Maintaining reliability in a long-term psychiatric study: an ongoing inter-rater reliability monitoring program using the longitudinal interval follow-up evaluation.

The Longitudinal Interval Follow-up Evaluation (LIFE), has been shown to be a valid and reliable instrument for characterizing the week-by-week course of anxiety disorders examined retrospectively over the period of 1 year. Due to the chronic nature of these disorders, there is a need for reliable, valid instruments for measuring course over periods of several years if we are to learn more about the natural history of these disorders. This paper describes a rater-monitoring program designed to ensure long-term inter-rater reliability and prevent "rater drift". In this program, clinical interviewers score taped interviews and are required to maintain a median intra-class correlation coefficient (ICC) of at least 0.80 with the other raters. Raters also assess tapes from previous years, to ensure that they are using the same diagnostic criteria as earlier generations of interviewers. A reliability study was conducted to compare psychiatric status ratings (PSRs) collected using biweekly telephone interviews with the semi-annual interviews. The ICCs for panic, agoraphobia, social phobia, and generalized anxiety disorder were very good to excellent. Another reliability study examined the PSRs of subjects who had been previously lost to follow-up. ICCs for panic, agoraphobia, generalized anxiety disorder and depression were good to excellent. These results show that the LIFE, when used in conjunction with an intensive training and rater monitoring system, is a reliable instrument for use in longitudinal studies of the course of anxiety disorders.

Patterns of personality pathology in patients with generalized anxiety disorder, panic disorder with and without agoraphobia, and social phobia.

The relationship between co-occurring personality disorders and anxiety disorders (panic disorder with or without agoraphobia, social phobia, and generalized anxiety disorder) was examined, taking into account the effect of major depression. This article describes findings for 622 participants in the Harvard/Brown Anxiety Research Project, a longitudinal follow-up study of DSM-III-R-defined anxiety disorders. A total of 24% of participants had at least one personality disorder, with avoidant, obsessive compulsive, dependent, and borderline most common. Generalized anxiety disorder, social phobia, and major depression were positively associated with the occurrence of one or more personality disorders, whereas panic disorder with agoraphobia was not associated. Major depression was associated in particular with dependent, borderline, histrionic, and obsessive compulsive personality disorders and social phobia was associated with avoidant personality disorder. Whereas some of our findings confirm results from earlier studies, others are somewhat inconsistent with previous results and indicate the need for further investigation.

Suicidal behavior in patients with current or past panic disorder: five years of prospective data from the Harvard/Brown Anxiety Research Program.

OBJECTIVE: The authors' goals were to examine predictors of suicidal behavior and provide guidelines for assessing suicide risk in patients with panic disorder. METHOD: Four hundred ninety-eight patients with panic disorder were followed for 5 years. Survival analysis was used to examine variables correlated with prospectively observed suicidal behavior. RESULTS: Subjects had a 0.06 probability of suicidal behavior during follow-up. Affective disorders, substance abuse, eating disorders, personality disorders, and being female were risk factors. Two subjects were suicidal in the absence of risk factors; both developed depression during follow-up. CONCLUSIONS: Panic disorder is not associated with suicidal behavior in the absence of other risk factors.

Personality disorders, history of trauma, and posttraumatic stress disorder in subjects with anxiety disorders.

The purpose of the study was to examine the association of personality disorders, history of trauma, and posttraumatic stress disorder (PTSD) in a large sample of subjects with anxiety disorders. Categorical and continuous indices of personality disorders were compared in three groups from the Harvard/Brown Anxiety Disorders Research Project (HARP): subjects with no history of trauma (n = 403), subjects with a history of trauma but no history of PTSD (n = 151), and subjects with a current or past diagnosis of PTSD (n = 68). Subjects with PTSD were more likely to meet criteria for borderline or self-defeating personality disorder than subjects in the other two groups. PTSD subjects also had higher scores on the continuous measures (total number of criteria met) for borderline and self-defeating personality disorder than the other two groups. The findings suggest that a diagnosis of PTSD rather than a history of trauma is associated with borderline and self-defeating personality disorder features. Alternative conceptualizations of axis II features in individuals with PTSD are discussed.

Psychosocial treatment prescriptions for generalized anxiety disorder, panic disorder, and social phobia, 1991-1996.

