Abstract
Aims
To examine the role of abstinence from gambling versus controlled gambling in recovery from pathological gambling (PG) in a community-based survey.
Design
Individuals with a lifetime history of PG identified in a community-based survey were divided into three groups based on their current levels of PG symptoms. These three groups were compared to each other on their past-year gambling involvement.
Setting
National general population twin survey conducted in Australia.
Participants
Overall, there were 4,764 participants in the community-based survey (mean age 37.7 years, 57.2% women). Among these were 104 participants with a lifetime history of PG; of the 104 with a lifetime diagnosis of PG, 28 had a past-year diagnosis of PG, 32 had past-year problem gambling, and 44 had no symptoms of PG in the past year (“recovery”).
Measurements
The measure of PG was based on the NODS (NORC DSM-IV Screen for Gambling Problems). Past-year participation in 11 different gambling activities was assessed, as well as the following composite indicators: any gambling, gambling versatility, the number of days and hours spent gambling, and the proportion of household income spent on gambling.
Findings
Ninety percent of those in the recovery group participated in some form of gambling in the past year.
Conclusions
In this general population survey, nearly all of the PG recoveries were achieved in the absence of abstinence. Controlled gambling appears to be a popular road to recovery in the community.
Abstinence has been the only acknowledged treatment goal in most pathological gambling (PG) interventions (1). For example, a popular form of treatment for PG, Gamblers Anonymous (GA), takes a very strong stance on the importance of complete abstinence from gambling for achieving recovery (2). In treatment studies, abstinence from all forms of gambling has traditionally been required for the treatment to be considered a success (3–5), although this position has been challenged by demonstrating that some individuals who undergo treatment for PG are able to learn to continue to gamble without problems (1,3,6).
The demonstration of recovery from PG in the absence of abstinence (“controlled gambling”) has typically been obtained in the context of PG treatment studies in which abstinence was the specified treatment goal. For example, in a one-year PG treatment follow-up study, 15 of 45 (33%) individuals who were asymptomatic had participated in some gambling in the past 6 months (6). More recently, studies of the efficacy of PG treatment with controlled gambling as a treatment goal also provide evidence of the viability of recovery from PG without abstinence (1).
Most recoveries from PG, however, are achieved without treatment (7–8). Over 80% of PG recoveries were attained without treatment in recent United States and Australian national surveys (7–8). Because those who seek treatment for PG are an unrepresentative subsample of individuals with PG, a community-based epidemiologic survey that includes individuals who have achieved recovery naturally as well as with the assistance of treatment may provide the best evidence to inform the debate about whether recovery from PG must be accompanied by abstention from gambling.
There is limited information about the extent to which PG recovery in the community is associated with abstinence versus controlled gambling outcomes. The aim of the present study was to examine the role of abstinence from gambling in recovery from PG in a community-based survey. In a previous paper, Slutske et al (8) identified 44 individuals from a national community-based survey who had a lifetime history of PG but who were asymptomatic in the past year. The level of gambling involvement among these recovered individuals was compared to that of two other groups with a lifetime history of PG who were symptomatic with PG or problem gambling in the past year.
Methods
Participants
Participants for this study were 4,764 members of the Australian Twin Registry (ATR) Cohort II (9), a general population sample of twin pairs. In 2004–2007, a telephone interview containing a thorough assessment of gambling behaviors was conducted with ATR Cohort II members (individual response rate of 80.4%; see 9 for more details). It is not always recognized that general population surveys of twins can be useful for more than just answering questions about genetic and environmental underpinnings of behavior. In addition to being a sample of twins, this study was also designed to address basic epidemiologic questions about gambling behavior.
Measures
All of the measures for this study, including PG and gambling involvement, were obtained via a structured psychiatric telephone interview. Prior to the interview, participants were mailed a respondent booklet that was used to assist with some of the interview questions. The respondent booklet contained response options that participants could select.
Pathological Gambling
The measure of PG was based on the NORC DSM-IV Screen for Gambling Problems (NODS) (10). The lifetime occurrence of and ages of onset and recency for each of the 10 DSM-IV (11) PG symptoms as well as periods of symptom clustering were assessed. The NODS DSM-IV diagnostic criteria were assessed for all participants who reported that they had ever gambled at least five times within a single 12 month period; the majority of participants, 77.5%, surpassed this gambling threshold. The lifetime prevalence of DSM-IV PG in the overall study sample was 2.2% (n=104). The test-retest reliability of the lifetime diagnosis of PG was very good (kappa = 0.67; Yule’s Y = 0.79) based on a 3.4 month retest interview completed with 166 participants in this study (8). The validity of DSM-IV PG diagnoses based on the NODS has been demonstrated in several studies -- 82% of individuals diagnosed with lifetime PG using the NODS (12) and 97% of individuals diagnosed with past-year PG using the NODS (13) also obtained a diagnosis of PG using the South Oaks Gambling Screen (SOGS)(14).
