INTRODUCTION

Simple definition of “alcohol craving” may be “a strong desire” to take alcohol. Cravings represent subjectively experienced, motivational states that are associated with on-going drug use in drug dependent individuals (1). Although many alcoholics consistently experience craving, researchers have not yet developed a common, valid definition of the phenomenon (2). A recent study which reviewed 18 models in past 60 years, suggested that no single model explains craving completely (3). Nevertheless, craving has been linked both to poorer outcomes following treatment and greater attrition during treatment (4-7). Alcohol craving is generaly considered as a core symptom of alcohol dependence and a strong predictor of relapse in alcohol-dependent adults (7-10).

There are several questionnaires available to quantify craving in adults (11). One instrument, the Obsessive Compulsive Drinking Scale (OCDS) appears to be emerging as the gold standard for this purpose (4). OCDS was developed to reflect the correspondence between key features of obsessive–compulsive anxiety disorder and key features of substance dependence (12). This model proposes that the intrusive and disruptive thoughts and images that accompany drug craving are analogous to anxiety-provoking obsessions, and that drug seeking, excessive consumption and resulting impairment are analogous to the repetitively performed and ritualized compulsions one engages in to reduce anxiety (11).

Several clinical, neurobiological, and neuropsychological data suggest that both obsessive thoughts about alcohol use and compulsive behaviour towards drinking are part of craving. Modell et al. (13) suggested that some aspects of alcohol craving (obsessive, recurrent and persistent thoughts about alcohol and compulsive drive to consume alcohol) have a phenomenological overlap with the obsessive-compulsive syndrome. Obsessive thoughts and compulsive drinking behaviors have been proposed as key factors associated with the loss of control over alcohol consumption experienced by alcohol-dependent patients (14). Modell et al. (15) modified the Yale–Brown Obsessive Compulsive Scale (16) for use in alcohol-dependent patients (Yale–Brown Obsessive Compulsive Drinking Scale for heavy drinking: YBOCS-hd). On this basis, Anton et al. (17) developed the Obsessive Compulsive Drinking Scale (OCDS) consisting of 14 queries as a self-rating instrument for quantifying cognitive aspects of alcohol craving with a good reliability, consistency and validity. Anton et al. differentiated in a dichotomous model between the obsessive and the compulsive subscale. The ease of use (it can be completed in 5 minutes), the reproducibility, the validity, and the analytic capacity make the OCDS a very effective and useful questionnaire during trials for the treatment of patients with alcohol problems, while also proving to be significantly related to the severity of alcoholism (4,12,18). The widespread use of the OCDS can also be seen from the OCDS versions validated in other languages, e.g. French (19), Japanese (20) and Italian (21). Although the Turkish version of the YBOCS-hd proved to be a reliable and valid instrument measuring craving in alcohol-dependent male individuals (22), until now, validation study of OCDS in Turkish population had not been conducted.

The purpose of this study was to translate and to validate the Turkish version of the OCDS, to assess its reliability, internal consistency and factor structure, to compare it with the Italian (21), Dutch (23) and French (24) studies according to Dutch (23) and French (24) versions, which used 14 and 10 items solutions of the scale respectively.

METHODS

Settings and sample

The study was conducted in Bakırköy Research and Training Hospital for Psychiatry, Neurology and Neurosurgery, Alcohol and Drug Research, Treatment and Training Center (AMATEM) in Istanbul between August 2008 and March 2009. AMATEM is a specialized center for substance use disorders with 84 inpatient beds, and accepts patients from all over Turkey. The Ethical Committee of the hospital approved the study. Patient’s written informed consent was obtained after the study protocol was thoroughly explained.

One hundred and twenty consecutively admitted alcohol-dependent inpatients without history of any other substance abuse were considered for participation in the study. All participants met the DSM-IV diagnostic criteria for alcohol dependence. Excluding criteria were illiteracy, mental retardation or cognitive impairment and comorbid psychotic disorder. Five patients were excluded due to illiteracy and three patients due to cognitive deficits. Although none of the patients refused to participate in the study, 16 patients were excluded because they left some parts of the scales unfilled, did not give the forms back or left the treatment program prematurely; i.e. before filling the forms. A total of 155 alcohol-dependent inpatients participated in the study. Interviews with the study group were conducted after detoxification period, i.e. 4-6 weeks after the last day of alcohol use.

