolar Other Grand totals
Suicides
Hyporeactives
Reactives
Suicides
n
n
n
n
% Sensitivity
4 27 5 36
3 22 5 30
75.0 81.5 100.0 83.3
% Specificity
25 176 49 250
1 5 0 6
96.0 97.2 100.0 97.6
1928
L.H. Thorell et al. / Journal of Psychiatric Research 47 (2013) 1925e1931
Table 4 Sensitivity and raw specificity of electrodermal hyporeactivity for suicide and violent attempt in the whole sample and in the diagnostic groups. Categories
Diagnostic groups: Bipolar Unipolar Other Grand totals
Suicides or violent attempts
Hyporeactives
Reactives
Suicides or violent attempts
n
n
n
n
% Sensitivity
22 87 11 120
21 59 9 89
95.5 67.8 81.8 74.2
% Specificity
25 176 49 250
1 28 2 31
96.0 84.1 95.9 87.6
The STAI-Trait scores (n ¼ 680) ranged from 20 to 79 (Median ¼ 58.0; Semi-interquartile distance ¼ 7.5). No statistically significant relationship was found between electrodermally reactives and hyporeactives and the STAI-Trait scores (ManneWhitney U-test; z ¼ 1.23; p ¼ 0.22, N.S.). The Spearman correlation coefficient between the BDI and STAI-Trait sum scores was statistically significant (r ¼ 0.69, p < 0.0001). Twenty-six patients were tested twice, once during one episode, called “former” episode, and once in a later depressive episode. The time between the tests varied from 0 to 12 years, 7 patients had a further depressive episode within one year, 12 between one and five years later, 5 after six years later or more, and there were 2 patients with unknown intervals. Fig. 3 shows the habituation scores in a former and a later episode. Of 20 hyporeactive patients, 18 (90%) remained hyporeactive. All 6 reactive patients (100%) became hyporeactive. The difference in stability over episodes between patients exhibiting reactivity and hyporeactivity during their first depressive episode was statistically significant (Fisher’s exact probability test, p ¼ 0.001). 3. Discussion The present study shows that; the prevalence of electrodermal hyporeactivity in the entire sample was high and highest (80%) among bipolar patients (Table 2); the sensitivity and specificity for suicide (Table 3), and for suicide and violent suicide attempt taken together (Table 4) were also high, 83%, 98% and 74%, 88%, respectively; electrodermal hyporeactivity was independent of rated severity of depression, trait anxiety, gender and age; and that electrodermal hyporeactivity remained in a later episode, whilst reactivity changed into hyporeactivity in a later episode (Fig. 3). 3.1. Patient sample The patient sample was collected over a period of 17 years, from 1985 to 2002 and as such, constitutes the largest sample of electrodermal reactivity and suicidal behaviour data in depressed patients analysed so far. However, there are several limitations to the study which require consideration. During the collection period, the inclusion criteria for patients varied somewhat due to shifts of research foci. However, the central inclusion criteria remained constant, i.e. depressed patients treated within the framework of their antidepressive therapy at the Depression Ward. BDI scores for 33 patients (4%) and STAI scores for 103 patients (13%) were not available, but this was considered not to influence the main results or conclusions drawn. The distribution of symptom severity according to BDI is skewed by selection as many patients were admitted to the ward specifically because of the risk that they might commit or attempt
Fig. 2. Distribution of BDI scores in classes of severity of depression (n ¼ 750) and the share of hyporeactive patients within each class.
suicide: 28.5% committed suicide or made an attempt. Therefore, any generalizations about other patient groups are not possible. 3.2. Classification of suicidal behaviour The method of classifying suicidal behaviour into suicide, violent suicide, suicide attempt, violent suicide attempt and one common category for no or other suicidal behaviour is based on the principle using observable behaviour for the classification. However, it cannot be ruled out that an unknown number of patients in the category of “Other or no suicidal behaviours” may have made one or more suicide attempts without reporting them. Therefore, the classifications in the first four categories are more useful. 3.3. Electrodermal methodology The use of a fixed criterion of minimum amplitude for the electrodermal response was found preferable to minimum amplitude relative to the basal level of the skin conductance, since this reduces the risk of rejecting obvious orienting responses when the basal conductance level is high. The very short latency criterion of 0.5 s may have the effect of occasionally classifying some single hyporeactive patients as reactive. The test procedure remained virtually identical throughout the entire research period (Keller et al., 1991; see also Wolfersdorf et al., 1999). 3.4. Prevalence The high prevalence of electrodermal hyporeactivity is typical for inpatients on a depression ward (Thorell, 2009). The prevalence ranges from 56% to 80% in inpatients, and from 13% to 23% in outpatients in different studies (Thorell, 2009). The current study shows for the first time that patients with bipolar disorder display the highest prevalence (80%, 101 of 126) of electrodermal hyporeactivity compared to patients with other affective disorders (66%, 432 of 657). This is consistent with observations that patients with bipolar disorder carry the highest risk of suicide among affective disorders (Bernstein et al., 1982). However, it does not imply that such a high prevalence is typical for persons with bipolar disorder in general, since the sample comprises patients that were hospitalized, in part because of suicide risk.