In the DSM-IV, a diagnosis of lifetime PG requires that at least 5 of 10 symptoms are experienced, and there is no explicit requirement that the symptoms co-occur. None of the previous epidemiologic surveys of DSM-IV PG have incorporated symptom co-occurrence (“clustering”) as a part of the diagnosis of lifetime PG (10,15–17). Rather than rely on the assumption that multiple symptoms will tend to co-occur, we directly assessed whether a participant with more than one PG symptom had ever had a 12-month period of symptom co-occurrence. Most of the 104 individuals with a lifetime history of DSM-IV PG (79%) also reported that they had experienced a 12-month period in which at least 3 symptoms of PG had co-occurred (“symptom clustering”; 8).
Participants also were administered the SOGS (14). The SOGS was not used for the primary aims of this paper, but a single SOGS item (#6) was used as an indicator of gambling-related impairment. Participants were asked: “Do you feel that you have ever had a problem with your gambling?” Those who responded yes to this lifetime question were asked the following past-year question: “Do you feel that you have had a problem with your gambling in the last 12 months?” This “gambling problem recognition” question (8) was used as an alternate indicator of (low) level of impairment from gambling in the past year.
Gambling Involvement
A thorough assessment of lifetime and past-year gambling was conducted, including any involvement in 11 types of gambling (lottery, electronic gambling machines, instant scratch tickets, betting on horse or dog races, playing casino table games, keno, bingo, card games, betting on a sporting event, betting on games of skill, and internet casino gambling). In addition to participation in each of the specific gambling activities, two composite indices of gambling were also examined: participation in any of the 11 activities, and the number of different gambling activities that were practiced (“gambling versatility”).
Three indicators of past-year gambling involvement based on the amount of time and money spent on gambling were also assessed (18). The number of days spent gambling in the past year was based on the response to a question: “on how many days in total have you participated in any form of gambling in the last 12 months?” with 14 response options (presented in the respondent booklet) ranging from “never” to “every day”. Because different forms of gambling vary in terms of the time commitment required, the number of days on which gambling occurred may not fully capture the amount of time spent on gambling. Therefore, we also derived an index of the number of hours spent on gambling in the past year. This was computed from the item “in the last 12 months, on a typical day when you gambled, how much time would you spend on gambling?” with open-ended responses in hours or minutes recorded. The number of hours spent on a typical gambling day was multiplied by the number of days gambled in the past year to obtain the number of hours spent on gambling in the past year. The percent of yearly income spent on gambling in the past year was computed from the item “in the last 12 months, on a typical day when you gambled, how much money would you spend on gambling? (By money spent on gambling, I mean the total amount that you started out with at the beginning of the day minus the total amount that you ended up with at the end of the day)” with 10 response options (presented in the respondent booklet) ranging from “less than $1” to “10,001 or more”. The number of dollars spent on a typical gambling day was multiplied by the number of days gambled in the past year to obtain the number of dollars spent on gambling (net losses) in the past year. This was then divided by the yearly household income reported in a different section of the interview to obtain the percent of yearly income spent on gambling.
Participants were also asked about their gambling involvement during a period when they were spending more time gambling than they had during the last 12 months. These questions were identical to the questions about past-year gambling, except that the questions included the phrase “during that period when you were gambling the most” rather than “in the last 12 months.” For the purposes of this study, we focus on the number of days gambled and the number of hours gambled during the heaviest gambling period. Both of these items were converted to a one-year metric (i.e. number of days in a year, number of hours in a year) even though the heaviest gambling period may not have been of one-year duration. This facilitated comparisons with the past-year gambling questions.
Data analysis
PG recovery was defined as a lifetime history of DSM-IV PG without endorsement of any PG symptoms in the past 12 months -- 44 individuals fit into this category. Individuals in this recovery group were compared to 28 individuals with a past year diagnosis of PG and to 32 individuals with a lifetime diagnosis of PG who had some gambling problems (1–4 PG symptoms) within the past 12 months (“problem gambling”); all of the individuals in these two groups had their onsets of PG symptoms prior to the past year, and 58 of 60 had experienced at least 5 symptoms of PG prior to the past year.