The original OCDS was independently translated from English into Turkish by two experts in alcoholism. Consensus was reached on a common draft by these experts. This Turkish version was back translated into English by an independent translator. The final Turkish OCDS was then, first applied to 155 patients and administration was repeated again after 24 h to 136 of these 155 patients in a test–retest procedure to assess the retest reliability. All patients received the test to be completed in the morning; moreover, at the same time a visual analogue scale (VAS: 10 cm) for alcohol craving severity (frequency + intensity) was applied. The patient was asked to rate his current level of craving intensity from 0 (“no craving”) to 10 (“worst imaginable craving”) and level of craving frequency from 0 (“no craving”) to 10 (“most frequent craving”).

No patients exhibited acute withdrawal symptoms; psychotropic medication was allowed when indicated, but specific drugs endowed with anti-craving properties or able to prevent relapses were avoided. The daily amounts of alcoholic beverages consumed by the members of the study were expressed in drinks per day (one standard drink equal to 12 g of absolute alcohol).

Measures

All patients were assessed by using a semi-structured socio-demographic form. The diagnosis of alcohol or drug dependence in each participating patient based on the clinical examination, a screening interview based on the Structured Clinical Interview for DSM-IV (SCID-I) (25), Turkish version (26), conducted by a trained interviewer (CE).

Obsessive–Compulsive Drinking Scale (OCDS)

Several clinical, neurobiological, and neuropsychological data suggest that both obsessive thoughts about alcohol use and compulsive behaviour towards drinking are part of craving. Modifying an interview-based questionnaire (Yale–Brown Obsessive Compulsive Drinking Scale: YBOCS-hd), Anton et al., (17) developed a self-administered questionnaire consisting of 14 queries, the Obsessive–Compulsive Drinking Scale (OCDS), which includes items to evaluate both total craving and its obsessive and compulsive components. The ease of use (it can be completed in 5 minutes), the reproducibility, the validity, and the analytic capacity make the OCDS a very effective and useful questionnaire during trials for the treatment of patients with alcohol problems, while also proving to be significantly related to the severity of alcoholism (4,12,18).

The Penn Alcohol Craving Scale

Alcohol craving was measured with The Penn Alcohol Craving Scale (PACS) which is a 5-item measure that assesses frequency, severity and time spent thinking about alcohol, difficulty in resisting relapse opportunities, and strength of craving episodes (27). The PACS consists of five items each scored 0–6 in increasing severity of craving. Prior research has demonstrated the PACS to have greater predictive value for treatment outcomes compared to the Obsessive–Compulsive Drinking Scale or the Alcohol Urge Questionnaire (8). PACS was shown to be reliable and valid instrument for evaluating craving (27). Turkish version of the PACS is valid and reliable for screening severity of craving of alcohol dependent patients (28). Cronbach’s alfa for present study was found as 0.96.

Michigan Alcoholism Screening Test

The severity of dependence was assessed by using the Michigan Alcoholism Screening Test (MAST), (29) which was developed as a “rapid and effective screening for lifetime alcohol-related problems and alcoholism” for a variety of populations. Turkish version of the MAST is valid and reliable for screening severity of dependency of both alcohol and drug dependent patients (30). The Cronbach’s alpha was 0.75 in the present study.

Statistical analysis

As to the OCDS scores (ranging from 0—no symptoms, to 4—severe symptoms), obsessive and compulsive subscales OB: items 1–6, CP: items 7–14) and the total scale was considered separately. As in previous studies two different methods were used to calculate these three variables (23,24), both of them were followed in the Italian study (21) in order to compare their results with those of the Dutch (23) and French (24) research groups. We too used both calculation system and compared with all these three studies (21,23,24).

