The role of drinking restraint in alcohol dependence: validation of the temptation and restraint inventory in an alcohol dependent sample

OBJECTIVE: The Temptation and Restraint Inventory (TRI) is commonly used to measure drinking restraint in relation to problem drinking behavior. However, as yet the TRI has not been validated in a clinical group with alcohol dependence. METHOD: Male (n = 111) and female (n = 57) inpatients with DSM-IV diagnosed alcohol dependence completed the TRI and measures of problem drinking severity, including the Alcohol Dependence Scale and the quantity, frequency and week total of alcohol consumed. RESULTS: The factor structure of the TRI was replicated in the alcohol dependent sample. Cognitive Emotional Pre-occupation (CEP), one of the two higher order factors of the TRI, demonstrated sound predictive power toward all dependence severity indices. The other higher order factor, Cognitive Behavioral Control (CBC), was related to frequency of drinking. There was limited support for the CEP/CBC interactional model of drinking restraint. CONCLUSIONS: Although the construct validity of the TRI was sound, the measure appears more useful in understanding the development, maintenance and severity of alcohol-related problems in nondependent drinkers. The TRI may show promise in detecting either continuous drinking or heavy episodic type dependent drinkers.     

Drinking restraint versus alcohol expectancies: which is the better indicator of alcohol problems?

OBJECTIVE: The relationship between expectancies and other psychological constructs related to drinking is unclear. The current study assesses the power of drinking restraint, measured by the Temptation and Restraint Inventory (TRI), along with alcohol expectancy and drinking refusal self-efficacy, measured by the Drinking Expectancy Profile (DEP), as indicators of alcohol use and problem drinking. METHOD: Volunteer students (N = 359), consisting of 113 men with a mean (+/-SD) age of 25.31 +/- 10.61 years and 246 women with a mean age of 23.04 +/- 8.90, completed the TRI, DEP and Alcohol Dependence Scale (ADS) as well as frequency and quantity measures of self-reported drinking. RESULTS: Drinking Restraint was a stronger indicator of higher scores on the ADS, accounting for 54% and 45% of the variance for men and women, respectively, with Alcohol Expectancies and Drinking Refusal Self-Efficacy being the better indicator of frequency of alcohol consumption, accounting for 20% of the variance for men and 26% for women. Both measures were represented by similar variances within the quantity of drinking measure. CONCLUSIONS: Drinking restraint and alcohol expectancies were seen to measure kindred but unique cognitive subsets, providing further insight into the progression of alcohol problems. Alcohol expectancies and drinking refusal self-efficacy may be acquired early in the development of drinking behavior, as evidenced by stronger associations with risky drinking; drinking restraint and its associated loss of control factors appear to be more specifically related to problem drinking. Alcohol expectancy and drinking refusal self-efficacy may thus have broader use in the assessment of drinking behavior.     

Gender comparison of alcohol exposure on drinking occasions

OBJECTIVE: This study was designed to assess the influence of gender, age, drinking style and other selected demographic variables on the rate of alcohol beverage consumption on drinking days. The data were used to predict possible differences in ethanol exposure levels (peak blood alcohol concentration [BAC]) among the subgroups examined. METHOD: A representative sample (N = 2,627) of the U.S. adult population was surveyed using computer-assisted telephone interviewing to assess alcohol use (past 12 months) in quantity, frequency and type of beverage, as well as the time (minutes) usually required to consume the typical quantity of alcohol on drinking days. RESULTS: Of the 1,833 current drinkers (1,028 men, 805 women), women reported consuming a mean of 2.2 standard drinks (1 drink = 12 g ethanol) on typical drinking occasions (days); men reported consuming a mean of 3.2. The duration of the drinking episode was similar for women (122 minutes) and men (126 minutes). The hourly rate of drink consumption was thus lower for women (approximately 1.1 drinks/hour) than for men (approximately 1.6 drinks/hour). For both men and women, there was an age-related increase in the frequency of drinking but a decrease with age in predicted peak BACs and in the alcohol intake per drinking episode. Peak BACs achieved during typical drinking episodes were estimated to be quite similar for men (0.037) and for women (0.036) when prediction equations were based on equal rates of alcohol elimination in both genders. The estimated peak BACs were lower in women than in men when the prediction equation assumed a higher rate of ethanol elimination in women. CONCLUSIONS: Information on the rate of alcohol beverage consumption on drinking days facilitates prediction of BACs typically experienced in men and women of different demographic group memberships. The exposure values obtained will depend upon as yet unsettled norms for ethanol elimination rates in men and women.     

