Performance characteristics and clinical utility of diagnostic criteria proposals in bereaved treatment-seeking patients.

BACKGROUND: Persistent complex bereavement disorder (PCBD) is a protracted form of grief included in DSM Section 3 indicating a need for more research. Two other criteria sets [prolonged grief disorder (PGD) and complicated grief (CG) disorder] are also currently in use by researchers. This study evaluates rates of diagnosis of each proposed criteria set in a clinical sample of bereaved individuals participating in clinical research. METHOD: Two groups in which persistent grief was judged to be present or absent completed an assessment instrument that included items needed to diagnose PCBD as well as PGD and CG. One group included grief treatment-seeking participants in our multicenter National Institute of Mental Health (NIMH)-sponsored study who scored ⩾30 on the Inventory of Complicated Grief (ICG) and the other comprised bereaved adults enrolled in clinical research studies who scored <20 on the ICG. Rates of diagnosis were determined for proposed PCBD, PGD and CG criteria. RESULTS: PCBD criteria diagnosed 70 [95% confidence interval (CI) 64.2-75.8] % of the grief treatment-seeking group, PGD criteria identified 59.6 (95% CI 53.4-65.8) % of these individuals and CG criteria identified 99.6 (95% CI 98.8-100.0) %. None of the three proposed criteria identified any cases in the bereaved comparison group. CONCLUSIONS: Both proposed DSM-5 criteria for PCBD and criteria for PGD appear to be too restrictive as they failed to identify substantial numbers of treatment-seeking individuals with clinically significant levels of grief-related distress and impairment. Use of CG criteria or a similar algorithm appears to be warranted.

Adult separation anxiety differentiates patients with complicated grief and/or major depression and is related to lifetime mood spectrum symptoms.

BACKGROUND: Increasing literature has been focused on complicated grief (CG) and its distinctiveness from other potentially loss related mental disorders such as major depression (MD). In this regard, symptoms of separation distress seem to play a key role. The aim of this study was to compare the clinical features of CG to those of MD and of CG+MD, with particular attention to separation anxiety. METHODS: Fifty patients with CG (26 with and 24 without MD) and 40 with MD were consecutively recruited. Assessments included: SCID-I/P, Inventory of Complicated Grief (ICG), Adult Separation Anxiety Symptom Questionnaire (ASA-27), Work and Social Adjustment Scale (WSAS), Mood Spectrum-Self Report (MOODS-SR)-lifetime version. RESULTS: Patients with MD reported significantly higher ASA-27 scores than patients with CG either alone or with MD. In all groups, ASA-27 total scores were significantly correlated with the MOODS-SR total scores and with those of its depressive component and rhythmicity domain. No significant differences were reported in the WSAS scores. LIMITATIONS: Major limitations are the small sample size and the use of lifetime instruments. CONCLUSIONS: Our results suggest a correlation between adult separation anxiety symptoms and lifetime mood spectrum symptoms both in patients with CG and MD. Further studies are needed to better understand the role of adult separation anxiety in the development of these disorders and for their nosographic autonomy as well.

Composite International Diagnostic Interview screening scales for DSM-IV anxiety and mood disorders.

BACKGROUND: Lack of coordination between screening studies for common mental disorders in primary care and community epidemiological samples impedes progress in clinical epidemiology. Short screening scales based on the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI), the diagnostic interview used in community epidemiological surveys throughout the world, were developed to address this problem. METHOD: Expert reviews and cognitive interviews generated CIDI screening scale (CIDI-SC) item pools for 30-day DSM-IV-TR major depressive episode (MDE), generalized anxiety disorder (GAD), panic disorder (PD) and bipolar disorder (BPD). These items were administered to 3058 unselected patients in 29 US primary care offices. Blinded SCID clinical reinterviews were administered to 206 of these patients, oversampling screened positives. RESULTS: Stepwise regression selected optimal screening items to predict clinical diagnoses. Excellent concordance [area under the receiver operating characteristic curve (AUC)] was found between continuous CIDI-SC and DSM-IV/SCID diagnoses of 30-day MDE (0.93), GAD (0.88), PD (0.90) and BPD (0.97), with only 9-38 questions needed to administer all scales. CIDI-SC versus SCID prevalence differences are insignificant at the optimal CIDI-SC diagnostic thresholds (χ2 1 = 0.0-2.9, p = 0.09-0.94). Individual-level diagnostic concordance at these thresholds is substantial (AUC 0.81-0.86, sensitivity 68.0-80.2%, specificity 90.1-98.8%). Likelihood ratio positive (LR+) exceeds 10 and LR- is 0.1 or less at informative thresholds for all diagnoses. CONCLUSIONS: CIDI-SC operating characteristics are equivalent (MDE, GAD) or superior (PD, BPD) to those of the best alternative screening scales. CIDI-SC results can be compared directly to general population CIDI survey results or used to target and streamline second-stage CIDIs.