L.H. Thorell et al. / Journal of Psychiatric Research 47 (2013) 1925e1931
1929
Table 5 Distribution of BDI scores according to BDI classes of depression severity and electrodermal reactivity. Categories
Reactivity: Hyporeactive Reactive Grand totals
Minimal depression 0e9
Mild depression 10 e16
Moderate depression 17e29
Serious depression 30e63
Totals
n
%
n
%
n
%
n
%
n
%
45 23 68
66.2 33.8 100.0
84 37 121
69.4 30.6 100.0
221 97 318
69.5 30.5 100.0
161 82 243
66.3 33.7 100.0
511 239 750
68.1 31.9 100.0
3.5. Sensitivity and specificity The act of suicide is, in the end, the result of intrinsic characteristics (suicidal propensity) in combination with external factors (inadequate or absent suicide risk assessment, insufficient antidepressant therapy, failing suicide preventive measures, access to the means and opportunity to commit suicide). Therefore, the usual definition of sensitivity for suicide was replaced by one which defined sensitivity as the percentage of true positives in the entire suicide group. Assuming that depressed patients who commit suicide constitute a group with very high suicidal propensity, the percentage of hyporeactives among patients who commit suicide can be taken as an index of sensitivity for suicidal propensity (Thorell, 2009). In accordance with previous findings (Thorell, 2009) the sensitivity and specificity for suicide in depressed patients was high in the present patient sample. However, electrodermal hyporeactivity is found in other diagnostic groups: for example, in 50% of patients with schizophrenia (Harris and Barraclough, 1997). But, there are no reports of an association between suicidal behaviour and hyporeactivity in patients with schizophrenia or other diagnoses such as panic disorder and general anxiety (unpublished). Hyporeactivity is seen in 17% of the hospital staff (Thorell, 2009) and around 12% in students (unpublished). The present study is also the first report on suicide and electrodermal hyporeactivity in dysthymic patients. The prevalence of electrodermal hyporeactivity in the dysthymic group was 56% (49 of 87) and the sensitivity for suicide was 100%: four patients committed suicide, all being electrodermally hyporeactive. However, since the number of suicides is very small this estimate of sensitivity should be regarded cautiously and further studies are required to confirm this finding. The specificity is computed here without considering the high frequency of suicides and violent attempts in the depressive group. The reason is that the clinical implication, to distinguish between high and low risk patients, is unaffected by that choice. The present
computation evokes interesting theoretical issues regarding the interpretation of the concept of specificity. For example, is the observed high raw specificity due to a low background frequency of suicides or to a high frequency of suicide preventing reactivity? In any case a negative test result, i.e. being electrodermally reactive, may be a strong indication of very low risk of suicide and suicide attempt with death intent. The observations of high sensitivity and specificity of electrodermal hyporeactivity for suicide support repeated findings reviewed by Thorell (2009). 3.6. Central dysfunction Jandl and co-workers reported faster habituation of both the Event Related Potential Novelty P3, a central indicator of the orienting reaction and of the electrodermal responses to the same tone stimuli in a habituation test in patients with major depressive disorder with a history of a severe suicide attempt compared to those who had not (Jandl et al., 2010). This finding offers additional support to the thesis that electrodermal orienting hyporeactivity is due to a common central dysfunction. Thorell suggested in 2009 that orienting hyporeactivity stems from an ‘interest limiting precognitive set’ in non-demanding life situations (Thorell, 2009). This may be thought of as an unconscious reduction in the subject’s awareness of events occurring around them in a passive context, as opposed to the active attention necessary for demanding/complex tasks. It is however beyond the scope of this article to discuss possible neurophysiological/neuroanatomical features manifested as electrodermal hyporeactivity. The similar distributions of electrodermal hyporeactivity among women and men indicate that it is independent of gender. Hence, the difference observed in the prevalence of suicides in men and women is most probably the result of other factors (Hawton, 2000). 3.7. Stability
Fig. 3. Habituation scores (HS) of 26 patients tested in two different depressive episodes. The shadowed area represents hyporeactivity in both episodes.