OBJECTIVE: Pharmacologic prescriptions for anxiety disorders have changed significantly in the last decade. This article investigates whether psychosocial treatments, as reported by 362 subjects in the Harvard/Brown Anxiety Disorders Research Program from 1991 to 1996, changed as well. METHOD: Subjects were interviewed in 1991 and 1995-1996 to determine which psychosocial treatments (behavioral, cognitive, dynamic, or relaxation or meditation) they had received. RESULTS: The percentage of subjects who received each type of psychosocial treatment either declined or remained the same from 1991 to 1995-1996. Dynamic psychotherapy remained the most frequently used method of these four. The percentage of subjects receiving any such method declined. CONCLUSIONS: Behavioral and cognitive treatment, two empirically validated forms of psychotherapy, were less frequently used than dynamic psychotherapy, which lacks such validation. All use of verbal treatment methods declined from 1991 to 1995-1996.

Comparing primary and secondary generalized anxiety disorder in a long-term naturalistic study of anxiety disorders.

This study explores the potential differences in comorbidity and course between primary generalized anxiety disorder (GAD), which develops before other anxiety disorders, and secondary GAD. As part of the Harvard/Brown Anxiety Research Project (HARP), a naturalistic, long-term, longitudinal study of 711 subjects from a variety of clinic settings with DSM III-R defined anxiety disorders, 210 subjects with GAD were identified. Of these, 78 (37%) had primary GAD, and 84 (40%) had secondary GAD; of the remainder, 28 (13%) had no other anxiety disorder and 20 (10%) developed GAD within a month of another anxiety disorder and were excluded from the analysis. All subjects were comorbid for at least one other anxiety disorder. Primary GAD subjects were more likely to be in episode at intake (90% vs. 77%, P = .04) and less likely than secondary GAD subjects to have current or past agoraphobia without panic disorder (3% vs. 11%, P = .04), social phobia (19% v. 52%, P = .001), simple phobia (14% v. 30%, P = .02), or post traumatic stress disorder (5% vs. 20%, P = .01). Subjects with primary GAD were also less likely to have current or past alcohol use disorders (17% vs. 37%, P = .004) or major depressive disorder (60% vs. 76%, P = .03). There were no significant differences in either treatment approaches or remission rates for primary compared to secondary GAD. Whether GAD first occurs before or after another anxiety disorder, it is similar in terms of prevalence, treatment, and course. The only significant differences between primary and secondary GAD lie in the rates of comorbidity of both other anxiety disorders and non-anxiety disorders, including major depression and substance abuse. These results support the concept of GAD as a valid, separate and distinct entity, whether it occurs primarily or secondarily.

Patterns of remission and relapse in obsessive-compulsive disorder: a 2-year prospective study.

OBJECTIVE: This study examined the course of illness in patients with obsessive-compulsive disorder (OCD) over a 2-year period. METHOD: Sixty-six patients with a primary diagnosis of DSM-III-R OCD were followed prospectively for 2 years. Baseline information was collected on demographic characteristics, Axis I and II diagnoses, and severity of OCD symptoms. Follow-up measures obtained at 3, 6, 12, and 24 months after baseline assessment included information on symptomatic and diagnostic status as well as behavioral and somatic treatments received. RESULTS: The probability of full remission from OCD over the 2-year period was 12%. The probability of partial remission was 47%. After achieving remission from OCD, the probability of relapse was 48%. No factors were identified that significantly predicted full or partial remission. Seventy-seven percent (N = 51) of the subjects received a serotonin reuptake inhibitor (SRI) for > or =12 weeks, and 68% (N = 45) received medium-to-high doses of SRIs for > or =12 weeks. Only 18% received a full trial of behavior therapy. CONCLUSION: Despite exposure to at least 1 adequate trial of an SRI, the likelihood of full remission of OCD in this study was low. Results of this study also suggest that behavior therapy may be under-utilized.

Prospective study of fluoxetine treatment and suicidal behavior in affectively ill subjects.

OBJECTIVE: There has been speculation in the literature about a link between fluoxetine use and suicidal behavior. The authors of this study hypothesized that there is no elevation in risk of suicidal behavior associated with use of fluoxetine. METHOD: The data come from the National Institute of Mental Health Collaborative Depression Study, a prospective, naturalistic follow-up of persons who presented for treatment of affective disorders. The analyses included data on 643 subjects who were followed up after fluoxetine was approved by the Food and Drug Administration in December 1987 for the treatment of depression. RESULTS: Nearly 30% (N = 185) of the study group was treated with fluoxetine at some point during the follow-up period. Relative to the other subjects, those who were subsequently treated with fluoxetine had onset of affective illness at a younger age and, after intake into the study and before 1988, had elevated rates of suicide attempts before fluoxetine treatment. A mixed-effects survival analysis that incorporated treatment exposure time, multiple treatment trials, and multiple suicide attempts per subject showed that relative to no treatment, use of fluoxetine and use of other somatic antidepressants were associated with nonsignificant reductions in the likelihood of suicide attempts or completions. Severity of psychopathology was strongly associated with elevated risk, and each suicide attempt after intake into the Collaborative Depression Study was associated with a marginally significant increase in risk of suicidal behavior. CONCLUSIONS: The results do not support the speculation that fluoxetine increases the risk of suicide. Rather, there was a nonsignificant reduction in risk of suicidal behavior among patients treated with fluoxetine, even though those subjects were more severely ill before treatment with fluoxetine.