Predictors and moderators of time to remission of major depression with interpersonal psychotherapy and SSRI pharmacotherapy.

BACKGROUND: Although many studies suggest that, on average, depression-specific psychotherapy and antidepressant pharmacotherapy are efficacious, we know relatively little about which patients are more likely to respond to one versus the other. We sought to determine whether measures of spectrum psychopathology are useful in deciding which patients with unipolar depression should receive pharmacotherapy versus depression-specific psychotherapy. METHOD: A total of 318 adult out-patients with major depression were randomly assigned to escitalopram pharmacotherapy or interpersonal psychotherapy (IPT) at academic medical centers at Pittsburgh, Pennsylvania and Pisa, Italy. Our main focus was on predictors and moderators of time to remission on monotherapy at 12 weeks. RESULTS: Participants with higher scores on the need for medical reassurance factor of the Panic-Agoraphobic Spectrum Self-Report (PAS-SR) had more rapid remission with IPT and those with lower scores on the psychomotor activation factor of the Mood Spectrum Self-Report (MOODS-SR) experienced more rapid remission with selective serotonin reuptake inhibitor (SSRI) pharmacotherapy. Non-specific predictors of longer time to remission with monotherapy included several panic spectrum and mood spectrum factors and the Social Phobia Spectrum (SHY) total score. Higher baseline scores on the 17- and 25-item Hamilton Depression Rating Scales (HAMD-17 and HAMD-25) and the Work and Social Adjustment Scale (WSAS) also predicted a longer time to remission, whereas being married predicted a shorter time to remission. CONCLUSIONS: This exploratory study identified several non-specific predictors but few moderators of psychotherapy versus pharmacotherapy outcome. It offers useful indicators of the characteristics of patients that are generally difficult to treat, but only limited guidance as to who benefits from IPT versus SSRI pharmacotherapy.

The structure of lifetime panic-agoraphobic spectrum.

The heterogeneity of the clinical presentation of panic disorder (PD) has prompted researchers to describe different subtypes of PD, on the basis of the observed predominant symptoms constellation. Starting from a dimensional approach to panic disorder, an instrument to assess lifetime panic-agoraphobic spectrum (PAS) available in interview or self-report form (SCI-PAS, PAS-SR) was developed which proved to have sound psychometric properties and the ability to predict delayed response to treatment in patients with mood disorders. However, the structure of the instrument was defined a priori and an examination of its empirical structure is still lacking. Aim of the present report is to analyse the factor structure of the PAS taking advantage of a large database of subjects with panic disorders (N=630) assessed in the framework of different studies. Using a classical exploratory factor analysis based on a tetrachoric correlation matrix and oblique rotation, 10 factors were extracted, accounting overall for 66.3% of the variance of the questionnaire: panic symptoms, agoraphobia, claustrophobia, separation anxiety, fear of losing control, drug sensitivity and phobia, medical reassurance, rescue object, loss sensitivity, reassurance from family members. The first two factors comprise the DSM-IV criteria for panic disorder and agoraphobia. The other factors had received limited empirical support to date. We submit that these symptoms profiles might be clinically relevant for tailoring drug treatments or psychotherapeutic approaches to specific needs. Future perspectives might include the use of these factors to select homogeneous subgroups of patients for brain-imaging studies and to contribute to elucidating the causes and pathophysiology of panic disorder at molecular level.

Relation between clinical features of the mitral prolapse syndrome and echocardiographically documented mitral valve prolapse.