In the Dutch method, sum scores of items 1 to 6 (maximum score: 24) is obsessive subscale (OB-D), whereas sum scores of items 7 to 14 (maximum score: 32) is compulsive subscale (CP-D). Thus total score of the OCDS-D (maximum score: 56) is calculated by adding OB-D and CP-D.

In the French method, one of the highest score from items 1 and 2 is taken to compute OB-F subscale. Similarly, for calculating CP-F subscale, highest scores from items 7 and 8, from 9 and 10 and from 13 and 14. Thus maximum score is 20 for OB-F, 20 for CP-F and 40 for OCDS-F (OB-F + CP-F).

The reliability of the OCDS was assessed using Cronbach α which evaluates the internal consistency of the questionnaire, based on the correlation between items. Pearson’s linear correlation analysis was employed to verify the correlations between variables (OB, CP and OCDS, calculated with both Dutch and French methods, and VAS). The same analysis was used to assess the test–retest correlations for OCDS, OB and CP subscales, and VAS. Goodness of fit with a normal distribution was tested by the Kolmogorov–Smirnov test. Principal component analysis was performed to detect the underlying dimensionality of the scale. The 14 OCDS items were standardized so that the method was performed on variables with means equal to 0 and variance equal to 1. The inter-item correlation matrix was factor analyzed. As a rotation method, Varimax with Kaiser Normalization was used. The eigenvalue-greater-than-one criterion was used to determine the number of relevant factors.

RESULTS

Sociodemographic variables and variables related with alcohol use are shown on Table 1 (Table1). Corrected Item-Total Correlations ranged between 0.40 and 0.69 (Table 2).

In the first evaluation, three factor solutions were found. Consistent with the original scale, items 1,2,4,5 and 6 computed “Obsessive factor” (explained 22.56% of variance), whereas items 7, 8, 11, 12, 13, and 14 computed “Compulsive factor” (explained 24.34% of variance). Third item originally from Obsessive factor (Degree of obsessive to interfere with social or work functioning) and 9 and 10th items (Degree of compulsions to interfere with social or work functioning) originally from Compulsive factor computed third factor, which explained 17.70% of variance (total of 64.6% of variance). These items were related with evaluation the function impairment related with obsessions and compulsive drinking, thus called as “Function factor” (Table 3). As a single factor, 14 items explained 41.55% of variance, 10 item 45.09% of variance. The resulting factors seem to describe the scale as a whole with its original distinction into two groups of items. Thus two factor solution was appropriate for Turkish version as in original scale.

Correlations between items, subscales and total score of OCDS: For each of the items, the corrected item-total correlation values were between 0.49 and 0.75 (p<0.001) for OCDS-D, whereas they were between 0.52 and 0.78 (p<0.001) for OCDS-F (Table 4).

Test–retest correlations and correlations of OCDS and subscales of OCDS with MAST, amount of drinks per day and PACS were shown on Table 4. Test–retest correlations was 0.64 for “Obsessive-D” scale, 0.74 for “Compulsive-D” scale, 0.75 for “OCDS-D”, 0.65 for “Obsessive-F” scale, 0.72 for “Compulsive-F” scale, 0.74 for “OCDS-F”. The subscales and total scores of the OCDS were correlated significantly with MAST, PACS and amount of drinks per day (p<0.001) (Table 5). The duration of the alcoholism history was not correlated with OCDS-D, OCDS-F or their subscales (not shown).

Means ± S.D. (calculated according to the Dutch (D) and French (F) methods) and Cronbach’s α values were compared to those found in the Dutch (23), French (24) and Italian studies (21). In alcohol dependents, the internal consistency coefficient (Cronbach’s alpha) was 0.83 for “Obsessive-D” scale, 0.84 for “Compulsive-D” scale, 0.89 for “OCDS-D”, 0.81 for “Obsessive-F” scale, 0.77 for “Compulsive-F” scale and 0.86 for “OCDS-F” (Table 6).

Pearson linear correlations between subscales, total OCDS and VAS scores calculated according to the Dutch (D) (23) and French (F) (24) methods (Table 7).

