Pollen counts and suicide rates. Association not replicated.

OBJECTIVE: To replicate a previously reported association between pollen counts and county suicide rates in the continental United States, across space and time. METHOD: The authors evaluated the relationship between airborne pollen counts and suicide rates in 42 counties of the continental United States, containing a pollen-counting station participating in the Aeroallergen Monitoring Network in the United States (N = 120,076 suicides), considering years' quarter, age group, sex, race, rural/urban location, number of local psychiatrists, and median household income, from 1999 to 2002. The county-level effects were broken into between-county and within-county. RESULTS: No within-county effects were found. Between-county effects for grass and ragweed pollen on suicide rates lost statistical significance after adjustment for median income, number of psychiatrists, and urban vs. rural location. CONCLUSION: Future research is necessary to reappraise the previously reported relationship between pollen levels and suicide rates that may have been driven by socioeconomic confounders.

High-frequency measurement of depressive severity in a patient treated for severe treatment-resistant depression with deep-brain stimulation.

Although there have been previous studies of deep-brain stimulation (DBS), we present, to our knowledge, the first example of high-frequency depressive severity measurement-based DBS treatment in particular and psychiatric treatment in general. Daily post-surgical e-mail prompts for a period of 6 months resulted in 93 administrations of a computerized adaptive test (CAT) of depression severity (CAT-Depression Inventory or CAT-DI) via the internet. There was an average of 3.37 weekly measurements with an average separation of 2.12 days. No additional incentive was provided to the patient for completing the adaptive tests. The patient is a 55-year-old female with six psychiatric hospitalizations for depression, two suicide attempts, marginal response to eight electroconvulsive therapy (ECT) treatments and 35 psychotropic medications. We report results after high-frequency stimulation of the superolateral branch of the medial forebrain bundle. The CAT-DI was used for daily assessments before, during and after (remotely in response to an e-mail prompt) the DBS procedure. Two follow-up Hamilton Depression Scales (HAM-Ds) were also collected. Response to treatment varied markedly, with a decrease from severe (>75) to mild (60), which is three times the size of the uncertainty level. Although the HAM-D scores decreased, they missed the more complete temporal pattern identified by CAT-DI daily monitoring. We demonstrated feasibility of daily depressive severity measurement at high levels of precision and compliance. Clinician ratings confirm the general pattern of treatment benefit, but mask the marked variability in mood and more marked periods of benefit and decline.

Assessing anhedonia in psychiatric patients.

In two studies of depressed, manic, schizophrenic, and normal subjects, a scale for measuring the intensity of subjects' pleasureable responses to normally emjoyable situations (the Pleasure Scale) evidenced good internal reliability and moderate agreement with the Chapman Anhedonia Scale and Indexes of depressive symptom severity. Only the depressed patients showed extremely anhedonic responses. Although more than half the depressed patients evidenced pleasure scores in the normal range, about 185 of them seemed more anhedonic than any norma subject. A mixture analysis resolved depressed patient scores into two distinct distributions: a normal-range distribution (88% of depressives) and an extremely anhedonic distribution (12%). The findings provide some support for the existence of a qualitatively distinct subtype of major depression that has been variously defined an "endogenomorphic" or "melancholic."

Lithium ion transport and affective disorders within families of bipolar patients. Identification of a major gene locus.

We have identified a genetic polymorphism in which one allele results in elevated RBC lithium ion ratios and also contributes to vulnerability to some forms of affective disorders. Interindividual variability in the lithium ion ratio (the ratio of the RBC to the plasma lithium ion concentration) is determined by variability in a lithium ion ratio (the ratio of the RBC to the plasma lithium ion concentration) is determined by variability in a lithium-sodium ion counterflow mechanism. Segregation analyses were conducted on lithium ion ratios in vitro in members of 120 normal families and on both the lithium ion ratio and the history of affective disorder in first-degree relatives of 31 bipolar patients. These analyses provided evidence for an autosomal major gene locus that influences the lithium ion ratio both in members of normal families and in relatives of bipolar patients. The allele at the major locus resulting in elevated lithium ion ratios was associated with an increased likelihood of psychiatric hospitalization among relatives of bipolar patients.

