VU Research Portal
Onset and recurrence of depressive disorders Dijkstra-Kersten, S.M.A.
2019
document version
Publisher's PDF, also known as Version of record
Link to publication in VU Research Portal
citation for published version (APA)
Dijkstra-Kersten, S. M. A. (2019). Onset and recurrence of depressive disorders: Contributing factors and prevention.
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ?
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
E-mail address:
vuresearchportal.ub@vu.nl
Download date: 13. Oct. 2021
CONTENTS
Chapter 1 General introduction 9
PART I Onset, presence and recurrence of depressive disorders: contributing factors
Chapter 2 Associations of financial strain and income with depressive 23 and anxiety disorders
Chapter 3 Somatization as a risk factor for incident depression 39 and anxiety
Chapter 4 Longitudinal associations of multiple physical symptoms 53 with recurrence of depression and anxiety
PART II Measurement of somatization
Chapter 5 Systematic review of measurement properties of 69 questionnaires measuring somatization in primary care
patients
PART III Prevention of recurrent depression
Chapter 6 Effectiveness of supported self-help in recurrent depression: 117 a randomized controlled trial in primary care
Chapter 7 Supported self-help for recurrent depression in primary care: 137 Who benefits most?
Chapter 8 General discussion 153
References 173
Summary (English) 191
Samenvatting (Nederlands) 197
Dankwoord 203
About the Author 207
List of publications 209
Associations of financial strain and income with depressive and anxiety disorders
Dijkstra-Kersten SMA Biesheuvel-Leliefeld KEM van der Wouden JC Penninx BWJH van Marwijk HWJ
Published in:
J Epidemiol Community Health 2015;69:660-665
Chapter 2
24 Chapter 2
ABSTRACT Background
Previous research has shown socio-economic inequality in prevalence and onset of depressive disorders. It is not yet clear whether perceived financial strain is associated with depressive and/or anxiety disorders, in addition to an objective indicator such as income. This study examines whether financial strain is associated with the prevalence and onset/recurrence of depressive and/or anxiety disorders, above income.
Methods
Data are from the Netherlands Study of Depression and Anxiety. Associations between financial strain, income and presence of depressive and/or anxiety disorder at baseline were assessed among 2937 subjects (18-65 years). Impact of financial strain and income on 4-year onset/recurrence of depressive and/or anxiety disorders were examined among 1250 subjects without a depressive or anxiety disorder at baseline. Depressive and anxiety disorders were determined with the Composite-International-Diagnostic-Interview. Financial strain and income were assessed in an interview.
Results
Subjects with mild or severe financial strain had higher odds of being depressed (OR=1.68, 95%CI=1.35 - 2.09; OR=3.88, 95%CI=2.58 - 5.81) or remitted (OR=1.56, 95%CI=1.24 - 1.96; OR=1.99, 95%CI=1.27 - 3.11) at baseline compared to healthy controls, after adjusting for income. Mild or severe financial strain was not associated with onset/recurrence of depressive and/or anxiety disorders during follow-up (OR=1.08, 95%CI=.83 - 1.42;OR=1.05, 95%CI=.64 - 1.73).
Conclusion
Financial strain was associated with having a depressive and/or anxiety disorder,
above the effect of income. Healthcare and social services should be alert to
this association, even for higher income households. However, financial strain
and income were not related with 4-year onset/recurrence of depressive and/or
anxiety disorders.
2
25 Financial strain, depression and anxiety
INTRODUCTION
Due to the current economic recession, more households have difficulties making ends meet (Statistics Netherlands 2013). Socio-economic characteristics such as low income, low education level, debt and unemployment are generally associated with higher rates of having mental health problems (Fryers et al. 2003;
Lorant et al. 2003; Jenkins et al. 2008). These are objective indicators of a person’s socio-economic status. In previous research, these indicators have been linked with poor mental health, and the strength of these associations vary with the type of indicator used (Fryers et al. 2003; Lorant et al. 2003; Lorant et al. 2007;
Wang et al. 2010; Kosidou et al. 2011. A systematic review by Fryers et al. (2003) for instance concluded that less education, lower income and unemployment provided substantial more evidence for social inequalities in prevalence of mental disorders than social status. A meta-analysis by Lorant et al. (2003) indicated a stronger relation of depression with income than with education. Subjective indicators however, such as perceived financial strain and subjective social class, have seldom been used. Perceived financial strain (henceforth called: financial strain) for instance might be a better predictor of mental health problems (Weich & Lewis 1998a; Butterworth et al. 2009). Financial strain may represent a tendency to worry (Weich & Lewis 1998b) and is not restricted to subjects with a low socio-economic status; subjects in a high income group or high socio- economic status may also perceive financial strain, for example as a result of an inappropriate standard of living (Wang et al. 2010). Financial strain could therefore be considered as a negative stressor which may increase the risk of mental health problems (Wang et al. 2010).
The current large scale study is one of the first to examine the effect of financial strain, as a subjective indicator of socio-economic status, and income, as an objective indicator of socio-economic status, on depressive and/or anxiety disorders in a multisite naturalistic cohort (Penninx et al. 2008). So far, only few studies focused on the effect of financial strain on depressive and/or anxiety disorders. They indicate that financial strain may be associated with an increased risk of depressive symptoms (Lorant et al. 2007; Dunn et al. 2008; Okechukwu et al. 2012), depressive disorder (Lorant et al. 2007; Wang et al. 2010) or common mental disorders (Weich
& Lewis 1998b). Only one study (Wang et al. 2010) used standardized clinical
assessments, while others used symptom-checklists or general practitioner
records. Furthermore, none of these studies examined anxiety disorders, although
anxiety disorders frequently co-occur with depressive disorders. Because of
these limitations and different methodologies, the exact relationships between
financial strain, income and depressive and/or anxiety disorders are still not clear.
26 Chapter 2
In order to examine the impact of financial strain on the prevalence and onset/
recurrence of depressive and/or anxiety disorders and to assess whether this effect is over and above the effect of income, this study used cross-sectional and longitudinal data. We hypothesized that perceiving more financial strain is associated with 1) higher prevalence and 2) higher 4-year onset/recurrence of depressive and/or anxiety disorders. It is furthermore hypothesized that the associations between financial strain and 3) the prevalence and 4) onset/
recurrence of depressive and anxiety disorders are independent from baseline income.