Simple phobia as a comorbid anxiety disorder.

This study sought to describe clinical and demographic characteristics differentiating patients with DSM-III-R simple phobias comorbid with one or more of five DSM-III-R index anxiety disorders as compared with those with the index diagnoses alone. From 711 subjects participating in a multicenter, longitudinal, naturalistic study of anxiety disorders, 115 subjects with comorbid simple phobias were compared with 596 subjects without simple phobias in terms of demographic data, comorbidity with other disorders, somatic and psychosocial treatment received, and quality of life. In addition, episode characteristics, types of simple phobias found, and course of illness were specified. Subjects with simple phobias had more additional comorbid anxiety disorders by history than did those without. Mean length of intake episode was 22.43 years and severity was typically moderate. Fears of heights and animals were the most commonly represented simple phobias. Subjects with uncomplicated panic disorder were less likely to have comorbid simple phobias than were subjects with other index diagnoses, and subjects with simple phobia were more likely to have comorbid posttraumatic stress disorder than were these without simple phobia. Subjects with and without simple phobias did not differ by somatic or psychosocial treatment received or in terms of quality of life. Simple phobia appeared in this study to be a chronic illness of moderate severity for which behavioral treatment methods of recognized efficacy were not being frequently utilized. Uncomplicated panic disorder may reflect some type of resistance to phobia development.

The psychosocial treatments interview for anxiety disorders. A method for assessing psychotherapeutic procedures in anxiety disorders.

The authors report on development, reliability, and findings of the Psychosocial Treatments Interview (PTI) to assess treatments reported by patients in a naturalistic study of the longitudinal course of anxiety disorders. The PTI ascertains frequency of different types of psychosocial treatments, based on patients' reports. The PTI showed good internal consistency and very good interrater reliability. At first 6-month follow-up, the most common modalities were supportive, medication discussion, and dynamic intervention. Combinations were common. Delivery of treatments differed by site. Overall, the PTI fills a methodological need for the assessment of the treatments reported by patients in naturalistic follow-up studies.

Does major depression result in lasting personality change?

OBJECTIVE: Individuals with a history of depression are characterized by high levels of certain personality traits, particularly neuroticism, introversion, and interpersonal dependency. The authors examined the "scar hypothesis," i.e., the possibility that episodes of major depression result in lasting personality changes that persist beyond recovery from the depression. METHOD: A large sample of first-degree relatives, spouses, and comparison subjects ascertained in connection with the proband sample from the National Institute of Mental Health Collaborative Program on the Psychobiology of Depression were assessed at two points in time separated by an interval of 6 years. Subjects with a prospectively observed first episode of major depression during the interval were compared with subjects remaining well in terms of change from time 1 to time 2 in self-reported personality traits. All subjects studied were well (had no mental disorders) at the time of both assessments. RESULTS: There was no evidence of negative change from premorbid to postmorbid assessment in any of the personality traits for subjects with a prospectively observed first episode of major depression during the interval. The results suggested a possible association of number and length of episodes with increased levels of emotional reliance and introversion, respectively. CONCLUSIONS: The findings suggest that self-reported personality traits do not change after a typical episode of major depression. Future studies are needed to determine whether such change occurs following more severe, chronic, or recurrent episodes of depression.

The infrequency of "pure culture" diagnoses among the anxiety disorders.