Methodology in psychiatric research. Report on the 1986 MacArthur Foundation Network I Methodology Institute.

This report discusses many of the issues raised during a two-day institute that focused on methodological problems encountered in psychiatric research. The topics range from the problems with psychiatric diagnosis and measurement to sampling issues and biases to specific statistical concerns. Attention is given to the need for improved communication and collaboration between psychiatric researchers and biostatisticians.

Pharmacologic evidence for specificity of pursuit dysfunction to schizophrenia. Lithium carbonate associated with abnormal pursuit.

Conflicting findings regarding the prevalence of abnormal smooth-pursuit eye movements in patients with major affective disorders call into question the specificity of impaired smooth-pursuit eye movements to schizophrenia. We report that pursuit is impaired in 88% of lithium carbonate-treated affective disorder patients whose pursuit was normal prior to receiving this drug. Over half of lithium carbonate-treated affective disorder patients in remission also showed impairment of smooth-pursuit eye movements. In conjunction with recent prevalence data on family members of psychiatric patients, the findings support the specificity of abnormal pursuit as a biological trait associated with schizophrenia, but not with the major affective disorders. The mechanisms by which lithium carbonate impairs pursuit are discussed.

A double-blind, placebo-controlled trial of lithium carbonate therapy for alcoholism.

The efficacy of lithium carbonate as a treatment for alcoholism was examined in a double-blind, placebo-controlled study of 104 men and women meeting DSM-III criteria for alcohol dependence. Subjects entered the study during inpatient treatment and were subsequently followed up for 12 months. Survival analysis disclosed essentially three categories of treatment response: one for noncompliant subjects (0% to 7% abstinent), one for compliant subjects not attaining therapeutic serum lithium levels (31% to 44% abstinent), and one for compliant subjects with therapeutic serum levels (67% abstinent). Two findings led us to believe that therapeutic serum levels of lithium were associated with better outcome over and above a behavioral compliance effect. First, in a dose-response analysis, serum lithium levels and abstinence rates were not linearly associated. Second, all subjects who started lithium carbonate therapy as inpatients were significantly less likely to relapse to drinking during the first month than were placebo-compliant subjects. There was no evidence that depressed alcoholics showed a better treatment response than nondepressed alcoholics or that lithium had any significant impact on the mood or social adjustment of alcoholics. Although the sample size and the difficulties of ascertaining placebo compliance caution against drawing firm conclusions, the data add further support to the hypothesis that lithium has an effect on drinking behavior not related to the treatment of affective symptoms.

Cerebrospinal fluid and urinary biogenic amines in depressed patients and healthy controls.

The National Institute of Mental Health-Clinical Research Branch Collaborative Study investigated 132 drug-free, severely depressed patients and 80 healthy controls. Forty-five percent of the depressed patients excreted markedly elevated levels of urinary epinephrine (E) and metanephrine (MET), while only 5% of healthy controls did so. Using gaussian mixture distributions, we identified two subgroups of depressed patients: one excreting normal levels and the other excreting high levels of urinary E, MET, norepinephrine, and normetanephrine. Cerebrospinal fluid homovanillic acid levels were low in a subgroup of depressed patients. When analyzed by subgroup, the elevated E + MET group had markedly lower cerebrospinal fluid homovanillic acid levels than controls, whereas depressed patients with normal catecholamine levels did not. Since it has been postulated that there are two subgroups of depressed patients, those with low 3-methoxy-4-hydroxyphenylglycol (MHPG) levels and normal 5-hydroxyindoleacetic acid (5-HIAA) levels and those with normal MHPG levels and low 5-HIAA levels, several analyses were performed to see if such a group could be identified. Our analysis failed to find evidence of a subgroup of depressives with low MHPG and normal 5-HIAA levels or normal MHPG and low 5-HIAA levels.