DISCUSSION

The results of the present study demonstrate that the OCDS has been successfully translated into Turkish. The Cronbach’s α analysis showed good reliability and construct validity for the Turkish version of the scale, comparable with the original and the European versions. Our sample size was larger (n=155) than those of the Italian (n=103), French (n=50) and Dutch (n=39) studies, with which the results of our study was compared. Turkish version of the both French (10 item) and Dutch (14 item) solutions were found to be compatible with original scales. The fact that the α values based on the calculation of Dutch method (23) and the French method (24) were both statistically significant, did not support the validity of grouping the score of four pairs of items and balancing the score range between OB and CP. This finding was consistent with the Italian study (21). Also the results of Schippers et al. (23) were relevant for this conclusion, as they found the same validity for the original and the substituted versions of the scale.

Anton et al. (17) created the Obsessive–Compulsive Drinking Scale (OCDS) by modifying a self-report questionnaire originally designed to measure obsessive cognitions and compulsive rituals (13). Their questionnaire assessed; the frequency and duration of distress, resistance to social–occupational problems caused by one’s thoughts, impulses and images of drinking, the degree to which drinking interferes with one’s daily functioning, and attempts to resist and to control one’s drinking. Factor analysis of the scale in different countries showed different numbers of factors. Australian study identified a 4-factor solution (compulsions, interference, obsessions and resisting obsessions) (31). Mexican study identified 2-factor solution in 12-item version of the scale (excluding the items on drinking habits) explaining 56.9% of the variance (obsessive thoughts related to drinking and interference/behaviors related to drinking) (14). In the present study, the results from the corrected item-total correlation analysis demonstrated that OCDS was a scale consisting of obsessive and compulsive components, with 14 items assessing the same phenomenon. A three-factor solution might better describe its structure. Principal component analysis of the Turkish version of the 14 OCDS items showed that there were three eigenvalues greater than 1 and that these three factors explained 64.60% of the variance. The first factor explained 24.34% of the variance and was represented by the Compulsive component. The second factor was taken into consideration, it was able to explain another 22.56% of the variance and represented by the Obsessive component. Finally, the third factor was able to explain another 17.70% of variance, thus reaching 64.60% of the cumulative variance. It could be easy to recognize that the first factor discriminated nearly all CP items, the second factor discriminated nearly all OP items, while the third factor discriminated function impairment related with most CP items. A possible explanation for these findings was that our study population was mostly made up of severely dependent patients, as demonstrated by the heavy amounts of alcohol daily consumed by them (see Table 1), and also of nearly half “type 2-like” alcoholics, who were expected to experience the highest levels of craving. Also regardless of the type of alcoholism, treatment seeking population were severely dependent population, which may all increase the probability of function impairment. Nevertheless, with respect to the findings of Janiri et al. (21) in the Italian study, who identified three-factor solution, we can confirm the three-factor solution for the scale, but the resulting factors seemed to describe the scale as a whole with its original distinction into two groups of items as Janiri et al. suggested in their population.

The correlation between OB, CP and OCDS total score was high, as was expected from their association in the construction of the scale. The correlations between the three variables and VAS were also significant, even to a higher extent than those obtained in the Italian study (21). Finding lower correlations, Anton et al. (4) suggested that OCDS may capture a broader dimension of alcoholism with respect to the analogue measures of craving, which usually leave the interpretation of its meaning to the patient. This was supported by several studies which indicated that the meaning of craving differs among substance-dependent subjects and professional caregivers (3), and this discrepancy could be explained by the multidimensionality of the concept of craving.

In line with the Italian (21) and Dutch (23) studies, the duration of the alcoholism history was not correlated with OCDS. In contrast with Italian study (21), the alcohol consumption prior to abstinence significantly influenced the OCDS scores, similarly with Anton’s study which demonstrated a correlation between OCDS and previous alcohol consumption (4). This discrepancy may be due to different enrolment procedures used in different studies. When the relation between OCDS and the alcohol intake was taken into consideration, the fact should be reminded that the two OCDS items investigated alcohol consumption, while all subjects, as those in present study, were currently abstinent.

