Measurement of depressive symptoms among cardiac patients:
Should sex differences be considered?
Abstract
Confounding of depressive and cardiac symptoms may hold implications for assessment.
This study investigated psychometric properties and sex differences in two depression
scales among cardiac patients. Cardiac inpatients from 11 hospitals were recruited, and
completed a mailed survey including the Beck Depression Inventory-II and Gotland Scale
of Male Depression one-year later.
The scales were significantly correlated and both were associated with social desirability.
Females scored higher than males on the fatigue factor only (p<.001).
Psychometric properties of the Beck Depression Inventory-II were more favorable in this
population. Practitioners must not overlook reports of fatigue in female cardiac patients.
Keywords
depression, gender, coronary heart disease, self-perception, well-being
2
Introduction
There is a growing epidemic of cardiovascular disease globally (Mendis et al., 2011). One
in five patients with established cardiovascular disease also suffers from clinicallysignificant depression (Rutledge et al., 2006; Thombs et al., 2006). Depression is more
prevalent in the cardiac population than in the general population (Naqvi et al., 2005). This
is alarming, given that depressive symptoms have been associated with 2-times increased
risk of morbidity and mortality among coronary patients (van Melle et al., 2004).
Recently, it has been observed that individual symptoms of depression may be
differentially associated with cardiovascular prognosis (Tully et al., 2011). Specifically,
somatic symptoms such as sleeping difficulties and fatigue have been demonstrated to be
particularly hazardous, when compared to cognitive symptoms such as shame and guilt
(Martens, 2010).
Epidemiological studies in the general population have established prevalence rates
of major depression averaging two times higher in females than in males, and they also
experience longer and more recurrent episodes (Kessler et al., 1993). Moreover, depressive
symptoms have been shown to be more prevalent and severe in females
(Shanmugasegaram et al, 2012). In accordance with the biopsychosocial model, reasons for
this sex disparity are multi-factorial and include genetic, psychosocial and environmental
etiologies (Nolen-Hoeksema et al, 1999; Piccinelli and Wilkinson, 2000). For instance,
females may be more likely to experience chronic strain, to have a low sense of selfmastery, and engage in ruminative thoughts than males (Nolen-Hoeksema et al., 1999).
Moreover, some research has shown sex differences in response to evidence-based
3
depression treatment among cardiac patients, with women less responsive to therapies than
men (André-Petersson et al., 2011; Berkman et al., 2003).
These findings have lead to questions regarding sex differences in the nature of
depressive symptoms (Linke et al., 2009). For example, work by Ketterer et al. (2004)
suggests that depressed female patients may be more likely to cry, while males are more
likely to deny depressive symptoms when compared to females. Depression assessment
and diagnosis rests predominantly on ‘conventional’ or ‘typical’ symptoms which may be
more prevalent in females. In cardiac samples, the Beck Depression Inventory (BDI) is
commonly-used; a measure consisting of two main factors, namely somatic and cognitive
symptoms, on which women generally score higher than men. However, it has been
suggested that males may express a different subset of depressive symptoms commonlyassociated with masculinity, such as self-abusive behaviors and aggression, which are not
assessed in the frequently-administered depression scales (Moller-Leimkuhler and Yucel,
2010). Evidence supporting the idea of male under-screening in current diagnostic tools of
depression has also come from suicide studies, considering suicide rates are much higher
for males than females in the general population (Rutz et al., 1995). Others have pointed to
the greater preponderance of externalizing disorders such as substance abuse and some
personality disorders among males (Compton et al., 2007).
Recent research has also suggested that depressed post-myocardial infarction
patients are significantly less likely to exhibit cognitive symptoms than psychiatric patients
(Martens et al., 2006). Coupled with concerns regarding potential confounding of somatic
depressive symptoms (i.e., psychomotor retardation, fatigue, hypersomnia) with cardiac
4
cardiovascular disease severity (Nicholson, Kuper & Hemingway, 2006; Ramsay, et al.,
2001), these questions have significant implications for depression measurement and
assessment in this population. The first objective of the present study was to compare the
psychometric characteristics of depressive symptom scales assessing conventional somatic
and cognitive (i.e., BDI-II), as well as atypical (i.e., Gotland Scale of Male Depression;
GSMD); Rutz et al., 1995) symptoms in the cardiac population. In particular, internal
reliability, the threat of socially-desirable responding, sensitivity, specificity, positive and
negative predictive values, as well as the factor structure of the scales were examined. The
second objective was to explore the nature of depressive symptoms in male and female
cardiac patients, and whether these symptoms differed significantly by sex.