METHODS Study sample
The present study is part of the Netherlands Study of Depression and Anxiety (NESDA), an on-going naturalistic multisite cohort study which investigates predictors, long-term course and consequences of depressive and anxiety disorders (Penninx et al. 2008). A total of 2981 subjects (18- 65 years) were recruited from community, general practice and secondary mental health care. The sample consisted of healthy controls, persons with a prior history of depression and/or anxiety and persons with a current depressive and/or anxiety disorder. Exclusion criteria were not being fluent in Dutch and a primary diagnosis of psychotic, obsessive compulsive, bipolar, or severe substance use disorder.
Baseline data collection took place between 2004 and 2007. Assessments were repeated every two years. The NESDA design is described elsewhere in more detail (Penninx et al. 2008). The research protocol was approved by the Ethical Committee of participating universities and has been performed in accordance with the ethical standards of the Declaration of Helsinki.
We used two samples from the NESDA-cohort. The first sample consisted of all subjects (N = 2981), regardless of their psychiatric status at baseline. In this sample, the association between financial strain and depressive and/or anxiety disorders was assessed (hypothesis 1) and whether this association was independent from income (hypothesis 3). Only data from the baseline assessment were used.
Forty-four subjects who did not provide all required data were excluded from
analyses (N = 2937). The second sample was restricted to 1525 subjects without a
current depressive or anxiety disorder at baseline (see ‘depressive and/or anxiety
disorders’ for definition) in order to examine the effect of financial strain on the4-year new onset/recurrence of depressive and/or anxiety disorders (hypothesis
2
27 Financial strain, depression and anxiety
2) and whether this effect was over and above the effect of income (hypothesis 4). Of these eligible subjects, 1299 participated in the 2-year and 4-year follow-up assessments, while 1250 of those provided all required data, and were used for our longitudinal analyses.
Measurements
Financial characteristics
At baseline, data on income and financial strain were collected in a face-to- face interview. The net household income per month was measured with 24 categories ranging from ‘less than 600 euro per month’ to ‘more than 5000 euro per month’ with steps of 300 euro. This variable was transformed to (approximate) interval level ranging from 500 euro to 6200 euro per month by replacing each category by its midpoint value. Financial strain was measured with the question:
‘In general: how is your financial status at the end of the month?’ Possible answers were: ‘usually money left’ (considered as no financial strain; reference category),
‘just enough money to manage’ (mild financial strain) and ‘not enough money to manage’ (severe financial strain).
Depressive and/or anxiety disorders
Depressive (i.e. major depressive disorder and dysthymia) or anxiety (i.e.
agoraphobia, social phobia, panic and generalized anxiety disorder) disorders were diagnosed at baseline, 2-year and 4-year follow-up with the Composite International Diagnostic Interview (CIDI, version 2.1; Wittchen 1994). If subjects reported any diagnosis, a Life Chart Interview (LCI; Lyketsos et al. 1994) was conducted. Life events were recalled to refresh memory using a calendar method, after which depressive and/or anxiety symptoms in each month were determined retrospectively. In addition, symptom severity was assessed for each month. Symptoms reported during LCI were only considered present when at least of mild severity.
The presence of a depressive and/or anxiety disorder at baseline was considered as the outcome for the first and third hypothesis (cross-sectional analysis).
Presence of a depressive and/or anxiety disorder at baseline was defined
according to the CIDI and LCI as all subjects with 1) a diagnosis of a depressive
and/or anxiety disorder in the previous month or 2) a diagnosis of a depressive
and/or anxiety disorder in the previous six months and having had substantial
depressive or anxiety symptoms in the year previous to baseline. Three categories
were created for this outcome measure: 1) healthy control: neither a lifetime nor
current depressive and/or anxiety disorder, 2) remitted; lifetime depressive and/
28 Chapter 2
or anxiety disorder but no current disorder, 3) current depressive and/or anxiety disorder.
Based on the CIDI, the onset of a depressive and/or anxiety disorder (regardless of the presence of an earlier diagnosis before baseline; the term ‘onset’ is used for both new onset and recurrence of depressive and/or anxiety disorders) anytime between baseline and the 4-year follow-up was used for our second and forth hypotheses (longitudinal analysis).
Socio-demographic characteristics
Socio-demographic characteristics included were: age, gender, number of years of education, partner status (yes/no) and employment status.
Statistical analyses
Baseline characteristics were compared across baseline psychiatric status using chi-square statistics and analyses of variance.
We used multinomial logistic regression analysis to assess the associations between financial strain, income and the presence of a depressive and/or anxiety disorder at baseline (analysis 1: cross-sectional). Subjects who had never had a depressive and/or anxiety disorder were used as reference group and were compared to subjects who were remitted and to subjects who had a current depressive and/or anxiety disorder at baseline. Logistic regression analysis was used to examine the effect of financial strain and income on the onset/recurrence of depressive and/or anxiety disorders during the 4-year follow-up (analysis 2:
longitudinal).
All analyses comprised three steps. Model 1 assessed the univariate associations
of income and financial strain with depressive and/or anxiety disorders. In model
2, both income and financial strain were included as main predictors. In model
3, we adjusted for socio-demographic characteristics (i.e. age, gender, education
level and partner status). Also, we checked if the presence of a lifetime diagnosis
of depressive and/or anxiety disorder modified the association between financial
strain and the onset/recurrence of depressive and/or anxiety disorders. This
interaction was not significant (p > .10) so results were not stratified. Furthermore,
no significant interaction-effect was found for financial strain and income in
analysis 1 and analysis 2.
2
29 Financial strain, depression and anxiety
RESULTS
Cross-sectional analysis
Baseline characteristicsAt baseline, of the sample of 2937 subjects with all relevant data, 308 subjects (10.5%) had a current depressive disorder, 553 subjects (18.8%) had an anxiety disorder and 571 subjects (19.4%) had both a depressive and anxiety disorder.
Socio-demographic and financial characteristics of subjects at baseline are shown in Table 1. Mean age was 41.9 (SD = 13.0) years and 66% of the subjects were female.
Financial strain and income at baseline were negatively but not strongly correlated (r = -.41, p < .001). Correlations between financial strain and income were not different for subjects with a current depressive and/or anxiety disorder compared to subjects without a depressive and/or anxiety disorder at baseline. Subjects with mild financial strain or severe financial strain reported lower income (median 1700 and 1300 Euros/month, respectively) than subjects with no financial strain (median 2700 Euros/month).
Associations between financial strain and prevalence of depressive and/or anxiety disorders
Subjects with financial strain were more likely (Table 2) to have a depressive and/
or anxiety disorder at baseline, compared to subjects with no financial strain (mild financial strain: OR = 1.68, 95%CI = 1.35 – 2.09; severe financial strain: OR = 3.88, 95%CI = 2.58 – 5.81, respectively). Also, compared to healthy controls, subjects with mild or severe financial strain had higher odds of being remitted at baseline than subjects with no financial strain (Table 3: OR = 1.56, 95%CI = 1.24 – 1.97; OR
= 1.96, 95%CI = 1.25 – 3.07, respectively).