This is the first time data is presented that suggests that electrodermal hyporeactivity remains stable over depressive episodes and that there is a trend from reactivity found in a “former” depressive episode to hyporeactivity in a subsequent depressive episode. The lower sensitivity found in the present study compared to the earlier meta-analysis (Thorell, 2009) is most probably explained by this movement from reactivity towards hyporeactivity in a later depressive episode. It now appears that brain dysfunction increases in patients with recurrent depressive episodes (Sheline, 2000). Therefore, we would expect that orienting hyporeactivity as a symptom of brain dysfunction would become more frequent in subsequent depressive episodes. It is concluded that the sensitivity for suicide and violent attempt in the present patient sample is most probably an underestimation compared to the high sensitivity in a previous metaanalysis (Thorell, 2009).
1930
L.H. Thorell et al. / Journal of Psychiatric Research 47 (2013) 1925e1931
A previous report showed that the electrodermal hyporeactivity persisted in remission after two years (Thorell and d’Elia, 1988). In combination with the observed stability of hyporeactivity over episodes seen in the present patient sample, it can be assumed that hyporeactivity may be chronic in many patients. In terms of the stability of electrodermal reactivity within a depressive episode, a statistically significant stability was found between tests at admission and tests at dismissal in 46 depressed patients (Straub and Hole, 1988). However, there were cases of patients who moved from reactivity to hyporeactivity and vice versa. The movement from reactivity to hyporeactivity may represent the emerging of a central dysfunction during the episode. Movement from hyporeactivity to reactivity is particularly interesting in terms of studying effective therapeutic options for hyporeactive patients. Considering (a) that all studies on electrodermal hyporeactivity and suicide in depression (Edman et al., 1986; Thorell, 1987 and the present) performed the habituation test at or soon after the time of admission, (b) the extraordinary high relationship between electrodermal hyporeactivity and suicidal propensity (Thorell, 2009) and (c) its high sensitivity and specificity for suicide (Thorell, 2009 and the present), we can conclude that there is a substantial stability of hyporeactivity over depressive episodes. For these reasons and for the need of early detection of suicide risk it is recommended that electrodermal reactivity is tested early in the depressive episode in order to be of value as a support in suicide risk assessment. 3.8. Marker for suicidal propensity The high sensitivity and raw specificity of the habituation test for suicide and violent suicide attempt and the stability over depressive episodes justifies the hypothesis that electrodermal hyporeactivity is a useful trait marker for suicide propensity in mood disorders, at least in depressive phases. 3.9. Independence of depression This is the first published study that demonstrates the independence of electrodermal hyporeactivity to rated severity of depression in a large patient sample. It is in line with the previous observation (Thorell, 2009) regarding the independence of ratings derived from the depression scale of the Comprehensive Psychopathological Rating Scale (CPRS) (Åsberg et al., 1978). Electrodermal hyporeactivity was also unrelated to trait anxiety as assessed by the STAI-Trait scale which, in turn, is correlated with the depressive symptomatology rating according to BDI. The independence of hyporeactivity from the severity of depressive disorders is further indicated by the finding that it is not influenced by successful antidepressive treatment (Noble and Lader, 1971; Dawson et al., 1977; Storrie et al., 1981; Toone et al., 1981; Janes and Strock, 1982). It is further strengthened by the observed stability of habituation of electrodermal responses after one year (Iacono et al., 1984) and in remission after two years (Thorell and d’Elia, 1988). The time course of hyporeactivity seems to be clearly different from that of the clinical depressive episode. A time-extended factor that is more enduring than that of the clinical depression seems to be involved behind suicidal propensity as estimated from electrodermal hyporeactivity. The time-extended suicidal propensity is supported by the repeated findings of a strong relationship between the occurrence of a previous suicide attempt and the later suicide (Coryell, 2006). Further indications that electrodermal hyporeactivity is independent of the time course of the depression is that it is not
