The prevalence and clinical characteristics associated with Diagnostic and Statistical Manual Version-5-defined anxious distress specifier in adults with major depressive disorder: results from the International Mood Disorders Collaborative Project

Abstract

Objectives:

The aim of the study was to evaluate the prevalence of and illness characteristics in adults with major depressive disorder (MDD) with anxious distress specifier (ADS) enrolled in the International Mood Disorders Collaborative Project, which is a collaborative research platform at the Mood Disorders Psychopharmacology Unit, University of Toronto, Canada and the Cleveland Clinic, Cleveland, Ohio, USA.

Methods:

Data from participants who met criteria for a current major depressive episode as part of MDD (n = 830) were included in this post hoc analysis. Diagnostic and Statistical Manual Version-5-defined ADS was operationalized as the presence of at least two out of three proxy items instead of two out of five specifiers.

Results:

A total of 464 individuals (i.e. 56%) met criteria for ADS. There were no between-group differences in sociodemographic variables (e.g. gender, employment, marital status). Greater severity of illness was observed in adults with ADS as evidenced by a higher number of hospitalizations, higher rates of suicidal ideation, greater depressive symptom severity, greater workplace impairment, decreased quality of life, and greater self-reported cognitive impairment.

Conclusions:

Our findings underscore the importance of evaluating ADS in adults with MDD as its presence identifies a subpopulation with greater illness-associated burden and hazards.

Introduction

Epidemiological and clinical studies provide replicated evidence that anxiety symptoms and disorders are commonly observed in adults with major depressive disorder (MDD) [Schaffer et al. 2012]. The pertinence of anxiety symptoms/disorders is underscored by convergent evidence indicating that MDD subpopulations with high baseline anxiety measures have a lower probability of, and longer time to achieving, remission [Fava et al. 2008]. It is also observed that adults with MDD and prominent anxiety as a dimension and/or comorbidity are more susceptible to treatment- emergent adverse events and consequently, lower rates of treatment acceptance [Fava et al. 2008].

The foregoing observations contribute to more unfavorable outcomes (i.e. illness trajectory and adherence) in adults with MDD and prominent anxiety. For example, anxiety features are highly associated with suicidal ideation, nonlethal self-harm and completed suicide [Fawcett, 2001]. The hazards posed by anxiety in adults with MDD provided the impetus for the Diagnostic and Statistical Manual Version-5 (DSM-5) to introduce and define the anxious distress specifier (ADS). Anxious distress is defined as the presence of at least two of the following symptoms during the majority of days of a major depressive episode or persistent depressive disorder: (a) feeling keyed up or tense; (b) feeling unusually restless; (c) difficulty concentrating because of worry; (d) fear that something awful might happen; (e) feeling that the individual might lose control of himself or herself. Clinicians are also encouraged to specify current ADS severity with Likert scores ranging from mild to severe depending on the number of criterion items met.

Herein, we sought to determine the prevalence of ADS amongst a well-characterized group of adults (i.e. ages 18–87) utilizing university-based mood disorder services with a diagnosis of MDD at one of two participating centers as part of the International Mood Disorders Collaborative Program (IMDCP). The overarching aim of this analysis is to provide empirical data regarding the clinical characteristics of individuals meeting the DSM-5-defined ADS criteria.

Methods

A total of 1861 individuals consented to be a part of the IMDCP between January 2008 and July 2013. The IMDCP is a multi-site, naturalistic, cross-sectional study of individuals presenting for treatment and/or evaluation in one of two tertiary care specialized centers (i.e. the Mood Disorders Psychopharmacology Unit (MDPU) located at the University Health Network, University of Toronto, Canada and the Cleveland Clinic for Mood Disorders Treatment and Research at Lutheran Hospital, Cleveland, OH, USA). Both the MDPU and the Cleveland Clinic are academic specialty research programs providing clinical service to adults (i.e. ages 18–87) with MDD or bipolar disorder. The MDPU is exclusively an outpatient program while the Cleveland Clinic offers both outpatient and inpatient services. Exclusion criteria for entry into the IMDCP are unwillingness or inability to comply with study assessment or to provide informed consent. The Research Ethics Board of the University Health Network, University of Toronto and the Institutional Review Board of the Cleveland Clinic Foundation at Lutheran Hospital approved the MDPU research platforms at both centers, respectively.