Associations of financial strain with depressive and/or anxiety disorders remained
while taking income into account. Furthermore, post-hoc subgroup analyses
in which associations between financial strain and depressive and/or anxiety
disorders were stratified according to income showed positive associations in all
three income groups. Regardless of income level, subjects with mild or severe
financial strain were more likely to have a depressive and/or anxiety disorder at
baseline compared to subjects with no financial strain.
30 Chapter 2
Table 1 | Baseline characteristics of subjects according to psychiatric status at baseline (N = 2937) Depressive and/or anxiety disorder at baseline (n = 1432) Remitted from depressive and/or anxiety disorder (n = 862 )
No lifetime or current depressive and/or anxiety disorder (healthy control) (n = 643)All subjects (N = 2937) Baseline characteristicsM(SD) or %vs. remittedvs. healthy controlM(SD) or %vs. healthy control M(SD) or %M(SD) or % Gender: female65.9p = .060p = .02569.7p <.00160.865.9 Age in years41.9 (12.2)p = .193p = .22742.6 (13.0)p = .03841.1 (14.6)41.9 (13.0) Partner status: married or partner64.6p <.001p <.00172.9p = .35875.069.3 Education in years11.6 (3.25)p <.001p <.00112.6 (3.20)p = .09412.8 (3.20)12.2 (3.3) Employment status: Working Early retirement, sickness benefit and others Not working
54.3 11.9 33.7
p <.001p <.00170.0 6.0 24.0
p = .59472.0 6.2 21.8
62.8 9.0 28.3 Income of household in Euros, median1900p <.001p <.0012100p = .00125002100 Financial status at end of the month Usually money left Just enough money to manage Not enough money to manage
42.0 40.9 17.0p <.001p <.00149.8 41.5 8.7p <.00163.5 31.6 5.0
49.0 39.1 12.0
2
31 Financial strain, depression and anxiety
Table 2 | Association between depressive and/or anxiety disorders, financial strain and income at baseline for current patients versus healthy controls (N = 2937)
Current patients versus healthy controls
Model 1 Model 2 Model 3
OR (95% CI) OR (95% CI) OR (95% CI)
Financial strain:
None 1.00 1.00 1.00
Mild 1.96 (1.60 – 2.40)** 1.71 (1.38 – 2.11)** 1.68 (1.35 – 2.09)**
Severe 5.17 (3.50 – 7.63)** 4.19 (2.80 – 6.26)** 3.88 (2.58 – 5.81)**
Income x 1000 Euros .74 (.69 - .80)** .84 (.77 - .91)* .90 (.82 - .98)*
** p < .01, * p < .05
Model 1: univariate, unadjusted Model 2: multivariate, unadjusted
Model 3: adjusted for gender, age, partner status and education level.
Statistics final model: Nagelkerke R2 = .08; χ2(14) = 220.36, p < .001.
Table 3 | Association between depressive and/or anxiety disorders, financial strain and income at baseline for remitted patients versus healthy controls (N = 2937)
Remitted patients versus healthy controls
Model 1 Model 2 Model 3
OR (95% CI) OR (95% CI) OR (95% CI)
Financial strain:
None 1.00 1.00 1.00
Mild 1.68 (1.35 – 2.09)** 1.56 (1.24 – 1.96)** 1.56 (1.24 – 1.97)**
Severe 2.23 (1.44 – 2.45)** 1.99 (1.27 – 3.11)** 1.96 (1.25 – 3.07)**
Income x 1000 Euros .85 (.79 - .93)** .91 (.83 - .99)* 0.89 (.81 - .98)*
** p < .01, * p < .05
Model 1: univariate, unadjusted Model 2: multivariate, unadjusted
Model 3: adjusted for gender, age, partner status and education level.
Statistics final model: Nagelkerke R2 = .08; χ2(14) = 220.36, p < .001.
Additional post-hoc multinomial regression analysis (results not in table) indicated that the association with financial strain was stronger for having both a depressive and an anxiety disorder (mild financial strain: OR = 1.40, 95%CI = 1.11 – 1.77; severe financial strain: OR = 3.59, 95%CI = 2.60 – 4.96) than for depressive disorder only (mild financial strain: OR = 1.17, 95%CI = .88 – 1.56; severe financial strain: OR= 2.11, 95%CI = 1.39 – 3.20) or anxiety disorders only (mild financial strain: OR = 1.29, 95%CI = 1.03 – 1.61; severe financial strain: OR = 1.83, 95%CI
= 1.28 – 2.62).
32 Chapter 2
Longitudinal analysis
During the 4-year follow-up, 402 of the 1250 subjects (32.2%) without a current depressive and/or anxiety disorder at baseline experienced a depressive and/or anxiety disorder: 13.5% experienced a depressive disorder, 7.9% experienced an anxiety disorder and 10.7% of subjects experienced both a depressive and anxiety disorder.
Associations of financial strain with the 4-year onset/recurrence of depressive and/or anxiety disorders
Subjects who reported mild or severe financial strain at baseline were not more likely to become depressed or anxious during follow-up, compared to subjects without financial strain (Table 4: OR = 1.08, 95%CI = .83 – 1.42; OR = 1.05, 95%CI
= .64 – 1.73, respectively). Income was however negatively associated with the onset/recurrence of depressive and/or anxiety disorders (OR = .87, 95%CI = .78 - .97). After adjusting for socio-demographic characteristics, this association was no longer statistically significant (p = .065). Post-hoc analysis in which results were stratified according to three income levels showed comparable results:
financial strain was not related with the onset/recurrence of depressive and/or anxiety disorders in either income category.
Table 4 | Associations of financial strain and income with the onset/recurrence of depressive and/or anxiety disorders during the 4-year follow-up (n = 1250)
Onset/Recurrence of depressive and/or anxiety disorders
Model 1 Model 2 Model 3
OR (95% CI) OR (95% CI) OR (95% CI)
Financial strain:
None 1.00 1.00 1.00
Mild 1.31 (1.02 – 1.68)* 1.18 (.90 – 1.53) 1.08 (.83 – 1.42) Severe 1.32 (0.82 – 2.12) 1.13 (.69 – 1.84) 1.05 (.64 – 1.73) Income x 1000 Euros .85 (.77 - .94)** .87 (.78 – .97)* .89 (.79 – 1.01) †
** p < .01, * p < .05, † p < .10 Model 1: univariate, unadjusted Model 2: multivariate, unadjusted
Model 3: adjusted for gender, age, partner status and education level.
