Search

Background: This study tested cortisol responses to a psychological stressor in controls (CT) versus patients who were diagnosed as alcohol dependent (AD) or alcohol and stimulant dependent (ADSD) by DSM-IV criteria and who were abstinent for 3 to 4 weeks from alcohol and illicit drugs. Alcohol increases cortisol secretion acutely and during withdrawal. However, there is little information about abnormalities of hypothalamic-pituitary-adrenocortical (HPA) reactivity in recovering alcoholics., Methods: Accordingly, we tested HPA function in the laboratory between 7:00 and 9:30 AM on control versus stress days. Stress consisted of a 20-min public speaking challenge with preparation and delivery of two short speeches, ostensibly evaluated for quality of delivery, whereas control involved relaxing for the same period. Cortisol was measured in saliva collected at baseline, stress or control, and recovery period, and also at home at 9:00 PM on one of the two days., Results: The three groups did not differ in diurnal patterns of cortisol secretion on the rest day and 9:00 PM sample, which indicated that AD and ADSD patients had intact diurnal HPA regulation at rest. During speech stress, the CT subjects showed the expected cortisol increase (p < 0.0001), whereas neither AD nor ADSD patients responded significantly. Cortisol values were not accounted for by covariates such as depression, posttraumatic stress disorder, glucose metabolism, or anthropometric or demographic characteristics., Conclusions: The apparent stress hyporesponsiveness of the AD and ADSD patients suggests a persistent disruption of HPA function, perhaps due to incomplete recovery from prior abuse, or to a preexisting alteration in neural systems that regulate HPA responses to stress.

Background: The relationships between severity of neuropsychological (NP) deficits and quantity and duration of alcoholic drinking remain controversial. Eckardt et al. (1998) proposed that NP deficits can be observed only if chronicity of alcohol abuse equals or exceeds 10 years. In this study we tested the hypothesis of Eckardt et al. and reexamined the relationship of NP performance and alcohol consumption., Methods: One hundred sixty-two alcoholics and 165 controls completed a NP test battery at least 3 weeks after the alcoholics attained sobriety. Chronicity varied from 4 to 9 years for 55 alcoholics and from 10 to 33 years for the remaining 107., Results: Compared to controls, both groups of alcoholics were impaired on the Shipley Vocabulary and Abstraction tests and on two versions of the Digit Symbol test, but there was no difference between the two alcoholic groups on any measure. Regression analyses that controlled for age and education showed that chronicity predicted less than 0.5% of the variance on NP measures. By contrast, a measure of recent alcohol consumption, the Quantity-Frequency Index, contributed significantly (approximately 5% of the variance) to the prediction of alcoholics' NP performance., Conclusions: These data provide weak support for a dose effect relationship between degree of NP impairment and level of alcoholic drinking in the past 6 months but no evidence for an influence of chronicity.

Previous studies have found alcoholics to be impaired on tests of cognitive efficiency. However, it is unclear to what extent individuals who abuse drugs in addition to alcohol exhibit similar deficits. To answer this question, 63 healthy control subjects were compared with 40 individuals who abused alcohol only, 24 individuals who abused alcohol and stimulants, 16 individuals who abused alcohol and marijuana, and 41 individuals who abused alcohol and depressants/narcotics, or alcohol and two or more other drugs. All subjects were administered tests of short-term memory, spatial orientation, visual-spatial perception, and problem-solving. Results from the study indicated that control subjects and individuals who abused both alcohol and marijuana performed significantly better than the other groups on most tests. These results were not attributable to differences on measures of affect or chronicity of alcohol consumption.