A total of 830 individuals who had a diagnosis of MDD were included in the analysis. We proxied the definition of ADS by identifying items in the IMDCP metrics that comport with criteria items of the ADS. For example, feeling keyed up or tense was proxied by the Montgomery–Asberg Depression Rating Scale (MADRS) item 3 (inner tension score ⩾ 3), difficulty concentrating because of worry was proxied by MADRS item 6 (concentration difficulties score ⩾ 3), and fear that something awful may happen was proxied by item 10 (anxiety psychic score > 3) of the Hamilton Depression Rating 17-item Scale (HAM-D-17). See Table 1 for a summary.

Table 1.

Proxy to determine anxious distress specifier.

DSM-IV-TR diagnosis	DSM-5 operational definition for anxious distress specifier	Proxy items
Major depressive disorder	1. Feeling keyed up or tense	MADRS item 3 – inner tension [score ⩾ 3]
	2. Feeling unusually restless	–
	3. Difficulty concentrating because of worry	MADRS item 6 – concentration difficulties [score ⩾ 3]
	4. Fear that something awful may happen	HAM-D-17 item 10 – anxiety psychic [score > 3]
	5. Feeling that the individual might lose control of himself/herself	–

Anxious distress specifier met: presence of at least two symptom criteria. DSM, Diagnostic and Statistical Manual of Mental Disorders; DSM-5, Diagnostic and Statistical Manual Version-5; HAM-D-17, Hamilton Depression Rating 17-item Scale; MADRS, Montgomery–Asberg Depression Rating Scale.

Individuals with MDD, with or without ADS, were compared on sociodemographic variables as well as measures of depression severity (i.e. HAM-D-17, MADRS and Clinical Global Impression – Severity [CGI-S]), cognitive dysfunction (i.e. Adult Attention Deficit Hyperactivity Disorder Self-Report Scales [ASRS] inattention subscale), neuroticism (i.e. NEO Five-Factor Inventory [NEO-FFI]), workplace productivity (i.e. Endicott Workplace Productivity Scale [EWPS]), and quality of life (i.e. Quality of Life Enjoyment and Satisfaction Questionnaire [Q-LES-Q]).

All data were initially captured with paper versions of all scales and then either manually entered or scanned with automated capture software (TELEFORM Version 8) prior to statistical analysis. Statistical analysis was conducted using SPSS for Windows, Version 20. The chi-square statistic was utilized for the comparison of prevalence rates as well as for comparison of other categorical factors (i.e. demographic, presence of psychotic features, comorbidity) between the groups. The Student’s t-test or the nonparametric equivalent Mann–Whitney U tests were applied for nonparametric comparisons.

Results

Data from a total of 830 participants were included in the analysis; 56% (n = 464) of individuals with MDD met criteria for ADS. Demographic information is summarized and presented in Table 2. There were no significant between-group differences in sex, employment status, or marital status. Significant between-group differences in race and education were documented.

Table 2.

Demographic characteristics.

	Major depressive disorder (n = 830)				Student’s t test	DF	p value
	Criteria for anxious distress specifier NOT met		Criterion for anxious distress specifier met
	(n = 366)		(n = 464)
	M	SD	M	SD
Age	42.5	14.8	44.4	15.0	−1.79	828	0.073
Sex	n	%	n	%	Chi-square test	DF	p value
Female	240	65.6	294	63.4	0.44	1	0.512
Male	126	34.4	170	36.6
Race	n = 284		n = 346
White	247	87.0	311	89.9	12.30	3	0.006
Black/African-American	9	3.2	22	6.4
Asian	15	5.3	7	2.0
Other	13	4.6	6	1.7
Employment status	n = 301		n = 356
Employed	143	47.5	147	41.3	3.99	3	0.263
Student	32	10.6	33	9.3
Unemployed or disabled	94	31.2	127	35.7
Other (homemaker, retired)	32	10.6	49	13.8
Marital status	n = 303		n = 355
Single	122	40.3	124	34.9	2.01	2	0.366
Married or cohabiting	129	42.6	163	45.9
Divorced or separated or widowed	52	17.2	68	19.2
Education	n = 336		n = 408
High-school diploma or less	54	16.1	109	26.7	13.01	3	0.005
Some college or university	76	22.6	90	22.1
Postsecondary diploma or degree	140	41.7	145	35.5
Some postgraduate or higher	66	19.6	64	15.7
Sample size varies due to missing data

DF, degrees of freedom; M, mean, SD, standard deviation.