The relationship between alcohol expectancies and drinking restraint in treatment seeking alcohol dependent patients

The allied psychological drinking constructs of alcohol expectancy and drinking restraint have shown considerable promise, both independently and in combination, in understanding problematic drinking. However, previous research examining the relationship between these two constructs has only occurred in pre-dependent, convenience samples. This study examined the role of both alcohol expectancies and drinking restraint in an alcohol dependent sample. 143 DSM-IV alcohol dependent participants (93 males, 50 females) completed measures of drinking restraint (Temptation and Restraint Inventory, TRI) and alcohol expectancy (Drinking Expectancy Profile, DEP), along with quantity and frequency of consumption and dependence severity (Alcohol Dependence Scale, ADS). The results showed that although alcohol expectancy and drinking restraint do share common underlying properties, there was unique variance attributed to the prediction of dependence severity and consumption. The results also failed to replicate the drinking restraint model observed in non-clinical samples. It was concluded that alcohol expectancy and drinking restraint models appear to show superior utility in pre-dependent populations. The implications of these findings in refining social cognitive models of alcohol misuse are discussed, with a particular focus on prevention.     

Drinking patterns of older Americans: National Health Interview Surveys, 1997-2001

OBJECTIVE: The objective of this study was to estimate the quantity and frequency of alcohol consumption among current drinking Americans age 60 years and older, by gender and age. METHOD: Five years (1997-2001) of cross-sectional National Health Interview Survey data were pooled. Quantity (number of drinks consumed, on average, on drinking days), frequency (number of drinking days per year) and a composite quantity-frequency measure (average number of drinks per day) were defined. Age trends (between age 60 and 84 years) were tested using logistic regression. Analyses were weighted to produce national estimates. RESULTS: Among 40,556 adults age 60 years and older, 52.8% (n = 8,136) of men and 37.2% (n = 8,710) of women were current drinkers. Over increasingly older age groups of current drinking men and women (1) proportions consuming higher quantities of alcohol (two drinks or more) decreased (p trend < .001), whereas proportions consuming lower quantities (one drink) increased (p trend < .001); and (2) proportions drinking least frequently (< 12 days per year) and most frequently (260-365 days per year) increased (p trend < .05), whereas proportions drinking at intermediate frequencies remained stable or decreased. Differences in patterns for quantity alone and frequency alone were obscured by the composite quantity-frequency measure. CONCLUSIONS: In a nationally representative survey, quantity and frequency of alcohol consumption showed strikingly different patterns of cross-sectional age-related change. In epidemiologic studies, quantity and frequency have been differentially associated with medical conditions prevalent in older populations. Investigators studying alcohol consumption in older people should consider reporting results separately by quantity and frequency.     

Epidemiological trends in drinking by age and gender: providing normative feedback to adults

OBJECTIVE: The purpose of this research was to evaluate drinking rates as a function of age and gender and to disseminate current estimates of U.S. population drinking norms based on age and gender. METHODS: Participants included 42,706 men and women 18 years and older who provided information about their drinking from the National Epidemiologic Survey on Alcohol and Related Conditions [National Alcohol Survey on Alcohol and Related Conditions (NESARC, 2001) dataset collected between 2001 and 2002 from a representative, non-institutionalized sample. RESULTS: Results revealed greater frequency and typical quantity of alcohol consumption among men versus women. Age differences in drinking frequency suggests a sharp increase with legal drinking age followed by a period of reduced frequency, in turn followed by gradual increase up to retirement age. Age differences in typical drinking quantity suggest a sharp increase with legal drinking age followed by a gradual linear decline in number of drinks per occasion. Age differences in typical quantity were more pronounced among men. CONCLUSIONS: Analyses provide epidemiological trends in drinking rates by age and gender, and emphasize the importance of within group differences when examining drinking rates. Discussion focuses on explaining how to incorporate norms information in prevention and treatment.     

Gender and the assessment of at-risk drinking: evidence from the GENACIS Canada (2004-2005) telephone survey version of the AUDIT

The alcohol use disorders identification test (AUDIT) is widely used in general population surveys as a method of determining prevalence of hazardous drinking. However, its interpretation has been questioned particularly regarding the unequal contribution of the items to the total score, specifically, that the drinking frequency item contributes disproportionately to the score and may lead to inappropriate identification of some drinkers as hazardous drinkers. To explore these issues further as well as possible gender differences in the applicability of the AUDIT, we conducted analyses using a modified version of the AUDIT (AUDIT(M)) as part of a general population survey that used random digit dialing and computer-assisted telephone interviewing. Item and factor analyses were performed separately for men and women, and the impacts of excluding the frequency of drinking item in the measurement of mean scores, percentages and types of problems for men and women were examined. We found that the AUDIT(M) items loaded onto three distinct dimensions for both men and women: frequency of drinking; usual quantity and frequency of heavy-episodic drinking; problem consequences from drinking. In addition, we found that excluding the frequency question may give a more meaningful estimate of the percent of drinkers actually at risk of experiencing problems from drinking for both men and women. Finally, although our analyses identified only minor gender differences in the structure of the AUDIT and good sensitivity for identifying problem drinkers among both men and women, significant gender differences in the types of problems experienced suggest that use and interpretation of the AUDIT should routinely take gender into consideration.     