Four measures of anger were investigated in sober male and female alcoholics and nonalcoholic peers. The relationships among anger variables, past drinking behavior, and substance abuse consequences in alcoholics were explored. Additionally, the interrelationships among anger, depression, and anxiety in the groups were examined, and the relationships between an overall dysphoria index and drinking behavior and substance abuse consequences were determined. 104 alcoholics (sober 21 to 45 days) and 70 community controls, aged 21 to 56, were given the Spielberger Anger Expression Inventory, the Beck Depression Inventory, and the Spielberger State Anxiety Inventory. Alcoholics scored higher than controls on Trait Anger, Anger-In, and Anger-Out, but not on State Anger. There were no main effects of sex. Anger-In was significantly negatively correlated with the Quantity-Frequency Index in alcoholic males. Anger-In was significantly positively correlated with depression in male and female alcoholics and with substance abuse consequences in the latter group. The depression measure was significantly correlated with consequences in female, but not in male alcoholics. These data have treatment implications, especially for female alcoholics.

The relation between childhood behavior disorders (CBD), attentional difficulties (ADD), and antisocial personality disorder (ASP) was examined in a clinical population of alcoholics and polysubstance abusers. Polysubstance abusers reported increased CBD symptomatology, increased attentional problems, a higher rate of ASP, and a greater number of both current and lifetime ASP symptoms relative to alcoholics without polysubstance abuse. An examination of gender effects revealed significant differences on self-reported adult attentional problems, with females endorsing a greater number of items. As a group, females exhibited the highest rate of ASP. Males showed a significant correlation between current ASP symptoms and both attention and conduct problems in early childhood.

Several tests of visuospatial cognition are known to be sensitive to chronic alcohol abuse, but the consequences of combined abuse of alcohol and other drugs on these measures in not known. To address this issue, groups that had abused only alcohol, alcohol and marijuana, or alcohol and multiple other drugs (Alc/Poly) were compared to community controls. Testing occurred after at least 3 weeks of treatment for the drug abusers. On all measures of visuospatial perception and construction and on all measures of visuospatial learning and memory, all groups of alcoholics were impaired relative to controls, but there were no significant differences among the groups that abused alcohol. By contrast, on all measures of geographical knowledge that required place localization, subjects in the Alc/Poly group were impaired while subjects who abused only alcohol or alcohol and marijuana performed as well as controls. Measures of alcohol consumption, mood or childhood or adult attention deficit were not consistently correlated with test performance.

More than one-half of alcoholics may exhibit deficits in the processing of visual-spatial information, learning and memory, problem-solving and abstracting capabilities, and the regulation of goal-directed behavior. Research using nonalcoholic subjects suggests that some of these deficits may be familial and may predate the development of alcoholism. Results using alcoholic subjects are inconclusive, however, perhaps reflecting the effects of long-term alcohol consumption on cognitive performance. Factors that may influence the interpretation of results include psychiatric disorders, emotional instability, and type of alcoholism. Studies are needed to determine whether children from alcoholic families who become alcoholic themselves prove to be those who were previously most deficient in cognitive abilities.

Neurocognitive functions were compared in alcoholic, schizophrenic, and dually diagnosed (schizophrenic and alcoholic) patients and community control volunteers. Participants (N = 52, n = 13 in each group) completed a brief battery including two measures of neuropsychological function and two affective measures. The neuropsychological tests included the Trail-Making Test Forms A and B and a face-recognition test. The affective measures included assessment of anxiety and depression. The major objective was to compare the patterns of neurocognitive performance among the four groups. The groups did not differ in mean age (35.37) or years of education (12.46). Compared with the community control and alcoholic groups, the schizophrenic groups reported higher mean depression (12.92 vs. 5.54) and mean state anxiety scores (60.73 vs. 46.04). Control subjects were generally, although not always significantly, superior to the other groups. Contrary to expectations, alcoholic patients were not consistently better than the schizophrenic groups, and dually diagnosed participants were not more impaired than schizophrenic patients without substance abuse histories. Thus, the interaction of schizophrenia and substance abuse did not have an additive effect on neurocognitive performance assessed in this study. More research using this complete four-group design and a larger neurocognitive battery is needed.