The presence of anxious distress was associated with greater depression and illness severity as measured by HAM-D-17, MADRS, and CGI-S (t = −21.25, degrees of freedom [DF] = 821, p < 0.001, t = −20.48, DF = 826, p < 0.001, and t = −10.32, DF = 695, p < 0.001, respectively). Individuals with ADS reported significantly higher levels of somatic, cognitive, and behavioral anxiety compared to those without ADS (t = −7.13, DF = 444, p < 0.001, t = −7.77, DF = 434, p < 0.001, t = −5.53, DF = 440, p < 0.001, respectively). Cognitive dysfunction (i.e. as measured with the ASRS-inattention subscale) was significantly greater in those with ADS (t = −3.96, DF = 427, p < 0.001). Adults with MDD and ADS were also more likely to have higher scores on measures of neuroticism (i.e. NEO-FFI) (t = −5.42, DF = 464, p < 0.001) (Table 3).

Table 3.

Illness characteristics.

	Major depressive disorder (n = 830)
	Criteria for anxious distress specifier NOT met		Criteria for anxious distress specifier met
	(n = 366)		(n = 464)
	M	SD	M	SD
Lifetime markers of illness severity
	M	SD	M	SD	Mann–Whitney U test	Z-test	p value
Age at onset	25.24	14.98	25.8	15.1	77945.00	−0.631	0.528
Illness duration	17.0	13.8	18.3	14.9	76933.00	−0.94	0.347
Time to treatment	5.97	9.64	7.2	10.3	25117.0	−1.376	0.169
	n	%	n	%	Chi-square test	DF	p value
History of hospitalization for mood episode	129.00	46.10	266.00	76.40	61.31	1	< 0.001
History of suicide attempt	106	29	151	32.6	2.05	2	0.358
Current illness severity
	M	SD	M	SD	Student’s t test	DF	p value
HAM-D-17	14.5	6.5	23.4	5.6	–21.25	821	< 0.001
MADRS	21.1	10.1	33.9	7.8	–20.48	826	< 0.001
CGI-S	4.4	1.3	5.2	0.9	–10.32	695	< 0.001
TAQ – Total	112.0	48.7	151.7	52.1	–8.07	418	< 0.001
TAQ – Somatic	35.9	23.2	53.3	28.3	–7.13	444	< 0.001
TAQ – Cognitive	43.3	18.1	56.7	18.0	–7.77	434	< 0.001
TAQ – Behavioral	32.7	16.1	41.0	15.7	–5.53	440	< 0.001
NEO – Neuroticism	29.9	7.6	33.6	7.0	–5.42	464	< 0.001
NEO – Extraversion	20.9	7.9	20.0	7.3	1.26	464	0.208
NEO – Openness	28.2	7.1	27.5	6.8	1.19	464	0.234
NEO – Agreeableness	32.0	5.8	31.0	6.4	1.61	464	0.107
NEO – Conscientiousness	27.5	8.0	26.5	8.2	1.31	464	0.191
ASRS	30.0	11.0	35.3	12.1	–4.73	427	< 0.001
Inattention subscale	17.1	6.9	19.9	7.5	–3.96	427	< 0.001
Hyperactivity subscale	12.9	5.8	15.4	6.1	–4.38	427	< 0.001
Function and quality of life
	M	SD	M	SD	Student’s t test	DF	p value
Sheehan Disability Scale	17.2	8.5	22.4	6.4	–7.26	442	< 0.001
Endicott Work Productivity Scale	31.8	18.1	39.9	17.0	–3.25	201	0.001
Rosenberg Self-Esteem Scale	25.8	5.9	28.3	5.4	–4.68	465	< 0.001
Quality of Life Enjoyment and Satisfaction Questionnaire	45.9	16.6	35.3	15.4	6.79	421	< 0.001
Psychiatric comorbidities
	n	%	n	%	Chi-square test	DF	p value
Suicidal ideation (past month)	97	36.9	137	52.3	12.61	1	< 0.001
Panic disorder	51	14.0	102	22.1	8.67	1	0.003
Agoraphobia	61	17.0	119	25.9	9.27	1	0.002
Social phobia	43	11.8	79	17.1	4.57	1	0.032
Specific phobia	15	4.1	17	3.7	0.10	1	0.752
Obsessive compulsive disorder	22	6.0	37	8.0	1.20	1	0.273
Posttraumatic stress disorder	19	5.2	59	12.7	13.59	1	< 0.001
Alcohol dependence current	41	11.2	35	7.5	3.29	1	0.070
Alcohol dependence lifetime	82	22.4	76	16.4	4.82	1	0.028
Alcohol abuse current	19	5.2	12	2.6	7.63	2	0.022
Alcohol abuse lifetime	35	9.6	43	9.3	4.99	2	0.083
Nonalcohol psychoactive SUD – dependence – lifetime	48	13.2	74	16.1	1.39	1	0.238
Nonalcohol psychoactive SUD – dependence – current	19	5.4	35	7.6	1.69	1	0.193
Nonalcohol psychoactive SUD – abuse – current	9	2.5	18	3.9	1.29	2	0.525
Psychotic disorder not otherwise specified	0	0.0	1	0.2	0.79	1	0.374
Anorexia nervosa – current	0	0.0	1	0.2	0.79	1	0.374
Bulimia nervosa – current	11	3.0	9	1.9	0.98	1	0.323
Generalized anxiety disorder – current	110	30.1	243	52.4	41.31	1	< 0.001
Attention deficit hyperactivity disorder	15	4.2	20	4.3	0.02	1	0.892