Temptation, restriction, and the regulation of alcohol intake: validity and utility of the temptation and restraint inventory

OBJECTIVE: The Temptation and Restraint Inventory (TRI) is a measure of drinking restraint, the preoccupation with controlling alcohol intake. It contains two higher order factors pertaining to the temptation to drink and the restriction of alcohol intake. We tested the convergent and discriminant validity of the TRI in comparison to the Preoccupation with Alcohol Scale (PAS), the Reasons for Limiting Drinking Scale (RLD) and the Drinking-Related Locus of Control Scale (DRIE), each of which measures constructs analogous to the components of the TRI. We also tested the utility of the TRI, as compared to these related measures, for predicting typical drinking and alcohol problems. METHOD: A sample of 296 (50% male) moderate to heavy drinking adults completed the TRI, the PAS, the RLD, the DRIE and measures of typical (weekly) drinking and alcohol problems (SMAST). RESULTS: Correlations indicated that the TRI had excellent convergent and discriminant validity. The TRI's temptation factor (Cognitive and Emotional Preoccupation) was strongly and positively correlated with the PAS and the DRIE, and was weakly and positively correlated with the RLD. The TRI's restriction factor (Cognitive and Behavioral Control) was positively and moderately correlated with the DRIE and moderately to weakly correlated with the PAS and the RLD. In regressions predicting weekly drinking and SMAST scores, the TRI was particularly effective in accounting for the variance in SMAST scores. CONCLUSIONS: In comparison to measures of conceptually related constructs, the components of the TRI showed excellent convergent and discriminant validity. The TRI accounted for large portions of the variance in alcohol-related outcomes, using fewer items than measures of similar constructs.     

Using the Bivariate Dale Model to jointly estimate predictors of frequency and quantity of alcohol use

OBJECTIVE: This study demonstrates the usefulness of the Bivariate Dale Model (BDM) as a method for estimating the relationship between risk factors and the quantity and frequency of alcohol use, as well as the degree of association between these highly correlated drinking measures. METHOD: The BDM is used to evaluate childhood sexual abuse, along with age and gender, as risk factors for the quantity and frequency of beer consumption in a sample of driving-while-intoxicated (DWI) offenders (N = 1,964; 1,612 men). The BDM allows one to estimate the relative odds of drinking up to each level of ordinal-scaled quantity and frequency of alcohol use, as well as model the degree of association between quantity and frequency of alcohol consumption as a function of covariates. RESULTS: Individuals who experienced childhood sexual abuse have increased risks of higher quantity and frequency of beer consumption. History of childhood sexual abuse has a greater effect on women, causing them to drink higher quantities of beer per drinking occasion. CONCLUSIONS: The BDM is a useful method for evaluating predictors of the quantity-frequency of alcohol consumption. SAS macrocode for fitting the BDM model is provided.     

Alcohol problem recognition as a function of own and others' perceived drinking

A self-regulation model was applied for predicting alcohol problem recognition. With the rate of others' alcohol use serving as a standard, problem recognition was predicted to increase the more one perceived one's drinking rate to be above others' use. Within the context of a classroom-administered survey, 707 nonabstaining undergraduates reported their drinking rates and estimated other students' drinking rates for annual drinking frequency, average weekly drinking quantity, and recent heavy drinking frequency. The independent roles of one's own and others' drinking rates, as well as the discrepancy between these two rates in predicting problem recognition, were examined. Findings were consistent with a self-regulation account. Across measures, only own drinking rate independently predicted problem recognition. Others' drinking rate interacted with own drinking rate in predicting problem recognition on the quantity measure. Specifically, the more heavy drinkers perceived their own quantity to be above others', the higher their problem recognition. Finally, gender effects were observed. Although men reported higher problem recognition than women, women reflected more on their drinking rates for problem recognition.     