Some conceptual and statistical issues in analysis of longitudinal psychiatric data. Application to the NIMH treatment of Depression Collaborative Research Program dataset.

Longitudinal studies have a prominent role in psychiatric research; however, statistical methods for analyzing these data are rarely commensurate with the effort involved in their acquisition. Frequently the majority of data are discarded and a simple end-point analysis is performed. In other cases, so called repeated-measures analysis of variance procedures are used with little regard to their restrictive and often unrealistic assumptions and the effect of missing data on the statistical properties of their estimates. We explored the unique features of longitudinal psychiatric data from both statistical and conceptual perspectives. We used a family of statistical models termed random regression models that provide a more realistic approach to analysis of longitudinal psychiatric data. Random regression models provide solutions to commonly observed problems of missing data, serial correlation, time-varying covariates, and irregular measurement occasions, and they accommodate systematic person-specific deviations from the average time trend. Properties of these models were compared with traditional approaches at a conceptual level. The approach was then illustrated in a new analysis of the National Institute of Mental Health Treatment of Depression Collaborative Research Program dataset, which investigated two forms of psychotherapy, pharmacotherapy with clinical management, and a placebo with clinical management control. Results indicated that both person-specific effects and serial correlation play major roles in the longitudinal psychiatric response process. Ignoring either of these effects produces misleading estimates of uncertainty that form the basis of statistical tests of hypotheses.

Quantification of scalp hair--a computer-aided methodology.

A method for obtaining a quantitative assessment of hair density is described. First, a photographic image of the scalp is digitized onto a high-resolution computer graphics screen. Second, the frequency of each of 256 gray levels (one for each of 500 vertical X 500 horizontal = 250,000 locations on the screen) is obtained and the frequency histogram of gray levels is displayed. Third, a statistical procedure, gaussian mixture analysis, is used to resolve the frequency distribution into two normally distributed component distributions. The first component distribution describes the range of gray levels that are typically associated with hair. The second component distribution describes shades of gray that are typically associated with scalp. The statistical model provides a precise measure of the proportion of the head that exhibits gray levels in each of the two component distributions (hair or scalp). The proportion of the first component distribution is a scale-independent measure of hair density. The difference in this quantity before and after treatment provides an accurate quantitative determination of the change in hair density and hence of the efficacy of treatment.

Lack of a bimodal distribution of ventricular size in schizophrenia: a Gaussian mixture analysis of 1056 cases and controls.

The finding of clinical and laboratory differences between schizophrenic patients with large and small cerebral ventricles has led to the widespread assumption that large ventricles are a marker that characterizes a subgroup of patients with schizophrenia. We reviewed all published English language ventricle-to-brain ratio (VBR) studies in which individual data points were available (schizophrenics: n = 691, medical controls; n = 205, normal volunteers: n = 160). Using a univariate normal mixture model to examine the distribution of ventricular size in each group, we found no evidence of a mixture of Gaussian distributions (i.e., "bimodality") within any of the three groups. The same analysis was then performed on the combined sample of schizophrenic patients and normal and medical controls, respectively. In each case the improvement in fit of a mixture of normal distributions compared to a single component normal distribution was significant. The data do not support the notion that ventricular enlargement is a discontinuous marker of a subtype of schizophrenia.

Ion transport and adrenergic function in major affective disorder.

Altered functional levels of norepinephrine (NE) have been implicated in the etiology of affective disorders. Abnormalities of membrane ion transport have been postulated to underlie such neurotransmitter imbalances. To evaluate the role of these systems in the pathophysiology of specific affective syndromes, we have examined the NE metabolite MHPG and ion transport function in a cross-section of affective patients in different mood states. Analysis of those data reveals that plasma MHPG levels are dependent on mood state, with bipolar patients in the manic phase having significant elevations of this metabolite. In contrast, the red cell: plasma lithium ratio (LR) was independent of mood state and significantly elevated in a group of bipolar patients.