Method
Design and Procedure
As part of a larger study comparing cardiac rehabilitation enrollment following
different referral strategies (Grace et al., 2011), 2657 cardiac inpatients from 11 hospitals
in Canada were recruited. After consent was obtained, clinical data was extracted from
medical charts, and a self-report survey was completed by the patients. One year later,
participants were mailed a follow-up survey. It included the BDI-II (Beck and Steer, 1996),
and the GSMD (Zierau et al., 2002). The Marlowe-Crowne Social Desirability Scale
(Reynolds, 1982) was also administered considering reluctance to acknowledge
stigmatized behavior such as depressive symptoms, and that males under-report emotional
distress relative to significant others, while the same is not true of females.
Sample
5
Participants were cardiac inpatients recruited between 2006 and 2008 who
completed a survey 1 year post-hospitalization. A study flow diagram illustrating the
retention of patients is shown in a supplemental Figure. Eligibility criteria for the larger
study were as follows: a) being diagnosed with acute coronary syndrome or having
undergone a percutaneous coronary intervention or coronary artery bypass grafting; b)
being eligible for cardiac rehabilitation (Stone et al., 2009); and c) proficiency in English,
French, or Punjabi.
Instruments
Self-reported sociodemographic characteristics were measured within the inhospital survey. Clinical variables, such as comorbidities and smoking status, were also
obtained from the questionnaire. Data obtained from the medical charts included date of
birth, sex, blood pressure, lipids, and cardiac conditions / procedures.
The one year follow-up survey included the Duke Activity Status Index, a social
desirability scale and depressive symptoms scales. The Duke Activity Status Index is a
brief 12-item self-administered survey to determine patients’ functional capacity (Hlatky et
al., 1989).
The short-form of the original 33-item Marlowe-Crowne Social Desirability Scale
(Reynolds, 1982) was administered in the one year post-hospitalization survey. It is a
commonly-used assessment tool for social desirability bias, or the tendency to answer
questions in a manner that will be viewed favorably by others. The scale consists of 10
items, and participants selected “true” or “false” for each statement. Scores range from 0 to
10, with higher scores representing higher need for approval or social desirability.
6
Scales to measure depressive symptoms. Depressive symptoms were measured
using the BDI-II (Beck and Steer, 1996) and the GSMD (Zierau et al., 2002). Items in the
BDI-II assess somatic and psychological symptoms reflecting diagnostic criteria. The BDIII contains 21 items which are rated on a scale from 0 to 3. Scores ≥14 were used to
indicate elevated depressive symptoms (Low, 2006). It is a frequently-used self-report
measure in assessing depressive symptoms in the cardiac population (Frasure-Smith and
Lesperance, 2008).
To assess depressive symptoms thought to be more commonly-experienced in
males, the GSMD was developed. The scale (Zierau et al., 2002) assesses both traditional
depressive symptoms (e.g., sleeping and concentration problems), as well as externalizing
depressive symptoms expected to be predominant in males (e.g., abusive behaviors,
aggression). The 13-items are also rated on a 4-point scale, where total scores range from
of 0 to 39. Responses are summed to form a global score, where interpretative cutoffs
suggested by Rutz (1995) include no signs of depression (0-13), possible depression (1426), and clear signs of depression (26-39). Scores ≥13 commonly serve as a screen for
depressive symptoms. The GSMD has high internal reliability (α=.86; Zierau et al., 2002),
and demonstrates convergent validity with the Beck Hopelessness Scale (Innamorati et al.,
2011). These psychometric characteristics were determined following administration in
predominantly student samples however, and as the scale has not previously been
administered in a cardiac sample to our knowledge its performance in this context is
unknown. The most recent psychometric validation of this scale concluded it is a valid
7
instrument for measuring atypical “suicidality-related” symptoms of depression in male
and female patients (Innamorati et al., 2011; Pompili et al., 2012).
Statistical analyses. Characteristics of retained, ineligible and declined participants
were compared using one-way analysis of variance and χ2 as appropriate. Bivariate tests,
including χ2 tests for categorical variables and independent samples t-tests for continuous
variables were used to compare sex differences in the sociodemographic and clinical
characteristics of retained participants.