BACKGROUND: Anxiety disorders are known to commonly coexist in individuals, both with other anxiety disorders and with mental disorders from other groupings, such as affective disorders. We questioned how frequently anxiety disorders actually occur in isolation, as "pure cultures." METHOD: We examined diagnostic patterns among the 711 subjects entered into a large, multicenter study of anxiety disorders, the Harvard/ Brown Anxiety Disorders Research Program (HARP), which focused on panic, agoraphobia, generalized anxiety disorder, and social phobias as "index disorders" required for intake. RESULTS: We used various definitions for "pure culture." By all definitions, subjects with "pure culture" represented a minority, especially in cases of generalized anxiety disorder and social phobia, where comorbidity was virtually ubiquitous. "Pure culture" status was associated with later onset of illness and less chronicity. CONCLUSION: Future studies of anxiety disorder should aim to document the extensive comorbidity, rather than eliminate it by restrictive diagnostic exclusion criteria, lest they yield atypical or even misrepresented groups of patients. Clinicians should not stop at identifying only the "main" diagnosis but look for other, comorbid diagnoses that are often present.

Serum lithium levels and psychosocial function in patients with bipolar I disorder.

OBJECTIVE: This study compared the effect of two different serum lithium levels on the psychosocial functioning of patients with bipolar I disorder. METHOD: Ninety patients with bipolar I disorder were enrolled in a prospective, double-blind, maintenance trial of lithium. The patients were randomly assigned to treatment with doses of lithium adjusted to achieve a serum lithium concentration of either 0.8 to 1.0 mmol/liter (standard) or 0.4 to 0.6 mmol/liter (low). The Longitudinal Interval Follow-Up Evaluation was used to assess psychosocial functioning in the areas of work, interpersonal relationships, and global functioning. All observed values were analyzed with a mixed-effects analysis of covariance. Independent variables included treatment group (low or standard lithium serum level), relapse status, socioeconomic status, time from random treatment assignment to assessment, termination of protocol before or after relapse, length of remission before random treatment assignment, polarity of the last mood episode before random treatment assignment, and number of mood episodes in the 3 years before random treatment assignment. RESULTS: Relapse was associated with large negative effects on psychosocial functioning. Patients in higher socioeconomic brackets had better psychosocial functioning than did those in lower brackets. Patients receiving lithium doses that achieved standard serum levels had better psychosocial functioning than those receiving doses that achieved low serum levels; this effect was partially but not wholly mediated through relapse prevention. CONCLUSIONS: For patients with bipolar I disorder, standard serum lithium levels may enhance psychosocial functioning, above and beyond the effects of relapse prevention.

The relationship between fluoxetine use and suicidal behavior in 654 subjects with anxiety disorders.

BACKGROUND: In the past few years, there has been controversy over the relationship between suicidal behavior and fluoxetine use. This report examines the relationship between fluoxetine use and suicidal behavior in the Harvard/Brown Anxiety Disorders Research Program (HARP). METHOD: HARP is a naturalistic, prospective, longitudinal anxiety disorders study. Probabilities of suicidal behavior for 654 subjects were examined using life table analysis for the study group as a whole and stratified by depression status at intake. RESULTS: Subjects not using fluoxetine during follow-up had almost twice the probability of making a suicide attempt or gesture during the follow-up than subjects who were using fluoxetine, although this difference was not statistically significant. Subjects having episodes of major depressive disorder (MDD) at intake were more likely than those not having an episode to receive fluoxetine during follow-up (74/166 [45%] vs. 118/488 [24%], chi squared= 24.85, df= 1, p < .0001). Among those subjects having episodes of MDD at intake, there was a statistically significantly lower probability of suicide attempts/gestures for those taking fluoxetine than for those not using fluoxetine during follow-up (log-rank chi squared= 5.10, df= 1, p= .02). CONCLUSION: We found no evidence that fluoxetine use is associated with increased risk of suicide attempts or gestures. However, we did find that subjects with more suicide risk factors at intake were more likely to use fluoxetine than those without these risk factors.

Phenomenology and course of generalised anxiety disorder.

BACKGROUND: The diagnostic category of generalised anxiety disorder (GAD) was originally intended to describe residual anxiety states. Over the years clinical criteria have been refined in an attempt to describe a unique diagnostic entity. Given these changes, little is known about the clinical course of this newly defined disorder. This study investigates the longitudinal course, including remission and relapse rates, for patients with DSM-III-R defined GAD. METHOD: Analysis of the 164 patients with GAD participating in the Harvard Anxiety Research Program. Patients were assessed with a structured clinical interview at intake and re-examined at six month intervals for two years and then annually for one to two years. Psychiatric Status Ratings were assigned at each interview point. Kaplan-Meier curves were constructed to assess likelihood of remission. RESULTS: Comorbidity was high, with panic disorder and social phobia as the most frequently found comorbid disorders. The likelihood of remission was 0.15 after one year and 0.25 after two years. The probability of becoming asymptomatic from all psychiatric symptoms was only 0.08. CONCLUSIONS: This prospective study confirms the chronicity associated with GAD and extends this finding to define the one and two year remission rates for the disorder. Likelihood of remission for GAD and any other comorbid condition after one year was half the annual remission rate for GAD alone.