Mixture distributions in psychiatric research.

This paper describes the application of Gaussian mixture distributions to biological marker research in psychiatry. Mixtures of univariate and multivariate normal distributions can be used to determine if diagnostically similar psychiatric patients belong to biologically distinct subpopulations. The resulting biological subtypes may be important in understanding the etiology of psychiatric disorders. The general model and estimation procedure are described (EM algorithm; Dempster, Laird and Rubin 1977). The method is illustrated using two examples of biological data: (1) red cell membranes and monoamine oxidase activity data in normal individuals having no family history of psychiatric illness, the first-degree relatives of bipolar depressed patients and a heterogeneous patient population; and (2) smooth pursuit eye movements that classify relatives of schizophrenics, nonschizophrenics and normal controls into biologically distinct populations.

Assessment of the debrisoquin and dextromethorphan phenotyping tests by gaussian mixture distributions analysis.

From a sample of 149 unrelated Spaniards, individuals were phenotyped for their ability to hydroxylate debrisoquin and O-demethylate dextromethorphan. The distribution of urinary metabolic ratios for each test was analyzed by univariate gaussian mixture distributions analysis to determine the number of populations, the mean and standard deviation of the metabolic ratios for each population, and the proportion belonging to each population. For the 124 subjects phenotyped with both the debrisoquin and dextromethorphan tests a bivariate analysis was performed. The results demonstrate that both tests similarly separated this sample into two populations, with 10% belonging to poor metabolizer phenotypes. In addition, the correlation between the metabolic ratios from each test is significant, indicating that they are measuring the same biologic trait and the certainty of correctly identifying the debrisoquin oxidation phenotype of an individual is improved by using the results of both tests.

Efficacy of ECT: a meta-analysis.

The authors analyzed several rigorously controlled studies that compared the efficacy of ECT with that of simulated ECT, placebo, and antidepressants. The data from these studies were combined statistically (with the Mantel-Haenszel method for the combination of fourfold tables), showing ECT's clear superiority over all these other forms of treatment for severe depression. The authors similarly analyzed the data from several studies comparing the efficacy of unilateral nondominant ECT with that of bilateral ECT and found no significant difference in their efficacy.

Length of stay and recidivism in schizophrenia: a study of public psychiatric hospital patients.

OBJECTIVE: Psychiatric beds in public hospitals have decreased 80% since 1955, but admissions have risen correspondingly, largely because of high recidivism rates. Decreases in numbers of beds have been partly achieved by shortening the length of stay, which lessened by half between 1970 and 1980. This study was undertaken to determine whether duration of hospital treatment affects the rate and rapidity of relapse among schizophrenic patients. METHOD: Data on 1,500 patients from 10 state hospitals were gathered for 18 months after initial discharge. Predictor variables included age, sex, marital status, race, number of previous admissions, location of the facility, and length of stay. Data were analyzed by survival analysis with a Cox regression model for two times to initial relapse: 30 days and 18 months (outcome). RESULTS: Length of stay was significantly related to each time to relapse after the effects of number of previous admissions and age were partialed out. Facility location was not predictive, but intrahospital effects were tested by examining the data on the largest facility; again, length of stay significantly predicted relapse. CONCLUSIONS: Although the magnitude of the effect was small, the clinical significance of the findings is the greater likelihood that brief-stay patients will be rehospitalized within 30 days after discharge than will patients treated for longer periods. Brief hospitalization seems generally applicable to psychiatric populations, but there may be a small but important group of seriously mentally ill patients for whom other alternatives are possibly more appropriate and should be explored.

Regression toward the mean: more on the price of beer and the salaries of priests.