Mitral valve prolapse, the most common inherited cardiovascular condition, has been associated with a variety of signs, symptoms and electrocardiographic abnormalities, but the true spectrum of the mitral prolapse syndrome remains in doubt because clinical findings often contribute to patient identification and their prevalence in patient groups may be overstated because of ascertainment bias. Accordingly, clinical findings in 88 patients with echocardiographic mitral prolapse were compared with those in 81 of their adult first degree relatives with mitral prolapse (a group free of ascertainment bias) and in two control groups without mitral prolapse: 172 first degree relatives and 60 spouses. Comparison of relatives with and without mitral prolapse demonstrated true associations between mitral prolapse and clicks or murmurs, or both (67 versus 9%, p less than 0.001), thoracic bony abnormalities (41 versus 16%, p less than 0.001), systolic blood pressure less than 120 mm Hg (53 versus 31%, p less than 0.001), body weight 90% or less of ideal (31 versus 14%, p less than 0.005) and palpitation (40 versus 24%, p less than 0.01). In contrast, relatives with mitral prolapse showed no significant increase over normal relatives or spouses without mitral prolapse in prevalence of chest pain, dyspnea, panic attacks, high anxiety or repolarization abnormalities, but these features were all more common in women than in men (p less than 0.01 to less than 0.001). Thus, the true spectrum of the mitral prolapse syndrome encompasses a midsystolic click and late systolic murmur, thoracic bony abnormalities, low body weight and blood pressure and palpitation. Other suggested clinical features, including nonanginal chest pain, dyspnea, panic attacks and electrocardiographic abnormalities, have appeared to be associated with mitral valve prolapse because of ascertainment bias and an erroneous classification of differences between men and women as being due to mitral valve prolapse.

Standardized assessment for panic disorder research. A conference report.

Lively controversies related to panic disorder are under active investigation by research groups around the world. However, publications from different laboratories are difficult to compare since there has been little consistency in measures or even in types of assessment used to characterize and follow up patients. Participants in the recently convened National Institutes of Health Consensus Development Conference on the Treatment of Panic Disorder noted this problem and recommended establishment of procedures to ensure comparability of studies. We organized a conference of clinical investigators whose objective was to develop a standard assessment package. Participants represented biological and psychosocial panic disorder treatment research sites in the United States and Canada. The 2-day conference resulted in agreement on a battery of assessments considered essential for panic disorder studies. The purposes of our report are to disseminate the conference conclusions and to encourage adoption of the proposed standards by clinical researchers, journal editors, Public Health Service peer review committees, and the Food and Drug Administration. We also identify some problematic issues that require further work.

Cognitive behavioral treatment compared with nonprescriptive treatment of panic disorder.

BACKGROUND: The efficacy of cognitive behavioral treatment for panic disorder has been established in controlled studies. However, little is known about the efficacy of other psychological treatments. We report the results of a study comparing cognitive behavioral treatment with a focused nonprescriptive treatment for panic. METHODS: Three sessions of panic-related information were provided in each treatment, followed by 12 sessions of either nonprescriptive, reflective listening (non-prescriptive treatment) or a treatment package that included breathing retraining, muscle relaxation, cognitive reframing, and exposure to interoceptive and agoraphobic stimuli (cognitive behavioral treatment). RESULTS: Posttreatment and 6-month follow-up assessments revealed a good response to both treatments. We observed a high rate of panic remission and significant improvement in associated symptoms in subjects in each treatment group. CONCLUSION: These findings raise questions about the specificity of cognitive behavioral treatment.

The effect of psychobiological research on treatment outcome. A controlled study.

Recent advances in psychiatric research methodology promise major progress. Simultaneously, however, mounting concerns about ethnics of human experimentation have resulted in increased scrutiny and regulation that threaten scientific productivity. Virtually no systematic data have been gathered about the effects of research participation on treatment outcome or patient satisfaction. In this study 56 hospitalized depressed patients, who had agreed to participate in psychobiological research protocols, were then randomly assigned to treatment on a research unit or on standard adult inpatient (nonresearch) units. Research participants received more diagnosis-related somatic treatments, had a longer mean length of stay, and experienced trends toward greater symptom reduction and better consumer satisfaction. We conclude that research participation may be helpful to patients but that more systematic study is needed to help to resolve ethical questions and to assist risk-benefit evaluations.

Adolescent onset of the gender difference in lifetime rates of major depression: a theoretical model.