To test the first objective, the association of both depressive symptoms scales to the
Marlowe-Crowne Social Desirability Scale (Reynolds, 1982) was tested using Pearson’s
correlation. Internal consistency of the BDI-II and GSMD was tested using Cronbach’s α.
According to established conventions, coefficients ranging from 0.70 to 0.95 were
considered “acceptable”.
Dichotomous variables were created to reflect elevated depressive symptoms on
both the BDI-II and GSMD scales. Agreement between the two scales was tested via
Cohen’s κ. Interpretation of Cohen’s κ followed the guidelines set by Landis and Koch
(Landis and Koch, 1977). Sensitivity, specificity, positive and negative predictive values
were also calculated against the BDI-II. These analyses were also conducted with the data
stratified by sex.
Finally, a principal components analysis with Varimax rotation was used to assess
the factorial composition of all items on both scales. Criteria for choosing the rotated factor
solution included the Cattell’s scree plot criterion (Cattell, 1977), eigenvalues >1, simple
structure, and minimum 50% variance explained by the resultant factors.
8
To test the second objective, mean scores on the BDI-II and GSMD were computed
and compared by sex using t-tests. Independent samples t-tests were used to test for sex
differences in the extracted factors, as well as individual items in both scales using a
Bonferroni correction (p<.0125).
Results
Respondent Characteristics
Overall, 1784 (98.7%) completed the survey in English, 22 (1.3%) in French, and 1
in Punjabi. Table 1 (available in online supplement) displays the characteristics of the
retained, ineligible and declining participants at 1 year post-hospitalization. As shown,
retained participants were significantly more likely to be older, married, white, and to have
undergone coronary artery bypass graft surgery or percutaneous coronary intervention than
declining and ineligible patients.
Sex differences in the sociodemographic and clinical characteristics of retained
participants are also displayed in Table 1. With regard to the latter, females were
significantly older, and were more likely to be retired, unmarried, have ≤13 years of
education, and reside farther than 30 minutes drive to a hospital than males. As shown,
females also scored significantly higher than males on both the BDI-II and the GSMD.
Psychometric properties and level of agreement between the BDI-II and GSMD
overall and by sex. The correlation between each of the BDI-II and GSMD with the
Marlowe-Crowne Social Desirability Scale, and the internal consistency of the two scales
is shown for the overall sample and by sex in Table 2. Significant associations between the
Marlowe-Crowne Social Desirability Scale and both the depressive symptom scales were
9
found, and the correlations were somewhat higher in males than females. Both scales had a
high degree of internal consistency regardless of sex. The 2 scales were significantly
related to one another (r=0.77, p<.001).
Table 3 displays the level of agreement between the two depressive symptom
scales, as well as sensitivity, specificity and positive and negative predictive values. As
shown, Cohen’s κ indicated there was moderate agreement between the GSMD and the
BDI-II overall and by sex. Using the suggested cutoff score of > 13, the GSMD yielded
high specificity and predictive values (both negative and positive) when compared to the
BDI-II. However, the GSMD had lower sensitivity, where for both males and females the
BDI-II classified 156 more participants (113 males, 43 females) as having elevated
depressive symptoms when compared to the GSMD.
Factor structure of the BDI-II and GSMD. The resultant rotated factor loadings
are displayed in Table 4 in descending order. Results revealed that four factors provided
the most parsimonious solution for the data. The factors accounted for 52.8% of the
variance in participants’ responses.
Extracted factors were heuristically labeled for interpretation. The items which
loaded on Factor 1 consisted of only BDI-II items, and appeared to represent maladaptive
internalizing cognitions (i.e., cognitive features). Factor 2 included both BDI-II and GSMD
items, and appeared to collectively represent psychological and physical fatigue (i.e.,
somatic features). The third factor consisted of only GSMD items representing
maladaptive externalizing behaviors. Finally, the last factor included items such as
aggression, irritability, agitation, and lower stress threshold stemming from both scales.
10
Sex differences in depressive symptoms. The mean item scores are also displayed
in Table 4 by sex. Overall, sleeping problems (GSMD) and changes in sleeping pattern
(BDI-II) were the most endorsed depressive symptoms from each scale in both males and
females, whereas suicidal ideation (BDI-II) and antisocial behavior (GSMD) were the least
endorsed in both sexes (tie for abusive behavior [GSMD] in women).