Investigators must take great care in studying the temporal association between biological and clinical data. By selecting patients on the basis of deviant levels of a biological parameter, it often is impossible to separate the effect of regression toward the mean from a temporal association of clinical interest. Experimental strategies that focus exclusively on subjects with extreme values are particularly prone to misinterpreting the effects of regression toward the mean as a temporal change in the variable of interest. We cite specific examples of two previous studies of the relationship between the dexamethasone suppression test (DST) and clinical response in depression. Contrary to the conclusions of the investigators, normalization of the DST had no relation to clinical response in the first 5 weeks of treatment. The decrease of post-dexamethasone cortisol levels occurred regardless of clinical response, most likely due to the effect of regression toward the mean.

The development of study exit criteria for evaluating antimanic compounds.

BACKGROUND: There is increasing interest on the part of investigators and the public at large in finding ways to study and improve treatments for the seriously mentally ill without exposing such individuals to unnecessary risks. One group of particular interest in this regard are patients suffering from acute mania. We set out to define "exit" criteria or novel clinical endpoints that might help to assess the efficacy of antimanic compounds. We sought a method that would be safer, more economical, and less sensitive to nonspecific factors in the clinical environment while still allowing unambiguous assessment of efficacy. METHOD: From a pool of subjects being screened for or already participating in intervention studies, we retrospectively identified 76 admissions of patients with a manic or mixed episode according to DSM-IV. We fit a mixed-effects regression model to all available data obtained using the Bech-Rafaelsen Mania Scale from admission to day 28 of treatment. Using the estimated model coefficients, we obtained empirical Bayes (EB) estimates of each subject's trend coefficients based on (1) all available data and (2) data through day 11 of treatment for mania. RESULTS: We found a high correlation (r = .67) between EB estimates of final response at day 28 and actual day 28 scores on the Bech-Rafaelsen scale based on scores through day 11. When subjects were categorized as full, partial, or nonresponders according to their final Bech-Rafaelsen score, we were able to show that only 2 of the 23 predicted nonresponders became full responders, 27 of the 31 predicted full responders became full responders, and 16 of the 22 predicted partial responders became partial or full responders. CONCLUSION: We conclude on the basis of this chart review study that it should be possible to define exit criteria for trials assessing the efficacy of antimanic compounds on the basis of relatively short duration exposure to experimental treatment.

Evaluation of lithium therapy for alcoholism.

The efficacy of lithium therapy for alcoholism was examined in 84 alcoholic volunteers admitted for rehabilitation. Subjects were randomly assigned to a lithium or placebo treatment condition for 18 months, and were reinterviewed monthly to document their compliance, drinking status, and mood. Two separate factors contributed to abstinent outcome: therapeutic serum lithium levels (greater than or equal to 0.4 mEq/L) and medication compliance. Recommendations are proposed for studying compliance effects in future treatment studies.

A statistical model for the classification of imipramine response in depressed inpatients.

We present a statistical model, recently developed for application in mathematical economics, that yields empirical evidence for the existence of two distinct subtypes of depression. We reanalyze previously reported data on 65 depressed patients treated with imipramine and repeatedly rated on the Hamilton Rating Scale (HRS). The estimated model parameters suggest two underlying response processes. Patients in the first subgroup were initially more severely depressed (as measured by the total HRS score), but exhibited a rapid rate of symptomatic response over time. In contrast, patients in the second subgroup were initially less severely depressed, yet showed a much slower rate of improvement. These findings recommend a refinement of the clinical definitions of endogenous depression. While this model suggests that there are two underlying response processes, it does not classify individual patients. Subject classification was made possible by fitting an item-response model to the data. This model relates the 17 individual symptom ratings, at baseline, to the total post-treatment HRS scores. The results of this analysis suggest that the previously described relationship between high initial severity of depression and more rapid improvement over time is characteristic of subjects who exhibit initial motor retardation and decreased sexual interest. Graphical analysis clearly indicates that subjects who exhibit both motor retardation and decreased sexual interest at baseline have higher total HRS scores at baseline and show more pronounced improvement in total HRS scores in response to treatment with imipramine. Patients not exhibiting motor retardation and decreased sexual interest were less severely depressed initially and show virtually no clinical response over time.