Statistics final model: Nagelkerke R2 = .041; χ2(7) = 37.32, p < .001.
We reran the analyses for the 2-year outcomes and found comparable results:
The final model showed that subjects with mild or severe financial strain at
baseline did not have higher odds of becoming depressed or anxious during the
2
33 Financial strain, depression and anxiety
2-year follow-up (OR = 1.14, 95%CI = 0.85 – 1.53; OR = 1.13, 95%CI = 0.66 – 1.93 respectively). Again, the association between income and the onset/recurrence of depressive and/or anxiety disorders was not significant (OR = .90, 95%CI = .79 – 1.03).
DISCUSSION Main findings
We examined associations of financial strain and income with the prevalence and onset/recurrence of depressive and/or anxiety disorders in a large multisite cohort. The results indicated that financial strain and income were cross- sectionally associated with depressive and/or anxiety disorders (hypothesis 1).
This is consistent with previous studies (Fryers et al. 2003; Lorant et al. 2003;
Jenkins et al. 2008; Wang et al. 2010; Butterworth et al. 2009; Okechukwu et al. 2012). However, financial strain and income were not related with the onset/
recurrence of depressive and/or anxiety disorders during the 4-year follow-up.
Hence, hypothesis 2 was not accepted. Previous research showed mixed results of the impact of financial strain on the onset of depressive and anxiety disorders.
For example, Wang et al. (2010) found that financial strain was associated with an increased risk of a depressive disorder for subjects who worked in the past 12 months
1, while Butterworth et al. (2009) found no effect of financial strain after adjusting for depressive symptoms at baseline. Conversely, a cohort- study in the UK (Weich & Lewis 1998b) indicated that financial strain was independently associated with both onset and persistence of common mental disorders at 12-months follow-up. Moreover, post-hoc sensitivity analyses, in which employment status at baseline was included in the model, showed that employment status at baseline had no effect on the association between financial strain and the presence and onset/recurrence of depressive and/or anxiety disorders (i.e. <10% change in the regression-coefficient of the effect of financial strain on depressive and/or anxiety disorders), although employment status differed considerably according to psychiatric status at baseline.
Although financial strain was cross-sectionally associated with depressive and/
or anxiety disorders, according to our findings it does not predict the onset/
recurrence of depressive and anxiety disorders. An explanation may be that a different type of relationship is involved: instead of a causal effect of financial strain on depressive and/or anxiety disorders, there may be reverse causation,
1 Additional analysis showed that employment status did not interact with financial strain in the longitudinal analysis.
34 Chapter 2
which postulates that subjects might experience more financial difficulties as a result of their depressive and/or anxiety disorder. For instance, subjects with a depressive and/or anxiety disorder are a vulnerably group on the labor market;
they might become unemployed more easily, might be discriminated in job interviews, or might have had fewer career opportunities (Waghorn & Chant 2006; Edward & Munro 2008). This unfavorable position on the labor market also manifested in our study; subjects with a depressive and/or anxiety disorder were more often unemployed than subjects without a depressive and/or anxiety disorder. Furthermore, depressed or anxious subjects could be more prone to worry about their financial status. Secondly, our findings might be explained by the large time interval in which the effect of baseline financial strain on depressive and/or anxiety disorders was assessed. The effect of financial difficulties on depression fades over time, unless the financial difficulties persist (Butterworth et al. 2009; Mirowsky et al. 2001). The association between depression and financial strain will therefore decline as the interval between the measurements increases. It might also explain differences in findings between studies. Our study and studies by Butterworth et al. (2009) and Wang et al. (2010) used a follow-up of four to six years, while Weich and Lewis’ study (1998b) used a 1-year follow-up.
However, we checked the association between financial strain and the onset/
recurrence of depressive and/or anxiety episodes for the 2-year data and found comparable results.
Furthermore, this study examined whether the associations of financial strain with the prevalence and onset/recurrence of depressive and/or anxiety disorders were above the effect of income (hypotheses 2 and 4). Hypothesis 2 was confirmed: the association between financial strain and having a depressive and/
or anxiety disorder held after adjusting for income. Additionally, financial strain and income were only moderately correlated. Financial strain and income appear to be separate constructs. Since financial strain and income were not related with onset/recurrence of depressive and/or anxiety disorders, no conclusions could be drawn about the effect of financial strain over and above the effect of income;
hypothesis 4 could not be accepted.
Strengths and limitations
The first strength of this study is the large naturalistic cohort-study among almost
3000 subjects (NESDA), on which data the current study reports. This sample
consists of persons with a current depressive and/or anxiety disorder, a history
of depressive and/or anxiety disorder or at risk because of a family history, and
healthy controls. Participants were selected from general population, general
2
35 Financial strain, depression and anxiety
practices and mental health organizations. It therefore reflects various settings and developmental stages of psychopathology. Furthermore, the attrition rate of NESDA was relatively low: 85% of eligible subjects at baseline participated in the 4-year follow-up assessment, while 81% provided all required data.
A second strength is that onset/recurrence of depressive and/or anxiety disorders were assessed during the whole 4-year follow-up period by combining data from the 2-year and 4-year follow-up assessments. It is therefore unlikely that cases of depressive and/or anxiety disorder were missed. Other studies (Weich & Lewis 1998b; Butterworth et al. 2009; Wang et al. 2010) did not cover the entire follow- up period, which may have caused biased results for cases of depressive and/
or anxiety disorders may be missed. Also, the second analysis primarily focused on onset/recurrence of new depressive and/or anxiety disorders by excluding subjects who were depressed at baseline. Third, both subjective and objective indicators of subjects’ socio-economic status were used so their associations with depressive and/or anxiety disorders could be compared.
A limitation is that financial circumstances were only assessed at baseline. A change in financial strain or income during the 4-year follow-up could not be determined. According to Lorant et al. (2007), a change in socio-economic status was associated with a change in depressive symptoms and depressive disorders. Butterworth et al. (2009) for example found that although subjects who experienced financial strain at baseline did have elevated risk of depression after 4-year follow-up, subjects who experienced financial strain more recently had greater risks of being depressed after follow-up. These findings might explain why we only found a cross-sectional association of financial strain with depressive and/or anxiety disorders and no association with onset/recurrence of depressive and/or anxiety disorders during the 4-year follow-up.