In a previous study, OCDS scores were significantly correlated with measures for the Alcohol Dependence Scale, number of DSM-IV criteria met for alcohol dependence as well as the number of days in a week engaged in heavy drinking, indicating concurrent validity (14). In the present study, OCDS scores were correlated with severity of alcohol related problems, amount of drinks per day, and severity of craving measured with PACS and VAS. These suggest the concurrent validity of the OCDS’s Turkish version. OCDS score was found to be predictive of the hazard for heavy relapse in the following week (6) future total alcohol consumption during a treatment period of 12 weeks if obtained during a period of relative abstinence (4) and 12 months after treatment completion (7). Thus further studies evaluating the predictive value of the Turkish version of OCDS for relaps is essential.

In conclusion, results obtained in this study suggest that the Turkish version of the OCDS (both 10 item and 14 item solutions) are reliable and valid for alcohol dependent inpatients. Also the Turkish version of the scale was found to be compatible with the results of the Dutch, French and Italian studies.

REFERENCES

1. Sinha R, O’Malley SS. Craving for alcohol: findings from the clinic and the laboratory. Alcohol Alcohol 1999; 34:223-230.

2. Anton RF. What is craving? Models and implications for treatment. Alcohol Res Health 1999; 23:165-173.

3. Skinner MD, Aubin HJ. Craving’s place in addiction theory: contributions of the major models. Neurosci Biobehav Rev 2010; 34:606-623.

4. Anton RF, Moak DH, Latham P. The obsessive compulsive drinking scale: A new method of assessing outcome in alcoholism treatment studies. Arch Gen Psychiatry 1996; 53: 225–231.

5. Kranzler H, Mulgrew C, Modesto-Lowe V, Burleson J. Validity of the Obsessive Compulsive Drinking Scale (OCDS): Does craving predict drinking behaviors? Alcohol Clin Exp Res, 1999; 23:108–114.

6. Roberts JS, Anton RF, Latham PK, Moak DH. Factor structure and predictive validity of the Obsessive Compulsive Drinking Scale. Alcohol Clin Exp Res 1999; 23:1484-1491.

7. Bottlender M, Soyka M. Impact of craving on alcohol relapse during, and 12 months following, outpatient treatment. Alcohol Alcohol 2004; 39: 357–361.

8. Flannery BA, Poole SA, Gallop RJ, Volpicelli JR. Alcohol craving predicts drinking during treatment: an analysis of three assessment instruments. J Stud Alcohol 2003; 64:120-126.

9. de Bruijn C, Korzec A, Koerselman F, van Den Brink W. Craving and withdrawal as core symptoms of alcohol dependence. J Nerv Ment Dis 2004; 192:494-502.

10. Evren C, Cetin R, Durkaya M, Dalbudak E. Clinical variables related with relapse during 6 month follow-up in male alcohol dependents. Bulletin of Clinical Psychopharmacology 2010; 20:14-22.

11. Rosenberg H. Clinical and laboratory assessment of the subjective experience of drug craving. Clin Psychol Rev 2009; 29:519-534.

12. Anton RF. Obsessive–compulsive aspects of craving: Development of the obsessive compulsive drinking scale. Addiction 2000; 95 (Suppl. 2): 211–217.

13. Modell JG, Glaser FB, Cyr L, Mountz JM. Obsessive and compulsive characteristics of craving for alcohol in alcohol abuse and dependence. Alcohol Clin Exp Res 1992;16: 272-274.

14. Cordero M, Solís L, Cordero R, Torruco M, Cruz-Fuentes C. Factor structure and concurrent validity of the obsessive compulsive drinking scale in a group of alcohol-dependent subjects of Mexico City. Alcohol Clin Exp Res 2009;33: 1145-1150.

15. Modell JG, Glaser FB, Mountz JM, Schmaltz S, Cyr L. Obsessive and compulsive characteristics of alcohol abuse and dependence: quantification by a newly developed questionnaire. Alcohol Clin Exp Res 1992;16: 266-271.