Prepubescent boys are, if anything, more likely than girls to be depressed. During adolescence, however, a dramatic shift occurs: between the ages of 11 and 13 years, this trend in depression rates is reversed. By 15 years of age, females are approximately twice as likely as males to have experienced an episode of depression, and this gender gap persists for the next 35 to 40 years. We offer a theoretical framework that addresses the timing of this phenomenon. First, we discuss the social and hormonal mechanisms that stimulate affiliative needs for females at puberty. Next, we describe how heightened affiliative need can interact with adolescent transition difficulties to create a depressogenic diathesis as at-risk females reach puberty. This gender-linked vulnerability explains why adolescent females are more likely than males to become depressed when faced with negative life events and, particularly, life events with interpersonal consequences.

A multicenter investigation of fixed-dose fluoxetine in the treatment of obsessive-compulsive disorder.

OBJECTIVES: To determine the effectiveness of fluoxetine hydrochloride at fixed doses of 20 mg/d, 40 mg/d, and 60 mg/d in patients with obsessive-compulsive disorder (OCD) and to evaluate its safety. METHODS: Fixed-dose fluoxetine hydrochloride (20 mg/d, 40 mg/d, 60 mg/d) was compared with placebo in two randomized, double-blind, parallel, 13-week trials of identical design in 355 outpatients with OCD aged 15 to 70 years (DSM-III-R criteria; 1 year's duration or longer; depression secondary if present). RESULTS: Fluoxetine (all doses) was significantly (P < or = .001) superior to placebo on the Yale-Brown Obsessive Compulsive Scale (Y-BOCS) total score (mean baseline-to-end-point decrease, 4.6, 5.5, and 6.5 vs 0.9, respectively, studies pooled) and other efficacy measures (P < or = .01). A trend suggesting greater efficacy at 60 mg/d was observed. Most patients (79.2%) completed the study. Eight adverse events were statistically significantly more frequent with fluoxetine and one, with placebo. For some events, incidence tended to increase with increasing dosage; however, few patients discontinued treatment for any single event. CONCLUSION: Fluoxetine was associated with a statistically significant reduction in OCD severity, including time engaged in obsessional and/or compulsive behaviors. Adverse events infrequently led to study discontinuation.

Multifocal fibrosis involving the thyroid, face, and orbits.

Our patient had fibrous thyroiditis with unusual facial and retroorbital fibrosis. Systemic fibrosis, which is more typical of the multifocal fibrosclerosing syndromes, was not evident. Treatment with steroid alone, surgery, and radiotherapy failed to maintain a remission. The combination of steroid and cyclophosphamide therapy appears to have been successful.

Neuroendocrine and behavioral responses to challenge with the indirect serotonin agonist dl-fenfluramine in adults with obsessive-compulsive disorder.

Neuroendocrine and behavioral responses to a single 60-mg oral dose of the indirect serotonin agonist dl-fenfluramine were assessed in unmedicated adults with obsessive-compulsive disorder (OCD) and neuroendocrine results contrasted with those in normal control subjects. Net fenfluramine-induced prolactin release did not differ significantly between OCD patients and normal controls. Prolactin responses in the OCD group were not significantly correlated with baseline Yale-Brown Obsessive Compulsive Scale scores for either obsessions or compulsions, but were positively correlated with the baseline Hamilton Depression Scale score and Hamilton Anxiety Scale score. No clear difference in the severity of patients' obsessions or compulsions was found following challenge with fenfluramine versus placebo. Although the present study does not demonstrate a serotonergic abnormality in OCD, this may be more a reflection of limitations of the test procedures than evidence that central nervous system (CNS) serotonergic function is normal in the disorder.

Emotion-focused psychotherapy for patients with panic disorder.

OBJECTIVE: Recent studies have suggested that most patients treated for panic disorder receive forms of psychotherapy other than cognitive behavior therapy, even though there is little information about the efficacy of such treatments or how they compare to proven active treatments. The authors compared one of these other forms, emotion-focused psychotherapy (given to 30 patients with panic disorder), to results obtained with recommended standard treatment (either cognitive behavior therapy [N=36] or imipramine [N=22]). The authors also compared emotion-focused psychotherapy to results obtained in subjects given pill placebo (N=24). METHOD: Subjects met DSM-IV criteria for panic disorder with no more than mild agoraphobia. Treatment consisted of approximately 3 months of weekly visits followed by 6 monthly maintenance visits. Assessments were conducted after each treatment phase and at a follow-up visit after 6 months of no treatment. RESULTS: Emotion-focused psychotherapy was less effective for symptoms of panic disorder than treatment with either cognitive behavior therapy or imipramine; results obtained with emotion-focused psychotherapy after the acute and maintenance phases were similar to those seen with placebo. Treatment expectations were not different among the different groups. Patients receiving emotion-focused psychotherapy had the highest completion rate. CONCLUSIONS: The results suggest that emotion-focused psychotherapy (a supportive form of psychotherapy) has low efficacy for the treatment of panic disorder. However, emotion-focused psychotherapy may be superior to medical management in helping patients stay in treatment.