Cloninger's Type 1/Type 2 classification scheme was examined in 82 male alcoholics. Only 16% were classified by presence/absence of Type 1/Type 2 symptoms. When age of symptom onset was considered, 75.6% were classified into early (ESO) and late symptom onset (LSO) groups. Relatively few group differences were found: ESOs were younger, had lower SES, began drinking earlier, reported greater frequency of maternal alcoholism and more severe paternal alcoholism, and reported more antisocial behaviors. ESO/LSOs did not differ on neuropsychological performance, although both performed poorer then controls. Differences between these findings and those found with female ESO/LSO alcoholics are discussed and clinical implications considered.

Objective: To analyze visuospatial cognition in recently detoxified alcoholics from the perspectives of three ways of conceptualizing spatial information processing: egocentric versus allocentric orientation, featural versus configural analysis, and categorical versus coordinate spatial judgements., Method: Twenty-eight chronic alcoholics (19 men, 9 women) were compared to 20 (10 men, 10 women) controls of comparable age and education on a battery of tests of visuospatial scanning, construction, mental imagery, and anterograde and remote spatial memory. Tests were administered 21-40 days after alcoholics entered treatment., Results: Alcoholics displayed impairment in visuospatial scanning, construction, utilizing and manipulating information from visual images and on three tests of anterograde spatial memory, but remote spatial memory was not significantly affected. Their deficits were evident on some measures of allocentric orientation, featural and configural analysis, but consistent deficits on egocentric orientation or categorical or coordinate spatial judgments were not seen., Conclusions: Deficits in spatial cognition exhibited by alcoholics do not seem to arise from dysfunction in any localized brain region. Small but potentially important impairments in fundamental aspects of spatial information processing such as scanning and use of visual imagery were found. The empirical basis and clinical significance of these deficits requires further study.

Using a component processes model, the current study examined the cognitive processes of alcoholic and community control subjects engaged in the acquisition of new context-bound (e.g., episodic) information. Of particular interest was whether alcoholics were inferior to controls in the efficiency with which information was acquired and whether there was a gender x group interaction in cognitive efficiency. Alcoholic (n = 16 females; n = 22 males) and community control (n = 21 females; n = 21 males) subjects participated in a serial learning task that consisted of three 12-item word lists. Standard administration protocols for serial learning tasks were used. On efficiency measures, there were significant differences between groups [F(1.75) = 8.51, p = 0.005] and sexes [F(1.75) = 4.05, p = 0.048]. There was also a group x sex interaction [F(1.75) = 7.73, p = 0.007]. Duncan multiple-range comparisons revealed alcoholic females to be significantly inferior to control females, but equivalent to male controls and alcoholics. These data are consistent with other studies revealing the sensitivity of cognitive efficiency to alcohol-related effects and extend previous findings to tasks involving episodic learning tasks.

This study was conducted to examine the role of childhood behavioral disorders (CBDs) and residual attention deficit disorder (ADDRT) in alcohol-related cognitive dysfunction in male and female subjects. Alcoholic (n = 44 females, 56 males) and control (n = 40 females, 40 males) subjects completed assessments that included measures of CBDs, ADDRT, and cognitive and psychosocial functioning. Cognitive tests were specifically designed to assess efficiency in function. As expected, alcoholics were inferior to controls in their cognitive efficiency [F(1,171) = 10.43, p = 0.0015]. Alcoholics reported more CBDs [F(1,161) = 12.02, p = 0.0007], regardless of sex. They also reported more ADDRT [F(1,173) = 44.12, p = 0.0001] than did controls. There were also sex [F(1,173) = 13.31, p = 0.0004] and group by sex effects [F(1,173) = 3.93, p = 0.05]. Female alcoholics reported more ADDRT symptoms than any other group. Regression equations conducted to clarify the relation between group, sex, CBDs, ADDRT, and cognitive efficiency indicated that the best predictor of cognitive efficiency was group classification (alcoholic versus control). That is, although symptoms of behavioral disorders were reported significantly more frequently by both male and female alcoholic subjects, these symptoms could not account for the cognitive impairment observed in either sex.