Drinking patterns and beverage preferences of liver cirrhosis patients in Mexico

The purpose of this study was to investigate the pattern of alcoholism in a special group of alcoholics (alcoholic cirrhotics) in a hospital-based population in west central Mexico and assess the role of regional spirits such as tequila. A complete alcohol drinking history and a structured questionnaire directed at investigating the pattern of alcohol consumption was applied to 124 adult patients with chronic liver disease caused by alcohol during January 1995 to January 1996. The mean age of onset was 27 +/- 3 years in women and 18 +/- 0.5 years in men. The mean alcohol intake per week was 749 +/- 192 g for women and 1113 +/- 151 g for men. On average, patients consumed alcohol for a mean of 24.5 years. The overall patient drinking preference was for tequila followed by 96 degree Gay Lusac (G.L.), alcohol, and beer. In a subset of 70 patients three phases of alcoholism could be identified (prealcoholic, critical, and chronic). Each phase had a mean duration of at least 11 years. Beer was the dominant beverage in the prealcoholic phase while tequila was consumed more often in the other phases. In the critical phase of alcoholism an average of 337 g of alcohol were consumed per week and in the chronic phase 1765 g/week. Tequila was the overall preferred beverage in this group of alcoholics. Other beverages included beer and straight alcohol with a clear trend from less to higher concentration of alcohol throughout the drinking history. Subtle gender differences in the patterns of alcoholism may be suspected. In this group of patients the role of tequila drinking is highlighted.     

Alcohol consumption and quitting smoking in the International Tobacco Control (ITC) Four Country Survey

Although greater alcohol consumption has been associated with decreased odds of quitting smoking in prospective studies, the aspects of drinking most strongly associated with quitting have not been fully explored and examination of potential confounder variables has been limited. Further studies are needed to inform efforts to enhance smoking cessation among the substantial portion of smokers who drink alcohol. The present study examines: (a) drinking frequency, average weekly quantity of alcohol consumption, and frequency of heavy drinking as prospective predictors of quit smoking behaviors, (b) difference across countries in this prediction, and (c) third variables that might account for the association between alcohol consumption and quitting smoking. Data were drawn from the International Tobacco Control Four Country Survey, a prospective cohort study of smokers in Australia, Canada, the UK, and the US. A total of 4831 participants provided alcohol data at one study wave and were re-interviewed 1 year later. Individuals who drank heavily (4+/5+ drinks for women and men, respectively) more than once a week had significantly lower rates of quitting smoking than all other participants, in part due to the fact that a significantly lower proportion of those making a quit attempt remained quit for more than 1 month at follow-up. The role of frequent heavy drinking did not differ by country or sex and was not accounted for by demographics, smoking dependence, or attitudes regarding quitting smoking. Neither drinking frequency nor weekly quantity of consumption showed robust associations with quitting behaviors. Results indicate further study of interventions to address heavy drinking among smokers is warranted.     

The role of drinking in suicidal ideation: analyses of Project MATCH data

OBJECTIVE: The associations of suicidal ideation with both the intensity (drinks per drinking day) and frequency of alcohol consumption were examined in a longitudinal study of treated alcoholics. METHOD: The data are from alcohol dependent subjects (1,187 men and 374 women) enrolled in Project MATCH, a multisite clinical trial of psychosocial treatments of alcoholism. Multivariate analyses for correlated data using generalized estimating equation approaches were performed to examine correlates of suicidal ideation at study entry and at 3-, 9- and 15-month follow-up. Analyses were stratified by gender and controlled statistically for depression and alcoholism severity. RESULTS: Suicidal ideation was common among women (15.5%) and men (9.9%) entering treatment, and at least 3.6% of women and 4.2% of men reported suicidal ideation at each follow-up. There were gender patterns in suicidal ideation. In women, intensity was associated with suicidal ideation, but even nonintense drinking became associated with suicidal ideation with more frequent drinking (intensity-frequency interaction). In men, intensity was also associated with suicidal ideation, whereas frequency was unrelated. Antisocial personality disorder in men but not in women, and depression in both groups, were also linked to suicidal ideation. CONCLUSIONS: Suicidal ideation is prevalent among treated alcoholics. Drinking is strongly associated with suicidal ideation and in women even light drinking, if it occurs regularly, is associated with suicidal ideation, with implications for suicide risk-recognition and intervention.     