A psychodynamic model of panic disorder.

Current psychiatric research on panic disorder and its treatment are heavily influenced by neurobiological and cognitive-behavioral models rather than psychodynamic propositions, and psychodynamic treatment is generally considered to be of little benefit in amelioration of symptoms. However, because neither of the current models fully explains the clinical psychopathology, etiology, or pathogenesis of panic disorder, there is a need for further model building. The authors suggest that a psychodynamic approach may add to the understanding of patients with panic disorder. They base their psychodynamic formulation on pilot interviews with nine patients with panic disorder, published reports of psychological characteristics of patients with panic disorder, and data from infant and animal research on temperament. Interview results included the following: 1) all of the patients described themselves as fearful, nervous, or shy as children, 2) they remembered their parents as angry, frightening, critical, or controlling, 3) they frequently indicated discomfort with aggression, 4) most described chronic feelings of low self-esteem, 5) their spouses were characterized as passive, kind, and nonaggressive, and 6) stressors associated with frustration and resentment preceded the onset of panic. The authors propose a model in which inborn neurophysiological irritability predisposes to early fearfulness. Exposure to parental behaviors that augment fearfulness results in disturbances in object relations and persistence of conflicts between dependence and independence, which predispose to fears of feeling trapped, suffocated, and unable to escape and/or feeling alone and unable to get help. Catastrophic fears of helplessness in the face of suffocation or abandonment are easily accessible.(ABSTRACT TRUNCATED AT 250 WORDS)

An empirical study of defense mechanisms in dysthymia.

OBJECTIVE: The psychodynamic approach to understanding dysthymia has rarely been empirically tested. In this pilot study the Defense Mechanism Rating Scales were used to examine psychodynamic data from patients with dysthymia and patients with panic disorder in order to test the hypotheses that 1) dysthymic patients would be similar to panic patients in endorsing primarily lower-maturity defense mechanisms, 2) dysthymic patients would use a distinct pattern of defense mechanisms, different from that of panic patients, and 3) dysthymic patients would endorse more frequently than panic patients four individual defenses that tend to handle anger and low self-esteem poorly: devaluation, passive aggression, projection, and hypochondriasis. METHOD: Twenty-two subjects meeting the DSM-III-R criteria for primary early-onset dysthymia and 22 subjects meeting the DSM-III-R criteria for primary panic disorder were interviewed on videotape and rated on the Defense Mechanism Rating Scales. RESULTS: The dysthymic subjects scored significantly higher on narcissistic, disavowal, and action defense levels and on the four individual defenses of devaluation, projection, passive aggression, and hypochondriasis, as predicted, as well as on two additional defenses, acting out and projective identification. Both groups tended to use lower-maturity defense mechanisms. CONCLUSIONS: The defense mechanism profile identified for dysthymia differs from that for panic disorder and supports particular psychodynamic hypotheses about chronic depression. It could be useful in devising treatment strategies and as a measure of treatment efficacy.

The panic-agoraphobic spectrum: a descriptive approach to the assessment and treatment of subtle symptoms.

OBJECTIVE: Psychiatric classification is still a topic of considerable discussion and debate in spite of major advances in the past two decades. The debate involves categorical versus dimensional approaches, cutoff numbers of symptoms to define a case, degree of impairment, objective diagnostic criteria versus more theoretically based criteria, episodic versus trait-like symptoms, and the role of atypical and subclinical symptoms. All of these issues have been raised for the anxiety disorders and depression. This article presents the conceptualization of a relatively novel and testable approach to the diagnosis and classification of panic and agoraphobia, the panic-agoraphobic spectrum, and pilot data on a new questionnaire to assess it. METHOD: Pilot testing of the Panic-Agoraphobic Spectrum Questionnaire was undertaken with 100 inpatients who had lifetime diagnoses of panic disorder, unipolar depression, comorbid panic and unipolar depressive disorders, or an eating disorder. The instrument emphasizes impairment related to 144 behaviors and experiences in seven panic-agoraphobic symptom domains. RESULTS: Patients with panic disorder scored highest on the questionnaire, and those with comorbid depression showed even greater severity of illness. The scores of the patients with eating disorders and of the depressed patients differed from those of the other groups but also differed from 0. CONCLUSIONS: The spectrum model of panic and agoraphobia is a flexible and comprehensive means of describing this clinical complex. The proposed model, complementary to the categorical approach, presumably expresses a unitary pathophysiology. Its usefulness is discussed in terms of its value for patient-therapist communication, outcome measures, identification of subtle personality traits, and subtyping of patients for research and treatment.