The mild generalized dysfunction hypothesis of alcohol abuse's deleterious effects on cognitive processes has gained support from a number of studies in which detoxified alcoholics have a lower mean performance level than peer controls on a variety of neuropsychological tests. This approach might obscure consistent but different patterns of preserved and impaired cognitive performance among subgroups of alcoholics, suggestive of alternative hypotheses. To test this possibility, neuropsychological test data from two large, independent samples of alcoholics (sample 1, n = 143; sample 2, n = 130) and controls (sample 1, n = 97; sample 2, n = 83) were subjected to separate centroid hierarchical cluster analyses. For both samples, the majority of alcoholics (94% and 94%) exhibited a pattern of impaired verbal and nonverbal performance and deficits in memory and perceptual motor skill, with normal motor skill. The alcoholics who did not fit this pattern showed more severe or wide-ranging impairments. These findings indicate that empirical support for the mild generalized dysfunction hypothesis of alcoholics' cognitive deficits is not an artifact of averaging.

To examine possible influences of premorbid and comorbid factors on the neuropsychological test performance of recently abstinent (3-5 weeks) drug abusers, we studied 24 alcoholics, 23 cocaine abusers, and 22 healthy controls of comparable age and education. Both alcoholics and cocaine abusers performed significantly more poorly than controls on most measures of learning and memory, problem solving and abstraction and perceptual-motor speed, but the groups did not differ on the measure of sustained attention. Correlational analyses revealed no significant relationships between measures of childhood and residual hyperactivity and neuropsychological performance; scores on the Beck Depression Inventory were related only to performance on the Wisconsin Card Sorting Test. The findings indicate that abuse of cocaine or alcohol is associated with deficits on neuropsychological tests which cannot be attributed to specific premorbid or comorbid factors such as depression or childhood or residual attention deficit disorder.

Female alcoholics experience greater affective symptomatology and are more frequently labeled with comorbid psychiatric diagnoses than male alcoholics. Subgroups of male and female alcoholics with antisocial symptomatology are strikingly similar across behavioral, psychosocial, and alcohol and drug use variables. Relative to controls, female alcoholics manifest deficits in standard neuropsychological testing as well as efficiency of information processing, interpersonal problem solving, neurophysiological measures, and neurophysiological assessment. The deficits are quite similar to those of male alcoholics. However, females may manifest a differential susceptibility to alcohol given the females' comparatively shorter drinking careers. There are no overall sex differences in relapse rates following treatment, although various personal and environmental factors play gender-specific roles in outcome. Patterns of recovery function with abstinence are similar for male and female alcoholics.

In an attempt to clarify the nature of the problem-solving deficits exhibited by chronic alcoholics, the California Card Sorting Test (CCST) and other measures of abstraction and problem solving were administered to 23 alcoholics and 16 nonalcoholic controls, equated for age, education and vocabulary. On the CCST, the alcoholics exhibited three types of deficits which appeared to be relatively independent. First, the alcoholics generated and identified fewer correct concepts than controls, although they executed concepts normally when cued by the examiner. Second, the alcoholics made more perseverative sorting responses and perseverative verbal explanations for their sorting behavior than did controls. Third, alcoholics provided less complete verbal explanations of the concepts that they correctly generated or identified. The differential importance of these factors on various measures of problem solving may help to explain the varied patterns of inefficient problem solving exhibited by alcoholics.

Visual-spatial learning in young adult nonalcoholic men and women with (FH+) and those without (FH-) a family history of alcoholism was investigated using nonsense shapes of high and low verbal association value. Four groups of FH+ and FH- men and women, 16 subjects in each group, were tested. The male FH+ group required significantly more trials to reach learning criterion and made more errors than the male FH- group; the FH effect was not significant in females although similar trends were present. Women exhibited significantly poorer visual-spatial learning compared to the men. High compared to low verbal association shapes were learned in fewer trials with fewer errors by all groups. There were no significant FH by verbal association interactions, indicating that the impaired visual-spatial learning in the FH+ men could not be ascribed to the meaningfulness dimension of the nonsense shapes. These results suggest that the impaired visual-spatial learning found in adult children of alcoholics (ACOAs) might be due in part to premorbid cognitive deficits.