correlated with levels of serotonin in liquor (Edman et al., 1986) and is not influenced by serotonergic antidepressant therapy (Noble and Lader, 1971; Dawson et al., 1977; Storrie et al., 1981; Toone et al., 1981). However, some indication exists of the contrary (Hellewell et al., 1999). It seems likely from our data that the central dysfunction behind the electrodermal hyporeactivity in depressed patients is most probably largely independent of the depressive state, but it cannot be ruled out that other interactions may exist. 3.10. Limitations Changes to research priorities during the long study period have meant that patient inclusion criteria have varied somewhat. Consequently, some data are missing for particular variables, but the conclusions from the study are considered not to be significantly affected. Although the dataset used in this study is large, the number of actual suicides is rather small. This is a consequence of the fact that suicide is a rare incident. Hence the data in Tables 1 and 3 should be interpreted with caution. Nevertheless, the results are in strong agreement with other reports. As with other research studies on this topic, there may be cases in which suicide attempts have been made, but not reported, which may result in an underestimation of the number of attempts and to some degree of uncertainty as to the classification of patients according to their suicidal behaviour. Further, the number of retested patients in a later depressive episode was small. The reason for this is that the retesting was fortuitous rather than planned. More systematic follow up of cases is required in future studies. 3.11. Comments Taken together, the findings from previous studies and this current study strongly support the thesis that electrodermal hyporeactivity represents a trait marker for suicidal propensity in depression that is independent of the type or severity of the depressive disorder. The high sensitivity and raw specificity of electrodermal hyporeactivity for suicide and violent suicide attempt underline its potential in clinical contexts, particularly early in the depressive episode. The electrodermal method is in fact the first clinically applicable objective method to support the clinical assessment of suicide risk in depressed patients. However, more data is required from prospective clinical studies to study the value of hyporeactivity as a predictive marker for suicidal propensity. Data presented in a report from the Swedish Board of Health and Welfare indicates, that health care systems are successful in managing suicide prevention (99.8%) where a high risk of suicide in a patient has been ascertained (Socialstyrelsen, 2010; Thorell and Eriksson, submitted for publication). Thus, a good assessment procedure for suicide risk, to which the test of electrodermal hyporeactivity is expected to contribute, seems to be a key to the success of future suicide prevention. Funding body agreements and policies There are no relevant agreements on financing to declare. Conflict of interest First author is Director of Research and owns shares in the Swedish company Emotra. Other authors declare no conflict of interest.
L.H. Thorell et al. / Journal of Psychiatric Research 47 (2013) 1925e1931
Acknowledgements We thank Peter Lauwasser who performed the psychophysiological measurements in the laboratory and Doris Herforth, who organized and computed original test data. Further, we thank Rita Goebel for administrative work. This study was supported by a grant from Deutsche Forschungsgemeinschaft (DFG Sonderforschungsbereich 129). References Åsberg M, Perris C, Schalling D, Sedvall G. The CPRS e development and applications of a Psychiatric Rating Scale. Acta Psychiatrica Scandinavica 1978;271:33e7. Astruc B, Torres S, Jollant F, Jean-Baptiste S, Castelnau D, Malafosse A, et al. A history of major depressive disorder influences intent to die in violent suicide attempters. Journal of Clinical Psychiatry 2004;65:690e5. Beck AT, Ward C, Mendelson M. Beck depression Inventory (BDI). Archives of General Psychiatry 1961;4:561e71. Bernstein AS, Frith CD, Gruzelier JH, Patterson T, Straube E, Venables PH, et al. An analysis of the skin conductance orienting response in samples of American, British, and German schizophrenics. Biological Psychology 1982;14:155e211. Boucsein W. Electrodermal activity. New York: Plenum Press; 1992. Boucsein W, Fowles DC, Grimnes S, Ben-Shakhar G, Roth WT, Dawson ME, et al. Publication recommendations for electrodermal measurements. Psychophysiology 2012;49:1017e34. Coryell WH. Clinical assessment of suicide risk in depressive disorder. CNS Spectrums 2006;11:455e61. Dawson ME, Schell AM, Catania JJ. Autonomic correlates of depression and clinical improvement following electroconvulsive shock therapy. Psychophysiology 1977;14:569e77. Edman G, Åsberg M, Levander S, Schalling D. Skin conductance habituation and cerebrospinal fluid 5-hydroxyindoleatic acid in suicidal patients. Archives of General Psychiatry 1986;43:586e92. Harris CE, Barraclough BM. Suicide as an outcome for mental disorders. British Journal of Psychiatry 1997;170:205e8. Hawton K. Sex and suicide. Gender differences in suicidal behavior. British Journal of Psychiatry 2000;177:484e5. Hellewell JS, Guimaraes FS, Wang M, Deakin JF. Comparison of buspirone with diazepam and fluvoxamine on aversive classical conditioning in humans. Journal of Psychopharmacology 1999;13:122e7. Iacono WG, Lykken DT, Haroian KP, Peloquin LJ, Valentine RH, Tuason VB. Electrodermal activity in euthymic patients with affective disorders: one-year retest stability and the effects of stimulus intensity and significance. Journal of Abnormal Psychology 1984;93:304e11. Jandl M, Steyer J, Kaschka WP. Suicide risk markers in major depressive disorder: a study of electrodermal activity and event-related potentials. Journal of Affective Disorders 2010;123:138e49. Janes CL, Strock BD. Skin conductance responding following major depressive episode remission. Psychophysiology 1982;19:566. Keller FM, Wolfersdorf M, Straub R, Hole G. Suicidal behaviour and electrodermal activity in depressive inpatients. Acta Psychiatrica Scandinavica 1991;83:324e8.