Treatment outcomes for primary care patients with major depression and lifetime anxiety disorders.

OBJECTIVE: Major depression occurs with generalized anxiety disorder and panic disorder in up to 60% of psychiatric and primary care patients. This comorbidity has been associated with greater severity of depression, poorer psychosocial functioning, and poorer treatment outcomes in psychiatric samples. This study examined the clinical outcomes for depressed primary care patients with and without a lifetime anxiety disorder. METHOD: A total of 157 primary care patients who met criteria for major depression were randomly assigned to standardized interpersonal psychotherapy or pharmacotherapy with nortriptyline and were assessed at baseline and at 4 and 8 months on severity of depression, psychosocial functioning, and health-related functioning. RESULTS: Depressed patients with a comorbid anxiety disorder presented with significantly more psychopathology and tended to prematurely terminate treatment more frequently than patients with major depression alone. Both standardized depression-specific treatments were effective for depressed patients with and without a comorbid generalized anxiety disorder, although time to recovery was longer for the former. Patients with lifetime panic disorder showed poor recovery in response to psychotherapy or pharmacotherapy. CONCLUSIONS: Standardized psychotherapy and pharmacotherapy are effective for patients with major depression with and without a generalized anxiety disorder. However, the longer time to recovery for the former group and lack of response to these treatments by patients with lifetime panic disorder suggest that primary care physicians should carefully assess history of anxiety disorder among depressed patients so as to select a proper intervention.

Diagnosis of nonpsychotic patients in community clinics.

OBJECTIVE: Diagnosis-specific, proven efficacious treatments are a major recent advance in psychiatry. Appropriate use of such treatments presupposes patients who meet the diagnostic criteria and clinicians who have accurately diagnosed the target disorder and comorbid conditions. Since little is known about whether these prerequisites are commonly met, the authors conducted a study at two community treatment sites to determine the frequency of various axis I diagnoses and the concordance between the diagnoses recorded in patient charts and those obtained by a structured interview. Given that a DSM diagnosis may not be sufficient to understand a patient's problems, the authors also obtained ratings of interpersonal functioning. METHOD: The subjects were 164 nonpsychotic patients who were seen at a rural (N=114) or urban (N=50) community treatment facility. Raters trained to reliably use the Structured Clinical Interview for DSM-IV (SCID) conducted diagnostic interviews. Clinical charts were reviewed to obtain clinical diagnoses. Patients completed questionnaires regarding interpersonal functioning. RESULTS: Most (N=145, 88%) of the patients met the SCID criteria for a current axis I diagnosis, and 53% (N=87) met the criteria for two or more disorders. Clinical and SCID diagnoses had poor agreement. Evidence was found for interpersonal dysfunction. CONCLUSIONS: Most patients met the diagnostic criteria for conditions for which there are proven treatments; however, inaccurate diagnosis proved common. This barrier to optimal treatment could be ameliorated with the use of structured interviews for common diagnoses. Scores on social/interpersonal measures support the premise that DSM symptoms provide only part of the relevant information about patients' conditions.

Self-reported anxiety, general medical conditions, and disability bed days.

OBJECTIVE: This study examined the effect of self-reported anxiety on the number of days persons with various general medical conditions spend in bed owing to disability. METHOD: Self-reported medical illness and disability data from a nationally representative household survey sample (N = 20,884) were analyzed. RESULTS: Among respondents with general medical conditions, those with self-reported anxiety had a nearly fourfold greater length of disability (mean = 18.0 bed days) than the nonanxious respondents (mean = 4.8 bed days). After adjustment for differences in demographic characteristics and burden of general medical illness, anxiety was associated with an additional 3.8 bed days. CONCLUSIONS: Self-reported anxiety in combination with general medical conditions may be associated with extensive functional impairment.