Chronic alcohol abuse is associated with a wide range of cognitive deficits, including impairment in perceptual-motor skills, visual-spatial functions, learning/memory, and abstraction and problem solving. In Korsakoff patients, despite normal IQs, these deficits are pronounced, especially in episodic memory and new learning. Procedural memory (learning of simple, repetitive skills), however, is relatively intact. New evidence suggests that orbitobasal frontal lobe dysfunction in addition to diencephalic lesions contributes to the deficits found in Korsakoff patients. A genetically controlled thiamine deficiency has been hypothesized to account for the development of the Wernicke-Korsakoff disorder. Prognosis for the recovery from Korsakoff's syndrome is poor, although evidence suggests that these patients should be able to learn simple repetitive tasks involving procedural memory. In non-Korsakoff sober alcoholics the aforementioned cognitive deficits are less severe but nonetheless significant. These performance deficits are not easily accounted for by the current neuropsychologic hypotheses, although neuroanatomic changes frequently accompany chronic alcohol abuse. Family history, although a significant predictor of later alcoholism, does not appear to play a major role in determining alcohol-related cognitive deficits. A history of childhood behavioral disorders is associated with greater cognitive impairment. The means by which its effect is expressed, however, is unclear. The most parsimonious explanation focuses on the possibility that these disorders are the manifestation of inherited temperament traits that increase vulnerability to a variety of negative outcomes including antisocial personality and alcoholism. Significant recovery of cognitive function can occur but may be contingent on factors such as age and continued abstinence.

Chronic alcoholics demonstrate cognitive deficits when compared with nonalcoholics. These deficits are typically attributed to the direct effects of ethanol and its metabolites on the central nervous system (CNS). There are other factors, however, that differentiate alcoholics from controls, such as personality or behavioral characteristics. These factors may affect neuropsychological performance and thus alter the interpretation of alcoholic cognitive deficits as resulting solely from alcohol's toxic effects. To investigate this question, male and female alcoholics and peer nonalcoholic controls were compared on personality, behavioral, and cognitive measures. Alcoholics had greater numbers of antisocial behaviors, childhood behavioral disorder symptoms (CBD), and affective symptomatology, and had poorer neuropsychological performance than controls. The three personality and behavioral factors were positively intercorrelated with each other, and were negatively related to cognitive performance. The CBD factor proved to be the most consistent predictor of neuropsychological performance for both alcoholics and controls, and males and females. While the behavioral factors differentiated alcoholics from controls and predicted performance, significant differences between the groups in cognitive performance still remained when these factors were taken into account.

Historically, there has been considerable interest in developing alcoholic typologies with the belief that typologies would facilitate treatment development and stimulate meaningful research. Most of the existing typologies can be classified into one of two broad camps: the behavioral typologies and the family history/genetic typologies. Recent work has focused on the integration of these two strategies into comprehensive typological systems. This chapter provides an overview of the major typologies, noting their various strengths and weaknesses. Common weaknesses among existing typologies are that female alcoholics have not been adequately studied and/or are stereotyped. Behavioral heterogeneity within the female alcoholic population is discussed and suggestions for typological development are noted.

This study was conducted to determine whether alcoholic and control subjects respond differently to manipulations that either enhance personal involvement (PI) or reduce negative affect (R, relaxation) on tests of neuropsychological function. In Phase 1, 48 male alcoholics and 36 male control subjects completed neuropsychological tasks under standard instructional sets. In Phase 2, subjects completed equivalent forms of these tests under one of three randomly assigned conditions; the PI condition in which subjects were encouraged to identify specific ways of improving their performance, the R condition in which subjects participated in a short relaxation exercise designed to reduce anxiety, or a No Treatment (NT) condition in which no attempt to manipulate the subjects' involvement or affect was made. Alcoholics were inferior to controls in both Phase 1 and Phase 2 [Fs (1,82) > 5.03, ps < 0.03]. The experimental manipulation differentially affected measures of negative affect and effort in the predicted direction. There were no group x condition interactions. Alcoholic and control subjects responded comparably to the experimental manipulations. This investigation, in combination with others using related manipulations, reinforces the hypothesis that alcohol-related cognitives dysfunction reflects an underlying deficit in brain states.