Another point of consideration is the combining of onset and recurrence of
depressive and/or anxiety disorders into one variable in the longitudinal analysis
instead of treating these separately. Assessing onset and occurrence separately
however, would have decreased our sample size considerably and would have
had a negative impact on the robustness of our analyses since, for instance, the
number of participants with (severe) financial strain who became depressed or
anxious during follow-up would be too small. Besides, we found that presence of a
lifetime depressive or anxiety disorder did not significantly modify the association
between financial strain and depressive and/or anxiety disorders.
36 Chapter 2
Implications and further research
This study showed, not unexpectedly, that persons who experience financial strain have higher odds of being depressed or anxious. This may have implications for mental healthcare services and social services. Perhaps even more attention should be paid to this association so that depressed or anxious subjects who experience financial strain, regardless of their income level, can get more support.
A treatment program to target specific high-risk groups for financial strain and depressive and/or anxiety disorder might be an important innovation.
The current study did not find significant associations of financial strain and income with the onset/recurrence of depressive and/or anxiety disorders during the 4-year follow-up. However, a meta-analysis (Lorant et al. 2003) suggested that in subjects with low socio-economic position, persisting depression may be a bigger problem than new depression. Further research could therefore explore the impact of financial strain and income on the persistence of depressive and/
or anxiety disorders. Also, more longitudinal cohort-studies that investigate both financial characteristics and mental disorders repeatedly are needed to study the question of causality. It is not yet fully clear whether financial difficulties predict mental disorders, whether mental disorder predicts financial difficulties, or whether there is a vicious cycle between financial difficulties and mental disorders.
CONCLUSION
Experiencing financial strain is associated with the presence of depressive and/
or anxiety disorders, over and above the effect of income. This finding suggests
that when a person experiences a depressive and/or anxiety disorder, a subjective
indicator like financial strain should be taken into account besides an objective
indicator like income. Financial strain and income however, were not related with
the onset/recurrence of depressive and/or anxiety disorders.
Somatization as a risk factor for incident depression and anxiety
Dijkstra-Kersten SMA Sitnikova K
van Marwijk HWJ Gerrits MMJG van der Wouden JC Penninx BWJH van der Horst HE Leone SS
Published in:
J Psychosom Res 2015;79(6):614-619.
Chapter 3
40 Chapter 3
ABSTRACT Objective
In this study we aim to examine somatization as a risk factor for the onset of depressive and anxiety disorders.
Methods
4-year follow-up data from the Netherlands Study of Depression and Anxiety (NESDA), a multisite cohort study of the course of depression and anxiety, were analysed. Participants (18-65 years) without a lifetime depressive or anxiety disorder at baseline were included (n = 611). Somatization was measured at baseline with the somatization subscale of the 4 Dimensional Symptoms Questionnaire. Onset of depression and anxiety was assessed with the CIDI interview at 2-year and 4-year follow-up.
Results
Somatization was a risk factor for the incidence of depression [Hazard Ratio per unit increase (HR); 95% Confidence Interval (CI): 1.13; 1.09 - 1.17] and anxiety [HR; 95% CI: 1.14; 1.09 - 1.18]. Associations attenuated but remained statistically significant after adjusting for socio-demographic characteristics, chronic disorders and baseline levels of (subclinical) depressive or anxiety symptoms [adjusted HR for depression; 95% CI: 1.06; 1.00 - 1.12, adjusted HR for anxiety; 95% CI: 1.13;
1.07 - 1.20].
Conclusion
Persons who somatize have an increased risk of becoming depressed or anxious
in subsequent years, over and above baseline levels of depressive or anxiety
symptoms. They may represent a target group for prevention of depressive and
anxiety disorders.
41 Somatization and onset of depression and anxiety
3
INTRODUCTION
Somatization is a common phenomenon in primary care and is defined as “the tendency to experience and communicate somatic distress and symptoms unaccounted for by pathological findings, to attribute them to physical illness, and to seek medical help for them” (Lipowski 1988). While experiencing some physical symptoms unaccounted for by pathological findings is a common phenomenon, experiencing multiple unexplained physical symptoms from different organ systems often implies somatization (Terluin et al. 2006). Physical symptoms, such as fatigue, dizziness and pain, are prevalent, often co-occurring and can range in severity with syndromes such as somatic symptom disorder and functional somatic syndromes (e.g. fibromyalgia) at the severe end of the clinical spectrum to self-limiting symptoms. They are especially prevalent in people who have mental health problems, such as depressive and anxiety disorders (Simon 1999; Kroenke 2003). Somatization could also constitute an important risk factor for the occurrence of these disorders and, therefore, persons who somatize may represent a target group for prevention of depressive and anxiety disorders, which continues to be an important public health goal (Lau & Rapee 2011;
Munoz et al. 2010; Tylee & Ghandi 2005). However, little is known about their actual risk to develop a depressive or anxiety disorder as to date most studies that have investigated this association have been cross-sectional, thus precluding conclusions about temporality (Kroenke 2003; Bair et al. 2003; Lieb et al. 2007).
Examining the relationship between somatization and the onset of depressive or anxiety disorders is important as the co-occurrence of somatization with mental health problems affects health outcomes, functioning and economic costs, thus increasing the burden of disease. The presence of somatization is associated with less improvement of depressive and anxiety symptoms, and is predictive of a poorer treatment response in depressed patients (Huijbregts et al. 2010).
In addition, the co-occurrence of symptoms of somatization and depression is associated with decreased physical, social and occupational functioning and an increase of health care services use (Tylee & Ghandi 2005).
Despite its obvious relevance for health care, few studies have investigated somatization as a predictor for the onset of depressive or anxiety disorders. Data from a primary care study showed that specific unexplained symptoms (e.g.
fatigue, pain or dizziness) presented to the GP were not predictive of subsequent
depressive or anxiety disorder after three months as recognised and registered by
GPs, although the presence of unexplained symptoms increased the concurrent
risk of depression or anxiety at least three-fold (van Boven et al. 2011). However,
in a community study somatization predicted subsequent depressive symptoms
42 Chapter 3
at 5-year follow-up in non-depressed women, but was not predictive in men, after correction for depressive symptoms at baseline (Terre et al. 2003). In addition, there is some evidence from community and primary care studies focusing on specific medically unexplained physical symptoms (e.g. fatigue) that these symptoms can increase the risk of subsequent psychiatric symptoms and disorders, including depression and anxiety (Skapinakis et al. 2004; van der Linden et al. 1999; Addington et al. 2001). Although some of these studies focused on medically unexplained physical symptoms, it is not clear whether these persons with medically unexplained symptoms somatize.