1931
Laux L, Glanzmann P, Schaffner P, Spielberger CD. Das State-Trait-Angstinventar. Theoretische Grundlagen und Handanweisung. Weinheim: Beltz Test GmbH; 1981. Lykken DT, Venables PH. Direct measurement of skin conductance: a proposal for standardization. Psychophysiology 1971;8:656e72. Noble P, Lader M. The symptomatic correlates of the skin conductance changes in depression. Journal of Psychiatric Research 1971;9:61e9. Sheline YI. 3D MRI studies of neuroanatomic changes in unipolar major depression: the role of stress and medical comorbidity. Biological Psychiatry 2000;48: 791e800. Socialstyrelsen (National Board of Health and Welfare). Självmord 2006e2008 anmälda enligt lex Maria [Suicides 2006e2008 reported according to the lex Maria]. Stockholm: Socialstyrelsen; 2010. Article 4e5. Storrie MC, Doerr HO, Johnson MH. Skin conductance characteristics of depressed subjects before and after therapeutic intervention. Journal of Nervous and Mental Disease 1981;69:176e9. Straub R. Klinische Psychophysiologie der Depression [Clinical psychophysiology of depression]. In: Wolfersdorf M, Kopittke W, Hole G, editors. Klinische Diagnostik und Therapie der Depression [Clinical diagnostics and therapy of depression]. Regensburg: Roderer; 1988. p. 142e59. Straub R, Hole G. Zur Psychophysiologie depressiver Syndrome [(On psychophysiology of the depressive syndrome]. In: Kopittke W, Wolfersdorf M, editors. 10 Jahre Weissenauer Depressionsstation zwischen Alltag und Forschung. Regensburg: Roderer; 1988. Straub R, Jandl M, Wolfersdorf M. Mental state and electrodermal activity in depressed patients during acute suicidal period. Psychiatrische Praxis 2003;30(Suppl. 2):s183e6. Thorell LH. Electrodermal activity in suicidal and nonsuicidal depressive patients and in matched healthy subjects. Acta Psychiatrica Scandinavica 1987;76: 420e30. Thorell LH. Valid electrodermal hyporeactivity for depressive suicidal propensity offers links to cognitive theory. Acta Psychiatrica Scandinavica 2009;119: 338e49. Thorell LH, d’Elia G. Electrodermal activity in depressive patients in remission and in matched healthy subjects. Acta Psychiatrica Scandinavica 1988;78:247e53. Thorell LH, Eriksson T. Skyddar fysiologiskt mätbar nyfikenhet den deprimerade mot självmord? [Is physiologically measurable curiosity protecting the depressed against suicide?]. Submitted for publication. Toone BK, Cooke E, Lader MH. Electrodermal activity in the affective disorders and schizophrenia. Psychological Medicine 1981;11:497e508. Vigouroux A. Etude sur la résistance électrique chez les mélancholiques. J. Rueff & Cie; 1890. Thése de Paris. Wolfersdorf M, Straub R, Hole G. Electrodermal activity in depressive men and women with violent or non-violent suicide attempts. Schweizer Archiv für Neurologie und Psychiatrie 1993;144:173e84. Wolfersdorf M, Straub R, Barg T, Keller F, Kaschka WP. Depressed inpatients, electrodermal reactivity, and suicide e a study about psychophysiology of suicide behaviour. Archives of Suicide Research 1999;5:1e10. World Health Organization. Mental disorders: glossary and guide to their classification in accordance with the ninth revision of the International Classification of Disease (ICD-9). WHO; 1978. World Health Organisation. The ICD-10 classifications of mental and behavioural disorders: clinical descriptions and diagnostic guidelines. Geneva: World Health Organisation. WHO; 1992.