This study compared the performance of alcoholics (18 male, 16 female) with community controls (15 male, 12 female) on the Adaptive Skills Battery (ASB), a test of interpersonal problem-solving. The ASB consists of 30 situations. Fifteen situations require subjects' typical responses and 15, their optimal responses. Male and female alcoholics were inferior to controls in their typical responses [F(1,57) = 45.22, p = 0.0001], but did not differ on the optimal responses (F less than 1). Further analyses indicated that decreased feelings of self-efficacy could not account for the alcoholic deficit. Females were superior to males in the optimal response condition [(F(1,56) = 9.90, p = 0.003]. No significant group x sex interactions were obtained. Performance on the ASB was not correlated with performance on traditional measures of problem-solving. Post-hoc correlational analyses revealed differential patterns between depression scores and performance for alcoholics and controls. These data suggest that (1) female and male alcoholics exhibit similar interpersonal problem-solving deficits, (2) alcoholic self-efficacy expectancies cannot account for the impairment, (3) the ASB appears to assess aspects of problem-solving not typically examined, and (4) the role of depression in alcoholic performance deserves continued empirical evaluation.

This study examined the incidence of Type 1/Type 2 symptoms and age of alcoholism onset in a population of alcoholic women in order to determine whether there were females similar to the Type 1 and Type 2 male alcoholics described by Cloninger and colleagues. Using a sample of 51 alcoholics divided into early and late alcoholism onset groups (less than 25 and greater than or equal to 25), the early onset group did not manifest merely Type 2 symptomatology nor did the late onset group manifest Type 1 symptoms without the occurrence of Type 2 symptoms, as would have been predicted by the Cloninger theory. Rather, the majority of women endorsed both types of symptoms regardless of their age of alcoholism onset, making classification by sympton type unsuccessful. However, combining symptoms across types (making a total of 7) produced 71% classification of subjects into early symptom onset (ESO) and late symptom onset (LSO) groups. Analyses indicated differences between the groups in each of five areas: (1) demographic and psychosocial variables, (2) alcohol abuse severity, (3) family history, (4) psychopathological symptoms, and (5) substance abuse history. In each case the ESO alcoholics resembled the characteristics of male Type 2 alcoholics reported in other studies. The findings indicate that modification of the Cloninger classification criteria to reflect age of symptom onset regardless of symptom "type" provides better classification of individual female alcoholics and suggests that female alcoholics are represented in both Type 1 and Type 2 alcoholic subtypes.

Verbal fluency deficits in female alcoholics (n = 48) were investigated using Thurstone's Word Fluency Test (TWFT) and a category fluency test. Overall, alcoholics performed less well than did controls (n = 36). Simple effects indicate that although controls performed less well on category fluency than on the TWFT, alcoholics did not differ between the two tests. Simple effects also show that controls' performance was superior to that of alcoholics on the category fluency test, but the groups did not differ on the TWFT. The pattern of results suggests that the process of shifting rapidly between categories is more difficult than perusing a mental lexicon. Correlations between these tasks and other tests of cognitive functioning suggest that the TWFT and the category fluency test may be assessing different types of verbal fluency.