The role of drug and alcohol abuse in recent increases in depression in the US.

Previous studies have reported an increase in depression among recent birth cohorts. Concurrent with the increase in rates of depression, there have been increases in rates of drug and alcohol abuse and dependence. This study sought to determine if the recent increase in rates of depression could be attributed to co-morbid alcohol and drug abuse. The data derived from two studies: (1) a sample of relatives of probands with affective disorder; and (2) a community survey of the US population. The piecewise exponential statistical model was applied to evaluate the association of gender, age, period and birth cohort with rates of major depressive disorder (MDD) separately for those with, and without, diagnoses of alcohol or drug abuse. Elevated rates of MDD occurred among those with co-morbid drug and alcohol abuse in both the family and community samples. However, there were also temporal increases in rates of MDD in those with no such co-morbidity. Specifically there were effects of age and gender for both studies; in addition, there was a period effect in the family study and a birth cohort effect in the community sample. The recent increases in depression in the US cannot be accounted for solely by concurrent increases in co-morbid drug and alcohol abuse. Temporal (period and cohort) effects on rates of depression occur in addition to the contribution of co-morbid drug and alcohol abuse or dependence.

Benzodiazepine use in anxiety disordered patients with and without a history of alcoholism.

BACKGROUND: People with a history of alcohol use disorders are thought to be at risk for misusing prescribed benzodiazepines. We examine the use of prescribed benzodiazepines in anxiety disordered subjects with and without a history of alcohol dependence or abuse. METHOD: A group of 343 subjects in the Harvard/Brown Anxiety Disorders Research Program (HARP) who were taking benzodiazepines at the time of entry into a prospective study of anxiety disorders serve as the study group. Subjects with (N=99) and without (N=244) a history of alcohol use or dependence (DSM- III-R) are examined for their reported total daily dose, p.r.n. use, or continued use of benzodiazepines. RESULTS: There is no significant difference in maximum daily dose or continued use of benzodiazepines over 12 months of follow- up. There is a clinically small but statistically significant difference in median daily dose during the second but not the first 6 months of follow-up for the alcohol history positive versus alcohol history negative groups. Additionally, there was significantly less reported use of p.r.n. benzodiazepines in the alcohol history positive versus alcohol history negative subjects during the second 6 months, but not the first 6 months, of follow-up. CONCLUSION: The presence or absence of a history of alcohol use disorders is not a strong predictor of the use of benzodiazepines in subjects with anxiety disorders over 12 months of prospective follow-up.

Prevalence of somatoform disorders in a large sample of patients with anxiety disorders.

The authors investigated the prevalence and characteristics of somatoform (SOM) disorders among 654 subjects with anxiety disorders who were part of the larger Harvard/Brown Anxiety Disorders Research Project. Thirty-six (5.5%) of the subjects had past or current SOM disorders. The subjects with SOM disorders were significantly more likely to have histories of posttraumatic stress disorder (22% vs. 8%, P = 0.01). The subjects with generalized anxiety disorder had significantly higher rates of SOM disorder (9.2% vs. 4.0%, P = 0.01). These results add support to the observation that SOM disorders are frequently comorbid with anxiety and depressive disorders.

Profile of a large sample of patients with social phobia: comparison between generalized and specific social phobia.

This study examines a large cohort of subjects with social phobia, as part of a larger naturalistic and longitudinal study of 711 subjects with anxiety disorders. We focused on 176 subjects who were in an episode of social phobia at intake. We were particularly interested in evaluating the diagnostic distinction between generalized and specific social phobia. We compared these two groups along demographic characteristics, comorbidities, psychosocial functioning (health, role functioning, social functioning, and emotional functioning) and global assessment scores. We found that generalized social phobics tended to have an earlier age of onset as compared to the specific group; however, this is not a statistically significant difference at this level of analysis. The two groups did not differ for the current comorbidities examined. We observed no differences in the treatment received by the two types of social phobia subjects, and the two groups functioned equally well in terms of health and fulfilling social roles. In addition, we examined adverse childhood events (i.e., death of a parent, childhood abuse) and found no evidence for any differential impact these events might have on the type of social phobia. Although we did observe significantly greater fear of public speaking among the specific compared to the generalized group, which may indicate a qualitative difference between the subtypes, our results suggest that for most parameters, generalized and specific social phobia represent a continuum of similar and overlapping entities.