There was several significant sex differences in individual scale items, and in all
cases females scored higher than males. Sex differences in factor scores were also tested
using Student’s t-tests. Females scored significantly higher than males on psychomotor
fatigue (Factor 2; p<.001), and there were no other sex differences.
Discussion
Systematic reviews have identified that limited evidence is available on the
psychometric properties of frequently-administered and endorsed depression scales in the
cardiac population (Thombs et al., 2007, 2008; Tully et al., 2011). This is the first study to
have investigated both conventional and atypical depressive symptoms in male and female
cardiac patients through an investigation of psychometric properties of depressive
symptom scales. This examination is particularly important given that depression in
cardiac disease may be atypical, and that female cardiac patients suffer a greater burden of
depression than men. Overall, the findings raise some important questions concerning the
measurement of depression in cardiac populations, and variations (i.e., sex differences) in
how individuals respond to questions about depression.
This study has shown that both the BDI-II and GSMD were significantly associated
with the Marlowe-Crowne Social Desirability Scale, confirming that high social
11
desirability (i.e., a reluctance to acknowledge stigmatized behavior) is associated with
reduced acknowledgement of depression symptoms in the cardiac population as well. This
calls in to question the interpretation of mean tendencies as well as individual differences
in depressive symptoms in this population. Second, it is known that males under-report
emotional distress relative to significant others, while the same is not true of females
(Ketterer et al., 2004). The correlation of both depression scales with the social desirability
scale was stronger in men than women in the current study, supporting previous research.
Further tests of depressive symptoms in the cardiac population should be undertaken with
validation of report by a significant other (Ketterer et al., 1996; Ketterer et al., 2004) to
determine the extent that social desirability bias may be affecting results, with particular
attention paid to sex differences.
Results of the PCA revealed 4 factors, two factors which were comprised of items
from both scales (psychomotor fatigue and aggression-irritability) and two solely
comprised of items from one depression scale (cognitive behaviours [BDI-II] and
externalizing behaviours [GSMD]). The first two and most predominant factors
corresponded with the conventional features of cognitive and somatic depressive
symptoms. Moreover, the BDI may also assess some of the more atypical symptoms,
specifically based on the ‘irritability’ and ‘agitation’ items. Furthermore, there was no
evidence of a sex difference in atypical symptoms in this cardiac sample. Concordant with
more recent research (Moller-Leimkuhler and Yucel, 2010), females were just as likely to
report antisocial and abusive behavior as males. Thus, the often-used BDI does not appear
12
to miss important symptoms in some patient groups (e.g. male cardiac patients), and the
GSMD does not offer much additional utility in depressive symptom assessment.
A key sex difference in the quality of depressive symptoms was identified, whereby
female participants more often reported various forms of psychomotor fatigue than males.
Indeed, this difference is consistent to Moller-Leimkuhler et al.’s (2004) study, where
GSMD items corresponding to psychomotor fatigue were the most strongly correlated
items in females. Moreover, although women seem to be more prone to psychomotor
fatigue than men, both men and women were more likely to endorse fatigue and sleeping
problems than other depressive symptoms. Somatic symptoms may be overlooked as they
can be attributed to the underlying cardiac illness, and therefore depression may go
undetected, particularly in women.
Caution is warranted when interpreting these results. First, there were some
retention biases in the sample, and therefore findings may not be generalizable to the larger
cardiac population. Moreover, there were sex differences in the sociodemographic and
clinical characteristics of the retained sample, consistent with other studies in cardiac
populations, which may also relate to depressive symptoms and could potentially serve as
an alternative explanation for the sex differences in symptoms found. Second, based on the
study design, no causal conclusions can be drawn. Finally, depressive symptoms were
assessed through self-report measures, and additional validation via a structured clinical
interview is warranted in future research.
In conclusion, the GSMD did not contribute to the assessment of depressive
symptoms in the cardiac population over-and-above the commonly-used BDI-II. The BDI-
13
II continues to be an important assessment tool, although researchers should be wary about
social desirability bias, particularly among males. Moreover, the results of this study
confirm the greater degree of depressive symptoms in females than males in the cardiac
population, and extend the evidence by demonstrating that women and men may show
different profiles of their depressive symptoms. Given the particularly hazardous nature of
somatic depressive symptoms in cardiac disease, and that some evidence suggests it is not
confounded with underlying physical illness (Smith et al., 2007), these findings suggest
that practitioners must not overlook reports of fatigue in female cardiac patients.