Besides focusing on a specific symptom, most previous studies either did not take psychiatric history into account, thus could not discriminate between incidence and recurrence, or did not use a clinical interview but rather self-report questionnaires to determine the presence of a disorder at baseline and follow- up. Therefore, in this study we aim to examine whether somatization is a risk factor for the subsequent onset of an incident depressive or anxiety disorder as determined by a clinical interview in a longitudinal prospective cohort study.
METHODS Design
This study used 4-year prospective data from the Netherlands Study of Depression and Anxiety (NESDA). The overall aim of the NESDA study (Penninx et al. 2008) is to examine the aetiology, course and consequences of depression and anxiety using biological and psychosocial research paradigms within an epidemiological framework. At baseline (2004-2007) 2981 participants aged 18 through 65 years were included in the study. The sample consisted of healthy controls;
persons with a prior history of depression and/or anxiety; persons with a current
depressive and/or anxiety disorder and persons at risk because of a family history
or subthreshold depressive or anxiety symptoms. Participants were recruited in
different settings: the community (n = 564), primary care (n = 1610) and mental
health services (n = 807). Exclusion criteria were the presence of a primary
psychiatric diagnosis other than depression or anxiety, which might interfere
with the course trajectory (i.e. psychotic disorder, obsessive compulsive disorder,
bipolar disorder or severe addiction disorder), and lack of fluency in the Dutch
language. Baseline and follow-up assessments consisted of a clinical interview,
written questionnaires and physical measurements. The baseline, 2-year, and
4-year follow-up measurements were used in this study. Further details on the
43 Somatization and onset of depression and anxiety
3
selection procedure and design of the NESDA study are provided elsewhere (Penninx et al. 2008). The research protocol was approved by the ethical review boards of the VU University medical center and all other participating institutes, and all respondents provided written informed consent.
Study population
Participants who never had an episode of a depressive disorder (major depressive disorder or dysthymia) or anxiety disorder (social phobia, panic with agoraphobia, panic without agoraphobia, agoraphobia, generalised anxiety disorder) were included in this study (n = 652). Of these participants, 614 (94%) participated in at least one follow-up measurement. Information on the central determinant (i.e.
somatization) was available for 611 of these remaining participants. These 611 participants represent our study sample.
Measures
Onset of depressive and anxiety disorders
The Composite International Diagnostic Interview (CIDI, lifetime version 2.1;
Wittchen 1994) was administered at the 2-year and 4-year follow-up assessments to determine the incidence (yes/no) of a depressive or anxiety disorder during the 4-year follow-up period. The CIDI is a highly reliable and valid assessment tool for depressive and anxiety disorders as defined by the DSM-IV criteria (Wittchen 1994).
Incident depression or anxiety was defined as the occurrence of a depressive or anxiety disorder at any point during the follow-up period.
When diagnosed with a depressive or anxiety disorder at either the 2-year or
4-year follow-up measurements, participants were asked to indicate the time of
onset of their disorder: less than two weeks ago, two weeks to one month ago,
one to six months ago, six to twelve months ago, in the last twelve months,
or more than one year ago. Time-at-risk was calculated from baseline until
the moment a participant had an event (i.e. onset of a depressive or anxiety
disorder), by taking the midpoint of each time interval for onset. For instance,
no depressive disorder was diagnosed at 2-year follow-up, while at the 4-year
follow-up measurement, a participant reported the time of onset of a depressive
disorder between six and twelve months ago. For this participant, time-at-risk
was set to 48 months - 9 months = 39 months. Time-at-risk was censored at 48
months when the participant had not developed a depressive or anxiety disorder
during the follow-up period. For participants who did not take part in the 4-year
follow-up measurement, time-at-risk was calculated over the 2-year follow-up.
44 Chapter 3
Participants who did not take part in the 2-year follow-up measurement, but did take part in the 4-year follow-up measurement, were asked about onset and recency of depressive and anxiety disorders over a 4-year period at the 4-year follow-up measurement and time to onset was calculated accordingly.
Somatization
The somatization subscale of the four dimensional symptoms questionnaire (4DSQ; Terluin et al. 2006) was used to measure somatization in NESDA. While experiencing one or few physical symptoms is considered normal, experiencing multiple unexplained physical symptoms from different organ systems often indicates somatization (Lipowski 1988; Terluin et al. 2006).The 4DSQ somatization subscale operationalises somatization as a high number and frequency of physical symptoms and has been validated against various measures (e.g. somatization subscale of the SCL-90; GP’s diagnoses) and in varying samples, including psychiatric and primary care samples (Terluin et al. 2006). This subscale consists of 16 symptoms (e.g. during the past week did you suffer from dizziness; painful muscles; headache), which are scored on a 5-point Likert scale (‘no’, ‘sometimes’,
‘regularly’, ‘often’, and ‘very often or constantly’). Responses were recoded (‘no’
= 0; ‘sometimes’ = 1; ‘regularly’ to ‘very often or constantly’ = 2) and item scores were summated to calculate scale scores (range 0-32). An overall score of 11 or higher is considered to be an indication of an elevated level of somatization as well as an elevated risk of impaired functioning. A score of 21 or higher indicates a highly elevated score and a very high risk of impaired functioning (Terluin et al.
2006).
Demographic variables
Age, gender, number of years of education, and work status were assessed by self-report questionnaires.
Chronic disorders
We also adjusted for the potential confounding effects of chronic disorders (yes/
no) in our analyses as these may be related to both the reporting of somatic symptoms and the occurrence of depression or anxiety (Creed et al. 2012).
Participants were asked to indicate whether they suffered from any of the sixteen
chronic medical disorders listed in the questionnaire and whether they were
receiving any medication or treatment for their disorder. Only disorders for which
treatment or medication was needed were included. The disorders included: lung
disease, heart conditions, diabetes, stroke, arthritis, cancer, hypertension, ulcers,
45 Somatization and onset of depression and anxiety
3
intestinal disorders, liver disease, epilepsy, allergies, thyroid disease, head injuries, other injuries and other chronic diseases. So-called functional somatic disorders (i.e. chronic fatigue syndrome, irritable bowel syndrome and fibromyalgia) were not included as these primarily consist of medically unexplained symptoms, which are the focus of our study.