This study investigated whether male and female alcoholics and controls differed in their ability to solve an ecologically valid measure of abstraction. Ninety-one alcoholics (39 females, 52 males) and 61 community controls (36 females, 26 males) completed a battery of abstraction tasks including a Piagetian-type task, the Plant Task. The Plant Task demands that subjects predict the outcome of an unseen plant given a specific treatment regime. No group differences in the ability to correctly predict the unseen plant's outcome were obtained. Alcoholics were significantly poorer than controls in isolating the relevant variable (chi 2 (1) = 8.24, p = 0.004). There were no sex differences in the ability to isolate the relevant variable. However, there was a sex effect in the number of irrelevant explanations (chi 2 (3) = 15.58, p = 0.001) with females producing more than males and female alcoholics more than any of the other three cells. These data suggest that (1) alcoholics may exhibit subtle abstraction deficits most readily observed in tasks examining process as opposed to final performance, (2) males and females differ in their response patterns regarding irrelevant material, and (3) further empirical study of the Plant Task may support its use in experimental and clinical settings.

To assess its construct validity and application, Cloninger's Tridimensional Personality Questionnaire (TPQ) was administered to 267 inpatient substance abusers. One hundred seventy-two of the patients were alcoholic, 47 abused stimulants, and 48 abused other drugs, primarily benzodiazepines, marijuana, or were polysubstance abusers. Of the sample 33.8% were female. Analyses were conducted to (1) determine whether TPQ scores could be used to classify alcoholics as either Cloninger Type I or Type II; (2) examine the intercorrelations of TPQ scales and identify differences in these patterns due to drug of choice; (3) replicate previously obtained gender differences on the TPQ. The results indicate that (1) TPQ scores did not classify the alcoholics into the expected frequencies for Type I and Type II; (2) the scales were not independent and did not give rise to any differences among drug user groups, but (3) did confirm previously reported gender differences. These results indicate that Cloninger's model and the TPQ must be subjected to careful empirical study prior to their widespread acceptance.

In two separate studies, episodic semantic (verbal) and figural-memory abilities as measured by the Wechsler Memory Scale (Russell Method) are compared in middle-aged detoxified male alcoholics and controls. Both studies gave rise to similar findings. Alcoholics recalled significantly less verbal and figural material at both immediate recall and 30-min delayed recall than did the controls. Contrary to expectation, compared to controls, alcoholics did not have a greater differential deficit for the figural material as compared with the verbal or a greater differential deficit on delayed than on immediate memory. The percentage of material recalled at the delayed testing, when taken as a function of the amount of material recalled immediately, did not differ for the two groups on the semantic or figural tasks in either study. Thus, rates of forgetting were equivalent in alcoholics and controls suggesting that the alcoholics' deficit lies in the initial acquisition process. Whether that deficit is in encoding or initial retrieval processes remains to be investigated. Implications for future research are given.

The effects of subject selection bias in research on cognitive deficits in sober alcoholics were studied in a sample of 523 subjects (98 controls, 276 ineligible alcoholics, 40 eligible alcoholics who declined to participate, and 144 eligible and participatory alcoholics). All subjects received the Shipley Institute of Living Scale (Vocabulary and Abstracting subscales) and measures of anxiety, depressive symptoms, childhood hyperkinesis and attentional deficit disorders were obtained. Results indicate that current guidelines for alcoholic subject selection are biasing analyses toward support of the null hypothesis. Declined alcoholics performed more poorly on the Shipley Abstracting than did the "Used" group, yet did not differ significantly from the Used alcoholic groups on depression or anxiety. Declined alcoholics did, however, report significantly fewer Hk/MBD symptoms than did other alcoholic groups. ANCOVAs that used the affective and childhood disorders as covariates did not alter the differences in cognitive performance described above.

Alcoholics' neurological deficits have been attributed to cognitive impairment associated with brain dysfunction. Because alcoholics perceive themselves as impaired in higher cognitive functions, we investigated the alternative hypothesis that their deficits are predicted by lowered expectancies for successful performance. Ss were 48 male, middle-aged Veterans Administration inpatient alcoholics and 36 community controls. Subjects completed a standard series of neuropsychological tests. Before taking each test, Ss were asked how well they expected to do on that test in relation to members of their community. As predicted, means for the overall measures of expectancies and performance were significantly lower for alcoholics than for control Ss. Although expectancies were significantly correlated with performance, regression analyses indicated they could not account for the obtained performance differences.