ENDNOTE: Supplemental Figure 1 and Supplemental Table 1 can be found online at:
http://www.yorku.ca/sgrace/onlinesupplementtopublications.html
14
References
André-Petersson L, Schlyter M, Engström G, Tydén P and Hedblad B (2011) Behavior in a
stressful situation, personality factors, and disease severity in patients with acute
myocardial infarction: baseline findings from the prospective cohort study SECAMI (the
Secondary Prevention and Compliance following Acute Myocardial Infar. BMC
cardiovascular disorders 11(1): 45.
Beck AT and Steer RA (1996) Beck depression inventory manual. San Antonio, TX:
Psychological Corporation.
Berkman LF, Blumenthal J, Burg M, Carney RM, Catellier D, Cowan MJ, Czajkowski SM,
DeBusk R, Hosking J, Jaffe A, Kaufmann PG, Mitchell P, Norman J, Powell LH,
Raczynski JM, Schneiderman N and Enhancing Recovery in Coronary Heart Disease
Patients (ENRICHD) Investigators (2003) Effects of treating depression and low
perceived social support on clinical events after myocardial infarction: The enhancing
recovery in coronary heart disease patients randomized trial. Journal of the American
Medical Association 289(23): 3106-3116.
Cattell RBS (1977) A comprehensive trial of the scree and KG criteria for determining the
number of factors. Multivariate Behavioral Research 12: 289-325.
Compton WM, Thomas YF, Stinson FS and Grant BF (2007) Prevalence, correlates, disability,
and comorbidity of DSM-IV drug abuse and dependence in the United States: Results
from the national epidemiologic survey on alcohol and related conditions. Archives of
General Psychiatry 64(5): 566-576.
15
Frasure-Smith N and Lesperance F (2008) Depression and anxiety as predictors of 2-year cardiac
events in patients with stable coronary artery disease. Archives of General Psychiatry
65(1): 62-71.
Grace SL, Russell KL, Reid RD, Oh P, Anand S, Rush J, Williamson K, Gupta M, Alter DA and
Stewart DE, on behalf of the CRCARE Investigators (2011) Effect of cardiac
rehabilitation referral strategies on utilization rates: A prospective, controlled study.
Archives of Internal Medicine 171(3): 235-241.
Hlatky MA, Boineau RE, Higginbotham MB, Lee KL, Mark DB, Califf RM, Cobb FR and
Pryor DB (1989) A brief self-administered questionnaire to determine functional
capacity. American Journal of Cardiology 64(10): 651-654.
Innamorati M, Pompili M, Gonda X, Amore M, Serafini G, Niolu C, Lester D, Rutz W, Rihmer
and Z and Girardi P (2011) Psychometric properties of the Gotland Scale for Depression
in Italian psychiatric inpatients and its utility in the prediction of suicide risk. Journal of
Affective Disorders 132(1-2): 99-103.
Kessler RC, McGonagle KA, Swartz M, Blazer DG and Nelson CB (1993) Sex and depression in
the national comorbidity survey I: Lifetime prevalence, chronicity and recurrence.
Journal of Affective Disorders 29(2-3): 85-96.
Ketterer MW, Denollet J, Chapp J, Thayer B, Keteyian S, Clark V, John S, Farha AJ and
Deveshwar S (2004) Men deny and women cry, but who dies? Do the wages of "denial"
include early ischemic coronary heart disease? Journal of psychosomatic research 56(1):
119-123.
16
Ketterer MW, Kenyon L, Foley BA, Brymer J, Rhoads K, Kraft P and Lovallo WR (1996)
Denial of depression as an independent correlate of coronary artery disease. Journal of
health psychology 1(1):93–105.
Landis JR and Koch GG (1977) The measurement of observer agreement for categorical data.
Biometrics 33(1): 159-174.
Linke SE, Rutledge T, Johnson BD, Vaccarino V, Bittner V, Cornell CE, Eteiba W, Sheps DS,
Krantz DS, Parashar S and Bairey Merz CN (2009) Depressive symptom dimensions and
cardiovascular prognosis among women with suspected myocardial ischemia: A report
from the national heart, lung, and blood institute-sponsored women's ischemia syndrome
evaluation. Archives of General Psychiatry 66(5): 499-507.