Depressive and anxiety symptoms
As noted in the introduction, symptoms of depression and anxiety are conceptually related to somatization, and many depression and anxiety measures in their turn tend to cover somatic items. Moreover, high but subclinical levels of symptoms of depression and anxiety are among the most important risk factors for the onset of depressive and anxiety disorders (Karsten et al. 2011). Therefore we adjusted for symptoms of depression and anxiety in our analyses. Depressive symptoms were measured using the validated Quick Inventory of Depressive Symptomatology (QIDS; Rush et al. 2002). The QIDS consists of sixteen items covering mood, cognitive and only few somatic features of a depressive episode. The items are scored on a scale of 0 to 3, where zero indicates the absence of the symptom in question. Items are summed to obtain a total score ranging from 0-27. A higher score is indicative of more depressive symptoms. The QIDS correlates less with the 4DSQ somatization scale than the full IDS (r = 0.58 versus r = 0.67). Anxiety
symptoms were measured with the subjective scale of the validated Beck AnxietyInventory (BAI), which consists of seven items covering subjective and cognitive features of anxiety (Beck et al. 1988; Kabacoff et al. 1997). Items are scored (0 to 3) on a 4-point Likert scale and added to obtain a total score, ranging from 0-21 (Beck et al. 1988). Again, a higher score is indicative of more anxiety symptoms.
The subjective scale of the BAI was used because of its lower correlation (r = 0.48, p < .001) with the 4DSQ somatization scale compared to the complete BAI (r = 0.69, p < .001), which also includes items on somatic features of anxiety.
Accordingly, the somatic subscale of the BAI covers only these somatic features of anxiety.
Statistical Analyses
Descriptive statistics were used to describe the sample characteristics. Moreover, Pearson correlations between baseline symptoms of somatization, depression and anxiety were calculated.
The associations between somatization and the occurrence of a subsequent
depressive or anxiety disorder were quantified with Hazard Ratios (HRs). Cox
regression analyses were used because they account for differences in time-at-
46 Chapter 3
risk (i.e. person time), as inevitably there was some loss-to follow-up. In all of the analyses, the somatization score was entered into the regression model as the central determinant of the incidence of a depressive or anxiety disorder during the follow-up period. The possible confounding effects of sociodemographic (age, gender, education level, and work status), somatic (presence of chronic disease) and psychiatric characteristics (subclinical depressive or anxiety symptoms) were investigated by including these variables in the regression model in a stepwise manner. Analyses were performed with SPSS (version 20) statistical software.
RESULTS
The baseline sample characteristics are presented in Table 1. Twelve percent of the baseline sample of 611 persons scored at or above the cut-off score of 11 on the somatization scale indicating an elevated level of somatization and being at risk for impaired functioning. Only 5 participants (0.8%) scored in the highest range (21-32). Mean somatization scores at baseline were low [M (SD) = 5.0 (4.6)].
Somatization was moderately correlated with anxiety symptoms (r= 0.48, p <
.001) and depressive symptoms (r = 0.58, p < .001) at baseline.
Table 1. Baseline characteristics of selected study sample (n = 611)
Baseline characteristics
All participants n = 611
Depressive and/or anxiety disorder during follow-up n = 95
No depressive or anxiety disorder during follow-up n = 516
Age, M (SD)* 41.0 (14.6) 36.5 (13.8) 41.8 (14.6)
Gender: female, n (%) 372 (61%) 65 (68%) 307 (60%)
Education level: years, M (SD) 12.9 (3.2) 12.7 (3.1) 13.0 (3.2) Work status, n (%)*
Working
Sickness benefit/occupational disabled Not working
445 (73%) 25 (4%) 141 (23%)
58 (61%) 7 (7%) 30 (32%)
387 (75%) 18 (4%) 111 (22%)
Chronic condition: yes, n (%) 207 (34%) 26 (27%) 181 (35%)
Depressive symptoms: QIDS, M (SD)* 3.42 (3.09) 6.16 (4.12) 2.91 (2.56) Anxiety symptoms: BAI subjective scale, M (SD)* 1.35 (1.97) 2.53 (2.50) 1.14 (1.76) Somatization: 4DSQ, M (SD)* 5.04 (4.65) 8.66 (5.87) 4.37 (4.04)
* Significant difference (p < .05) between participants with and participants without incident depressive and/or anxiety disorder during 4-year follow-up.
47 Somatization and onset of depression and anxiety
3
Attrition analyses showed no significant differences between participants (n = 611) and non-participants (n = 41) for age [M (SD) = 41.0 (14.6) vs. 43.1 (16.2)], gender (female: 61% vs. 68%) and anxiety symptoms at baseline [M (SD) = 1.35 (1.97) vs. 2.18 (3.20)]. There was a slight, but statistically significant difference in number of years of education [M (SD) = 12.9 (3.21) vs. 11.2 (2.43), p < .001], and depressive symptoms at baseline [M (SD) = 3.42 (3.09) vs. 4.53 (3.42), p = 0.03].
During the 4-year follow-up, 95 participants (16%) became depressed and/or anxious: 47 participants (8 %) developed a depressive disorder, 24 participants (4%) developed an anxiety disorder and 24 participants (4%) developed both a depressive and an anxiety disorder.
Somatization as a predictor for an incident depressive or anxiety disorder
The Hazard Ratios (HRs) of the association between somatization and the occurrence of an incident depressive disorder (with or without anxiety), incident anxiety disorder (with or without depression), and depressive and/or anxiety disorder during follow-up are displayed in Table 2. This table shows that somatization significantly increased the risk of a subsequent incident depressive disorder as one point increase in the somatization score increased the incidence rate of a depressive disorder with 13% (HR = 1.13; 95%CI = 1.09 to 1.17; p < .001).
After adjusting for baseline levels of depressive symptoms as measured by the QIDS, the association of somatization with the onset of depressive disorders was reduced but remained statistically significant (HR = 1.06; 95%CI = 1.01 to 1.12; p
= .032). Higher levels of somatization at baseline also significantly increased the incidence rate of an anxiety disorder (HR = 1.14; 95%CI = 1.09 to 1.18; p < .001, adjusted HR = 1.13; 95%CI = 1.07 to 1.20; p < .001). Similar results were found for the onset of depressive and/or anxiety disorders: Having a higher somatization score at baseline increased the risk of becoming depressed or anxious (HR = 1.14;
95%CI = 1.10 to 1.17; p < .001). Again, the Hazard Ratio decreased after adjusting
for baseline levels of depressive and anxiety symptoms but remained statistically
significant (HR = 1.07; 95%CI = 1.02 to 1.13; p < .01). The positive association
between somatization and onset of depressive and/or anxiety disorders is also
shown in Figure 1. The incidence of depressive and/or anxiety disorders was
higher in high somatization groups.