16. Goodman WK, Price LH, Rasmussen SA, Mazure C, Fleischmann RL, Hill CL, Heninger GR, Charney DS. The Yale-Brown Obsessive Compulsive Scale. I. Development, use, and reliability. Arch Gen Psychiatry 1989; 46:1006-1011.

17. Anton RF, Moak DH, Latham P. The obsessive compulsive drinking scale: A self-rated instrument for the quantification of thoughts about alcohol and drinking behavior. Alcohol Clin Exp Res 1995; 19:92–99.

18. Moak DH, Anton RF, Latham PK. Further validation of the Obsessive-Compulsive Drinking Scale (OCDS). Relationship to alcoholism severity. Am J Addict 1998; 7:14-23.

19. Chignon JM, Jacquesy L, Mennad M, Terki A, Huttin F, Martin P, Chabannes JP. Self-assessment questionnaire of alcoholic craving (ECCA Questionnaire): Behaviour and cognition in relation to alcohol: French translation and validation of the obsessive–compulsive drinking scale. Encephale 1998; 24:426–434.

20. Tatsuzawa Y, Yoshimasu H, Moriyama Y, Furusawa T, Yoshino A. Validation study of the Japanese version of the Obsessive-Compulsive Drinking Scale. Psychiatry Clin Neurosci 2002; 56:91-95.

21. Janiri L, Calvosa F, Dario T, Pozzi G, Ruggeri A, Addolorato G, Di Giannantonio M, De Risio S. The Italian version of the obsessive–compulsive drinking scale: Validation, comparison with the other versions and difference between type 1- and type 2-like alcoholics. Drug Alcohol Depend 2004; 74:187–195.

22. Ilhan IO, Demirbas H, Dogan YB. Validation study of the Turkish version of the Yale - Brown obsessive compulsive scale for heavy drinking in a group of male patients. Drug Alcohol Rev 2006; 25:357-360.

23. Schippers GM, De Jong CAJ, Lehert P, Potgieter A, Deckers F, Casselman J, Geerlings PJ. The Obsessive–Compulsive Drinking Scale: translation into Dutch and possible modifications. Eur Addict Res 1997; 3:116–122.

24. Ansseau M, Besson J, Lejoyeux M, Pinto E, Landry U, Cornes M, Deckers F, Potgieter A, Ades J. A French translation of the obsessive-compulsive drinking scale for craving in alcohol-dependent patients: a validation study in Belgium, France, and Switzerland. Eur Addict Res 2000; 6:51-56.

25. First MB, Spitzer RL, Gibbon M, Williams JBW. Stuructured Clinical Interview for DSM-IV Axis I Disorders (SCID-I), Clinical Version. Washington D.C. and London: American Psychiatric Press, Inc, 1997.

26. Corapcioglu A, Aydemir O, Yildiz M, Esen A., Koroglu E. DSM-IV Eksen I Bozuklukları (SCID-I) için Yapılandırılmış Klinik Görüşme, Klinik Versiyon. Ankara. Hekimler Yayın Birliği, 1999.

27. Flannery BA, Volpicelli JR, Pettinati HM. Psychometric properties of the Penn Alcohol Craving Scale. Alcohol Clin Exp Res 1999; 23:1289-1295.

28. Evren C, Flannery B, Çelik R, Durkaya M, Dalbudak E. Penn Alkol Aşerme Ölçeği (PAAÖ) Türkçe şeklinin yatarak tedavi gören erkek alkol bağımlısı hastalarda geçerliği ve güvenirliği. Bağımlılık Dergisi 2008; 9:128-134.

29. Gibbs LE. Validity and reliability of the Michigan Alcoholism Screening Test: A review. Drug Alcohol Depend 1985; 12: 279-285.

30. Coskunol H, Bagdiken I, Sorias S, Saygili R. Michigan Alkolizm Tarama Testinin geçerliliği. Ege Tıp Dergisi 1995; 34: 15-18.

31. Connor JP, Jack A, Feeney GF, Young RM. Validity of the obsessive compulsive drinking scale in a heavy drinking population. Alcohol Clin Exp Res 2008; 32:1067-1073.