Low GHA (2006) Screening for depression for cardiac events using the Beck Depression
Inventory-II and the Geriatric Depression Scale. Social Indicators Researc 82: 527-543.
Martens EJ, Hoen PW, Mittelhaeuser M, de Jonge P and Denollet J (2010) Symptom dimensions
of post-myocardial infarction depression, disease severity and cardiac prognosis.
Psychological Medicine 40(5): 807-814.
Martens EJ, Denollet J, Pedersen SS, Scherders M, Griez E, Widdershoven J, Szabó B,Bonnier
H, Appels A (2006) Relative lack of depressive cognitions in post-myocardial infarction
depression . Journal of Affective Disorders 94: 231 - 237.
Mendis S, Puska P and Norrving B editors (2011) Global Atlas on Cardiovascular Disease
Prevention and Control. World Health Organization, Geneva.Moller-Leimkuhler AM,
Bottlender R, Strauss A and Rutz W (2004) Is there evidence for a male depressive
syndrome in inpatients with major depression? Journal of Affective Disorders 80(1): 8793.
17
Moller-Leimkuhler AM and Yucel M (2010) Male depression in females? Journal of Affective
Disorders 121(1-2): 22-29.
Naqvi TZ, Naqvi SS and Merz CN (2005) Gender differences in the link between depression and
cardiovascular disease. Psychosomatic Medicine 67(Suppl 1): S15-S18.
Nicholson A, Kuper H and Hemingway H (2006) Depression as an aetiologic and prognostic
factor in coronary heart disease: a meta-analysis of 6362 events among 146 538
participants in 54 observational studies. European Heart Journal 27: 2763–2774.
Nolen-Hoeksema S, Larson J and Grayson C (1999) Explaining the gender difference in
depressive symptoms. Journal of Personality and Social Psychology 77(5): 1061-1072.
Piccinelli M and Wilkinson G (2000) Gender differences in depression: Critical review. British
Journal of Psychiatry 177: 486-492.
Pompili M, Innamorati M, Rihmer Z, Gonda X, Serafini G, Akiskal H, Amore M, Niolu C, Sher
L, Tatarelli R, Perugi G and Girardi P (2012) Cyclothymic – depressive – anxious
temperament pattern is related to suicide risk in 346 patients with major mood disorders.
Journal of Affective Disorders 136(3): 405-411.
Ramsay JM, McDermott MR and Bray C (2001) Components of the anger-hostility complex and
symptom reporting in patients with coronary artery disease: a multi-measure study.
Journal of health psychology 6(6):713–729.
Reynolds WM (1982) Development of reliable and valid short forms of the Marlowe-Crowne
Social Desirability Scale. Journal of Clinical Psychology 38(1): 119-125.
Rutledge T, Reis VA, Linke SE, Greenberg BH and Mills PJ (2006) Depression in heart failure:
A meta-analytic review of prevalence, intervention effects, and associations with clinical
outcomes. Journal of the American College of Cardiology 48(8): 1527-1537.
18
Rutz W, von Knorring L, Pihlgren H, Rihmer Z and Walinder J (1995) Prevention of male
suicides: Lessons from Gotland study. Lancet 345(8948): 524.
Shanmugasegaram S, Russell K, Kovacs A, Stewart DE and Grace SL (2012) Gender and sex
differences in prevalence of major depression in coronary artery disease patients: A metaanalysis. Maturitas 73(4): 305-311
Smith SM, Allwright S and O’Dowd T (2007) Effectiveness of shared care across the interface
between primary and specialty care in chronic disease management. Cochrane database
of systematic reviews (Online) 3(3): CD004910.
Stone JA, Arthur HM and Suskin N (2009) Canadian guidelines for cardiac rehabilitation and
cardiovascular disease prevention: Translating knowledge into action (3rd edition).
Winnipeg, MB: Canadian Association of Cardiac Rehabilitation.
Stromberg R, Backlund LG and Lofvander M (2010) A comparison between the Beck's
Depression Inventory and the Gotland Male Depression Scale in detecting depression
among men visiting a drop-in clinic in primary care. Nordic Journal of Psychiatry 64(4):
258-264.