48 Chapter 3
Table 2. Hazard Ratios for incidence of depressive disorder, anxiety disorder, and depressive and/or anxiety disorder at follow-up as predicted by baseline somatization score (n = 611)
Depressive disorder HR (95% CI)
Anxiety disorder HR (95% CI)
Depressive and/or anxiety disorder HR (95% CI) Somatization
crude association, per unit increase
1.13 (1.09 to 1.17)*
1.14 (1.09 to 1.18)*
1.14 (1.10 to 1.17)*
Somatization
adjusted for age, gender, education, working status 1.14 (1.10 to 1.19)*
1.17 (1.11 to 1.23)*
1.15 (1.11 to 1.19)*
Somatization + chronic disease
1.15 (1.10 to 1.20)*
1.17 (1.11 to 1.23)*
1.16 (1.11 to 1.20)*
Somatization
+ depressive and/or anxiety symptoms
1.06 (1.01 to 1.12)*
1.13
(1.07 to 1.20)*
1.07 (1.02 to 1.13)*
* p < .05; HR= Hazard Ratios; 95% CI = 95% Confi dence Interval
Somatisation score (quartiles)
8-32 (Q4) 5-7 (Q3)
2-4 (Q2) 0-1 (Q1)
Percentage
100,0%
80,0%
60,0%
40,0%
20,0%
0,0%
depressive or anxiety disorder
no depressive or anxiety disorder
Page 1
Figure 1. Incidence of depressive and/or anxiety disorders during 4-year follow-up (%), classifi ed according to somatization score (n = 611)
49 Somatization and onset of depression and anxiety
3
DISCUSSION
In this study we prospectively investigated somatization as a risk factor for the onset of incident depressive or anxiety disorders as determined by a clinical interview. In our study sample we found that somatization increased the risk of a subsequent incident depressive or anxiety disorder during four years, over and above baseline depressive and anxious symptoms.
Methodological considerations
The NESDA cohort study offers a unique opportunity to prospectively investigate the relationship between somatization and the onset of depressive and anxiety disorders as it includes detailed information on the history, phenomenology and correlates of these disorders and multiple assessments of depressive and anxiety disorders using clinical interviews over a 4-year time-period. However, some methodological considerations should be mentioned. At baseline we only had information on the presence and extent, but not on the history, duration and course of somatization. The symptoms reported by the participants were also not checked by a doctor to verify whether they could be explained by a somatic disorder. However, we did adjust for the presence of chronic somatic disorders that may explain symptoms in our analyses, and this had no effect on the strength of the hazard ratios. Unavoidably there was some attrition in our study; however, analyses showed that differences between participants and non-participants were small. Furthermore, inherent to the NESDA sampling methods (Penninx et al. 2008), our sample consisted of persons at high risk of depressive and/
or anxiety disorders as well as healthy controls. This study, therefore, reported higher incidence rates of depressive and anxiety disorders compared to the general population (de Graaf et al. 2013).
Our findings in context of the literature
Previous research has found a strong concurrent relationship between
somatization and depression. Moreover, somatization also seems to have negative
prognostic value as co-occurrence is predictive of a poorer treatment response
in depressed patients (Huijbregts et al. 2010). Several underlying mechanisms
have been proposed that may explain how these conditions are related. It has
been suggested that 1) depression and anxiety may be a reaction to somatization,
2) somatization may be part of, or a consequence of depression and anxiety, and
3) that all of these conditions are just different expressions and dimensions of a
common underlying form of distress (Henningsen et al. 2003). Findings pertaining
50 Chapter 3
to the first proposition, which can be considered to be the focus of the current study, have thus far been mixed. A study in a community sample found that after correction for depressive symptoms at baseline, somatization marginally predicted subsequent depressive symptoms at 5-year follow-up in non-depressed women, but were not predictive in non-depressed men (Terre et al. 2003). Furthermore, a prospective community cohort study in the UK reported that only a persistently high somatic symptom count was predictive of poor self-reported mental health while a high somatic symptom count at baseline was not predictive (Creed et al. 2012). Although each study has its methodological limitations, an important restriction across the board is that unlike in our study, psychiatric history of depression or anxiety was either not or not adequately taken into consideration.
Thus it was impossible to discriminate between recurrence and incidence, but it can also confound the association between somatization and subsequent depression or anxiety as psychiatric history is an important predictor of these disorders (Karsten et al. 2011). In our study, which was restricted to participants without a lifetime history of a depressive or anxiety disorder, we found that somatization prospectively predicted these disorders above and beyond baseline levels of depressive or anxiety symptoms, and, therefore, provided support for the first mechanism mentioned above.
Somatization and symptoms of depression and anxiety
Commonly used measures for depressive and anxiety symptoms such as the IDS and BAI also assess somatic symptoms that accompany or are part of depression and anxiety. To minimise difficulties in investigating the relationship between somatization, depression and anxiety, we used the QIDS, which includes fewer items on somatic features of depression and the subjective scale of the BAI which only focuses on subjective and cognitive aspects of anxiety. Furthermore, the majority of somatization measures currently available operationalise somatization as an increased somatic symptom count (e.g. PHQ-15, SCL-90 somatization subscale, 4DSQ somatization subscale). However, although somatic symptom count is certainly a core feature of somatization it is probably not very specific.
Increased somatic symptom count as measured with somatization scales may, in this case, be found to predict depression and anxiety, but questions remain, such as whether it is actually the ‘somatization’ (i.e. “the tendency to experience and communicate somatic stress and symptoms unaccounted for by pathological findings, to attribute them to physical illness, and to seek medical help for them”
(Lipowski 1988) that is predictive and to which condition do we attribute the
somatic symptoms. In line with DSM V (Dimsdale & Creed 2009; APA 2013), a
51 Somatization and onset of depression and anxiety
3
step forward may be to operationalise and develop more specific measures of somatization, which include cognitive criteria besides listing somatic symptoms (Rief & Broadbent 2007;Rief et al. 2010). The concept and operationalisation of somatization as well as its conceptual relationship with depressive and anxiety symptoms remains an important point of consideration and future study.
CONCLUSIONS
High physical symptom levels (somatization) predict incident depressive and
anxiety disorders during four years, over and above baseline levels of depressive
or anxiety symptoms, in a study sample without a lifetime history of depressive
and anxiety disorders. Somatization scores may, therefore, deserve extra clinical
attention in primary care. The 4DSQ offers such a symptom tool and is increasingly
used as such in countries around the world. Also, persons with high somatization
scores represent a potential target group for prevention of depressive and anxiety
disorders as they have a higher risk of becoming depressed or anxious.
Longitudinal associations of multiple physical symptoms with recurrence of depressive and anxiety disorders
Dijkstra-Kersten SMA Sitnikova K
Terluin B Penninx BWJH Twisk JWR van Marwijk HWJ van der Horst HE van der Wouden JC
Published in:
J Psychosom Res 2017;97:96-101.
Chapter 4
54 Chapter 4