Sample size varies due to missing data. ASRS, Attention Deficit Hyperactivity Disorder Self-Report Scale; CGI–S, Clinical Global Impression – Severity; DF, degrees of freedom; HAM-D-17, Hamilton Depression Rating 17-item Scale; MADRS, Montgomery–Asberg Depression Rating Scale; M, mean; SD, standard deviation; SUD, substance use disorder; TAQ, Trimodal Anxiety Questionnaire.

Individuals with anxious distress reported worse workplace productivity as well as greater disability (i.e. as measured by the EWPS) (t = −3.25, DF = 201, p = 0.001, t = −7.26, DF = 442, p < 0.001). Furthermore, individuals with MDD and ADS reported significantly lower levels of quality of life as measured by the Q-LES-Q (t = 6.79, DF = 421, p < 0.001) (Table 3).

Medical comorbidity data is summarized and presented in Table 4.

Table 4.

Medical comorbidity.

	Major depressive disorder (n = 830)				Student’s t test	DF	p value
	Criteria for anxious distress specifier NOT met		Criterion for anxious distress specifier Met
	(n = 366)		(n = 464)
	M	SD	M	SD
Body mass index (n = 264, n = 423)	27.8	7.03665	28.2	7.15398	−0.69	685	0.490
	n	%	n	%	Chi-square test	DF	p value
Endocrine and metabolic disorders	129	70.9	143	70.1	0.028	1	0.867
Cardiovascular disorders	86	43.2	117	51.5	2.947	1	0.086
Rheumatologic/autoimmune disorders	37	28.2	40	29.6	0.062	1	0.803
Infectious diseases	22	19.0	16	12.4	2.007	1	0.157
Renal diseases	13	10.7	11	8.6	0.306	1	0.580
Liver diseases	0	0.0	6	4.8	5.520	1	0.019
Pulmonary disorders	60	38.2	48	34.5	0.432	1	0.511
Dermatologic disorders	58	47.9	47	37.3	2.855	1	0.091
Hematologic diseases	28	23.5	28	22.4	0.044	1	0.834
Metabolic bone disorders	8	7.0	4	3.2	1.784	1	0.182
Genitourinary disorders	22	17.6	18	13.6	0.768	1	0.381
Gastrointestinal disorders	62	39.5	68	46.9	1.686	1	0.194
Neurological disorders	99	59.6	113	75.3	8.791	1	0.003
Neurological trauma or infections	18	15.0	19	14.6	0.007	1	0.932
Eye disorders	82	69.5	76	59.8	2.487	1	0.115
Ear, nose, throat and dental disorders	60	46.9	65	48.5	0.070	1	0.791
Neoplasm	6	5.3	15	11.5	3.041	1	0.081
Other medical disorders	186	86.9	155	81.6	2.178	1	0.140
Neurological disorders
Migraine	38	24.5	55	39.9	7.928	1	0.005
Headaches	29	20.1	29	22.1	0.165	1	0.685
Chronic pain	30	20.7	34	25.4	0.864	1	0.353
Sleep and wakefulness disorders	22	15.2	21	15.9	0.029	1	0.866
Sleep apnea	23	15.8	20	14.9	0.037	1	0.848
Epilepsy	2	1.4	2	1.5	0.012	1	0.914
Multiple sclerosis	0	0	2	1.5	2.153	1	0.142
Huntington’s disease	0	0	1	0.8	1.081	1	0.298
Other neurological disorders	9	6.3	7	5.2	0.146	1	0.702