Thombs BD, Bass EB, Ford DE, Stewart KJ, Tsilidis KK, Patel U, Fauerbach JA, Bush DE and
Ziegelstein RC (2006) Prevalence of depression in survivors of acute myocardial
infarction. Journal of General Internal Medicine 21(1): 30-38.
Thombs BD, Magyar-Russell G, Bass EB, Stewart KJ, Tsilidis KK, Bush DE, Fauerbach JA,
McCann UD and Ziegelstein RC (2007) Performance characteristics of depression
screening instruments in survivors of acute myocardial infarction: Review of the
evidence. Psychosomatics 48(3): 185.
19
Thombs BD, de Jonge P, Coyne JC, Whooley MA, Frasure-Smith N, Mitchell AJ, Zuidersma M,
Eze-Nliam C, Lima BB, Smith CG, Soderlund K and Ziegelstein RC (2008) Depression
screening and patient outcomes in cardiovascular care: A systematic review. Journal of
the American Medical Association 300(18): 2161-2171.
Tully PJ, Winefield HR, Baker RA, Turnbull DA and de Jonge P (2011) Confirmatory factor
analysis of the Beck Depression Inventory-II and the association with cardiac morbidity
and mortality after coronary revascularization. Journal of Health Psychology16(4):584595.
van Melle JP, de Jonge P, Spijkerman TA, Tijssen JG, Ormel J, van Veldhuisen DJ, van den
Brink RH and van den Berg MP (2004) Prognostic association of depression following
myocardial infarction with mortality and cardiovascular events: A meta-analysis.
Psychosomatic Medicine 66(6): 814-822.
World Health Organization (2011) Process of translation and adaptation of instruments.
Available at: http://www.who.int/substance_abuse/research_tools/translation/en/
(accessed 12 April 2011).
Zierau F, Bille A, Rutz W and Bech P (2002) The Gotland Male Depression Scale: A validity
study in patients with alcohol use disorder. Nordic Journal of Psychiatry 56(4): 265-271.
20
Table 2
Social Desirability and Internal Consistency of the Beck Depression Inventory II (BDI-II) and
Gotland Scale of Male Depression (GSMD) Stratified by Sex
Males
Females
Total
Social desirability
(Pearson’s r)
BDI-II*
GSMD*
-0.34
-0.33
-0.24
-0.24
-0.31
-0.30
Internal consistency
(Cronbach’s α)
BDI-II
GSMD
0.92
0.89
0.92
0.90
0.92
0.89
*All correlations have a significance of p < .001.
21
Table 3
Overall Agreement Regarding Frequency of Patients with Elevated Depressive Symptoms
Between the BDI-II and the GSMD and by Sex
Elevated depressive
symptoms
GSMD
All Participants
Elevated depressive symptoms
Subthreshold depressive symptoms
Total
Sensitivity
Specificity
PPV
NPV
Cohen’s К
Males
Elevated depressive symptoms
Subthreshold depressive symptoms
Total
Sensitivity
Specificity
PPV
NPV
Cohen’s К
Females
Elevated depressive symptoms
BDI-II
Subtreshold depressive
symptoms
Total
131
7.5%
156
9.0%
1454
16.5%
45.6%
97.2%
76.6%
90.1%
0.51
40
2.3%
1414
81.2%
287
83.5%
171
9.8%
1570
90.2%
1741
100%
94
7.1%
113
8.6%
207
15.7%
45.4%
97.5%
77.0%
90.6%
0.52
28
2.1%
1087
82.2%
1115
84.3%
122
9.2%
1200
90.8%
1322
100%
37
12
49
8.8%
2.9%
11.7%
Subthreshold depressive symptoms
43
327
370
10.3%
78.0%
88.3%
Total
79
337
416
19.1%
80.9%
100%
Sensitivity
46.3%
Specificity
96.6%
PPV
75.5%
NPV
88.7%
Cohen’s К
0.50
Notes. Depressive symptoms defined as a global score > than 13 in the GSMD and > 14 in the BDI-II. BDI-II =
Beck Depression Inventory-II; GSMD = Gotland Scale for Male Depression; NPV = Negative predictive value, PPV
= Positive predictive value.