The inheritance of alcohol consumption patterns in a general population twin sample: II. Determinants of consumption frequency and quantity consumed

Genetic models were fitted to self-report data on frequency of alcohol consumption and average quantity consumed when drinking, from 3,810 adult Australian twin pairs. Frequency of consumption is determined both by an abstinence dimension, which is strongly influenced by shared environmental effects but not by genetic effects, and by an independent frequency dimension, which is influenced by genetic effects in both sexes and possibly by shared environmental affects in men. Quantity of alcohol consumed is likewise determined by an environmental abstinence dimension and by an independent and partly heritable quantity dimension. The best-fitting model allowed for two routes to abstinence: those who were not abstainers by virtue of their position on the abstinence dimension could nonetheless become abstainers by their position on the second, frequency (or quantity) dimension. Heritability estimates were 66% in women and 42-75% in men, for frequency; and 57% in women and 24-61% in men, for quantity.     

Temptation, restriction, and alcohol consumption among American and German college students.

To assess relationships between alcohol consumption and two dimensions of drinking restraint (temptation and restriction), American and German college students were given the Khavari Alcohol Test (KAT) and the Temptation and Restraint Inventory (TRI). As hypothesized, drinking temptation was a positive predictor of students' alcohol consumption in both countries, but there was no main effect for drinking restriction. Also as hypothesized, American students who were high on drinking temptation drank greater (not smaller) quantities of alcohol if they were also high on drinking restriction. Conversely, German students who were high on drinking temptation drank more alcohol if they were also low on drinking restriction. The results point to cross-cultural similarities and differences in relationships between drinking temptation and restriction and actual alcohol consumption.     

Alcoholic men endorse more DSM-IV withdrawal symptoms than alcoholic women matched in drinking history

OBJECTIVE: Given gender differences in alcohol metabolism, drinking patterns and alcohol-related problems, we asked whether men and women recruited for research protocols from treatment programs would meet different subsets of alcohol dependence or withdrawal criteria or differ in current level of functioning. METHOD: The subjects were 66 men and 62 women meeting DSM-III-R or DSM-IV criteria for alcohol dependence. Gender differences were tested infrequency counts of criteria endorsed and Global Assessment of Functioning (GAF) scores. RESULTS: All seven alcohol dependence criteria were endorsed by 50% of the sample. There were no significant gender differences in frequency of individual criteria endorsed. However, more men than women tended to endorse the withdrawal criterion for alcohol dependence and the tremor criterion for alcohol withdrawal, whereas women had higher GAF scores. When subgroups of men and women were matched on alcohol consumption variables, significantly more men than women endorsed the withdrawal criterion for alcohol dependence and the anxiety criterion for alcohol withdrawal, and women still had significantly higher GAF scores than men. CONCLUSIONS: DSM criteria provide a similar characterization of alcohol dependence in male and female research volunteers. Despite this similarity, the DSM criteria were sensitive to gender differences, which can now be challenged with rigorous testing.     

Factor structure of the Alcohol Use Disorders Identification Test (AUDIT) for men and women in different age groups.

OBJECTIVE: Our main aim was to investigate the factor structure of the Alcohol Use Disorders Identification Test (AUDIT) in a Finnish population sample. METHOD: The AUDIT was completed by 3,125 men (mean age = 26.2 years) and 6,006 women (mean age = 26.1 years). RESULTS: At a cutoff score of 8 or more, 49.8% of the men and 23.9% of the women would be identified as potentially engaged in excessive alcohol use. Exploratory factor analyses suggested a two-factor solution for both men and women. However, the factor structure was not invariant between men and women or in the different age groups among men. CONCLUSIONS: This is one of the largest known general population studies on alcohol use in recent years in Finland. The findings support a two-factor solution, and it is suggested that the AUDIT cutoff scores should be tailored according to age, gender, and drinking culture.     

Drinking restraint, alcohol consumption and alcohol dependence among children of alcoholics

OBJECTIVE: Previous research has found drinking restraint to be a risk factor for alcohol use and alcohol-related problems in normative populations, but has not tested these relations in high-risk populations. The current study tested whether drinking restraint predicted alcohol-related outcomes in the same way for high-risk and low-risk individuals and tested whether there was a quadratic effect of drinking restraint on alcohol-related outcomes. METHOD: Data from an ongoing longitudinal study of children of alcoholics (COAs; n = 189) and controls (n = 192) were collected at two time points 5 years apart. RESULTS: The prospective findings extended previous cross-sectional literature by replicating the main effects of drinking restraint as a risk factor for subsequent drinking for controls. For COAs, however, higher levels of drinking restraint were associated with lower levels of later drinking. There was also a quadratic effect of drinking restraint in the prediction of alcohol dependence diagnoses, suggesting that those at the extreme levels of drinking restraint were least likely to develop alcohol dependence. CONCLUSIONS: The relation of drinking restraint to alcohol-related outcomes may be more complex than previously hypothesized because it may work in different directions for high- and low-risk individuals and may have a nonlinear relationship to diagnostic outcomes.