Sample size varies due to missing data. DF, degrees of freedom; M, mean; SD, standard deviation.

Discussion

The overarching finding of the analysis herein is that anxious distress (i.e. ADS-proxied definition) is common in patients with MDD and, when present, identifies a subpopulation of adults with MDD who evince greater illness severity on both patient- and clinician-reported measures. The observation that anxious distress identifies a subpopulation of MDD with a more unfavorable presentation represents a replication of existing data [Fava et al. 2004, 2008]. A novel aspect of our analysis is that relatively few studies have sought to determine the prevalence and illness characteristics associated with DSM-5-defined ADS.

An observation herein, which would need to be considered preliminary, is that individuals with MDD fulfilling ADS self-report significantly greater cognitive dysfunction. Cognitive dysfunction is prevalent, pervasive, and, in many cases, an enduring feature in adults with MDD. Moreover, self-reported cognitive impairment accounts for greater variability in workplace performance than does total depression symptom severity [McIntyre et al. 2015]. There is a need to identify determinants of cognitive dysfunction in MDD. Our results preliminarily suggest that the presence of ADS negatively affects measures of cognitive function [McIntyre et al. 2013]. A testable hypothesis may be that the unfavorable illness presentation, course, and treatment outcomes in a subpopulation of adults with MDD and prominent anxiety may be mediated by greater cognitive impairment.

There are several methodological limitations that affect inferences and interpretations of our results. Firstly, the overarching limitation is that all subjects received a post hoc diagnosis of ‘anxious distress’ rather than a diagnosis using DSM-5 criteria at the time of entry. In addition, we defined ADS as present when at least two symptoms amongst three were present rather than two out of five as described in DSM-5. A separate limitation is the heterogeneity of the sample composition with respect to demographic, illness severity, clinical characteristics, comorbidity, and treatment regimen. Nevertheless, we are of the view that these limitations may also be interpreted as strengths insofar as MDD is a highly heterogeneous presentation across diverse clinical settings. Thirdly, data for the analysis were collected cross-sectionally; in addition, we had incomplete data capture with missing data of varying degrees across our constituent scales and instruments. It should also be underscored that we did not have a comprehensive measure of cognitive function and, instead, we proxied cognitive function with the inattention subscale of the ASRS. Notwithstanding the foregoing limitations, the subjects enrolled in our database represent ‘real-world’ patients commonly encountered in a tertiary university-based setting, enhancing the ecological validity.

Our findings in this brief preliminary report align with the clinical impression that ADS is a common clinical presentation amongst adults with MDD and, when present, identifies a subpopulation of adults with MDD with greater illness-associated morbidity and burden. An important tenet of chronic disease management models is the use of measurement-based care and ‘treating to target’ [Chwastiak et al. 2014]. Clinicians are encouraged to screen for the presence of ADS in adults with MDD and to anticipate a more complicated illness presentation and, consequently, less favorable outcomes with single modality therapies, highlighting the need for integrated and combined pharmacological and psychosocial approaches.