22
Table 4
Principal Components Analysis of GSMD and BDI-II Items with Varimax Rotation, and
Individual Item Score Comparisons by Sex
Item
(scale source)
Self perceived failure
(BDI-II)
Worthlessness (BDIII)
Self dislike (BDI-II)
Self criticism (BDI-II)
Guilty feelings (BDIII)
Sadness (BDI-II)
Lost of interest (BDIII)
Indecisiveness (BDIII)
Punishment feelings
(BDI-II)
Suicidal thoughts
(BDI-II)
Pessimism (BDI-II)
Crying (BDI-II)
Tiredness (BDI-II)
Loss of energy (BDIII)
Inexplicable tiredness
(GSMD)
Sleep problems
(GSMD)
Changes in sleeping
pattern (BDI-II)
Loss of interest in sex
(BDI-II)
Loss of pleasure (BDIII)
Concentration
difficulty (BDI-II)
Loss of vitality (BDIII)
Changes in appetite
(BDI-II)
Antisocial behavior
(GSMD)
Factor 1
Maladaptive
cognitions
.72
Factor 2
Psychomotor
fatigue
.16
Factor 3
Factor 4
Males
Females
Externalizing Irritability (mean+SD) (mean+SD)
symptoms
.19
.09
0.26+0.83
0.27+0.62
.71
.30
.18
.11
0.22+0.52
0.26+0.59
.69
.65
.62
.18
.19
.14
.22
.16
.18
.13
.19
.14
0.23+0.55
0.28+0.59
0.19+0.44
0.27+0.60
0.29+0.62
0.22+0.48
.62
.59
.19
.33
.26
.20
.22
.19
0.15+0.40
0.35+0.67
0.19+0.43
0.43+0.80
.59
.33
.17
.16
0.27+0.58
0.35+0.71*
.58
.09
.22
.09
0.15+0.55
0.17+0.64
.58
.06
.20
.14
0.07+0.27
0.09+ 0.33
.58
.50
.19
.23
.31
.10
.80
.79
.19
.01
.12
.11
.20
.27
.15
.10
0.31+0.64
0.16+0.50
0.76+0.72
0.80+0.65
0.33+0.63
0.32+0.76*
0.94+0.76*
0.97+0.67*
.10
.68
.31
.29
0.68+0.82
0.91+0.90*
.12
.58
.17
.38
0.94+0.93
1.20+0.99*
.13
.58
.02
.27
0.81+0.85
1.09+1.01*
.25
.57
.03
-.05
0.73+0.92
1.07+1.20*
.47
.48
.17
.16
0.45+0.66
0.45+0.67
.42
.49
.16
.25
0.42+0.62
0.53+0.70*
.28
.44
.41
.41
0.50+0.75
0.54+0.81
.38
.42
.03
.17
0.42+0.65
0.60+0.79*
.24
.11
.64
.31
0.16+0.43
0.16+0.51
23
Behavioral pessimism
(GSMD)
Abusive behavior
(GSMD)
Complaintiveness
(GSMD)
Hereditary (GSMD)
Ambivalence (GSMD)
Worse in the morning
(GSMD)
Aggressiveness
(GSMD)
Irritability (GSMD)
Lower stress threshold
(GSMD)
Irritability (BDI-II)
Agitation (BDI-II)
.40
.18
.64
.26
0.22+0.54
0.27+0.60
.10
.06
.63
.10
0.19+0.50
0.16+0.50
.31
.19
.59
.22
0.16+0.45
0.19+0.48
.31
.34
.27
.18
.36
.31
.59
.42
.41
.23
.25
.38
0.14+0.46
0.31+0.62
0.34+0.65
0.17+0.54
0.36+0.68
0.40+0.74
.13
.12
.28
.76
0.39+0.66
0.39+0.69
.27
.18
.33
.24
.32
.25
.69
.68
0.52+0.75
0.57+0.76
0.57+0.81
0.65+0.79
.38
.41
.24
.17
.08
-.02
.61
.56
0.29+0.60
0.24+0.51
0.32+0.59
0.29+0.58
Proportion of variance
explained
18.92%
13.91%
10.65%
9.34%
Eigenvalue
6.43
4.73
3.62
3.18
Notes.*significant difference at p <.0125 (t-test). Scores range from 0-3 for all items of both depression scales.
BDI-II = Beck Depression Inventory-II, GSMD = Gotland Male Depression Scale.
24