It is not clear how long an individual must be symptom free to be considered recovered from PG. For treatment-outcome studies, a consensus of experts has suggested that a long-term follow-up of two years or more following completion of treatment is needed to establish that the treatment has been successful and that changes made in gambling involvement are stable (18). By this benchmark, most of the individuals classified as recovered in the present study would be considered stable recoveries; the average number of years the individuals in the recovery group had been completely symptom-free was 4.7 (range = 1–11); 93% had been symptom free for 2 or more years (41 of 44) and 48% (21 of 44) had been symptom free for 5 or more years (8). Only 8 individuals in the recovery group had ever sought treatment for their gambling problems, therefore 36 (82%) would be considered natural recoveries (8).
Differences in past year gambling involvement for the three lifetime PG groups were examined using generalized least squares regression for continuous outcomes and logistic regression for binary outcomes. (Note that the 104 twin individuals included in these analyses were obtained from 102 different twin pairs. Therefore, treating these twin data as non-independent observations was unnecessary.) These same analytic techniques were also used to examine differences between the three lifetime PG groups in demographic characteristics and level of lifetime PG symptomatology (Table 1). Two of the lifetime PG groups -- those with past-year symptoms -- were more likely to be male than the recovered group. This is because women were more likely to recover from PG than were men (8). The three lifetime PG groups differed in levels of lifetime PG symptomatology (although the differences were not statistically significant). Comparisons of the three lifetime PG groups and the corresponding estimated parameters (means and percentages) were adjusted for sex and levels of lifetime PG symptomatology.
Table 1.
Characteristics of three groups with a lifetime history of pathological gambling identified from an Australian national survey.
| PG symptom levels in the past year | |||
|---|---|---|---|
| past year PG (5+ sxs) | past year problems (1–4 sxs) | PG recovery (0 sxs) | |
| n=28 | n=32 | n=44 | |
| Characteristic | |||
| Age (M) | 37.9a | 37.8a | 37.2a |
| Male (%) | 71a | 78a | 57b |
| Married (%) | 46a | 59a | 52a |
| Lifetime PG symptoms (M) | 7.0a | 6.4a | 6.2a |
Note: PG = pathological gambling;
estimates within rows with a different letter significantly differ at p < 0.05.
Results
The past-year PG, past-year problem gambling and past-year recovery groups significantly differed in past-year gambling problem recognition (χ2 = 25.13, df = 2, p < .0001). Only 2% (1 of 44) of individuals in the recovery group responded affirmatively to the question “Do you feel that you have had a problem with your gambling in the last 12 months?” compared to 54% and 34% in the past-year PG and past-year problem gambling groups, respectively.
Ninety percent of those in the recovery group participated in some form of gambling in the past year. On average, those in the recovery group participated in 3.6 different gambling activities, gambled on 54.1 days, spent 54.6 hours and 6.8 percent of their yearly household income on gambling in the past year. The following pattern was observed for the rates of any past year involvement for most of the different gambling outcomes: PG group > problem gambling group > recovered group (see Table 2).
Table 2.
Past year gambling involvement among three groups with a lifetime history of pathological gambling identified from an Australian national survey.
| PG symptom levels in the past year | |||
|---|---|---|---|
| past year PG (5+ sxs) | past year problems (1–4 sxs) | PG recovery (0 sxs) | |
| n=28 | n=32 | n=44 | |
| Past year gambling involvement | % | % | % |
| any gambling | 100a | 100a | 90a |
| lottery | 91a | 94a | 76a |
| scratch tickets | 78a | 69a | 59a |
| electronic gaming machines | 93a | 78a,b | 62b |
| horse or dog races | 76a | 69a | 52a |
| casino table games | 49a | 52a | 31a |
| keno at a club, hotel, or casino | 60a | 37a,b | 24b |
| bingo at a club or hall | 26a | 19a | 8a |
| card games (not at a casino) | 46a | 5b | 19c |
| betting on sporting event | 31a | 36a | 16a |
| betting on games of skill | 24a | 12a | 10a |
| mean (SE) | mean (SE) | mean (SE) | |
| # different gambling activities | 5.7a (0.41) | 4.9b (0.38) | 3.6c (0.33) |
| # days gambled | 175.9a (15.83) | 113.2b (14.72) | 54.1c (12.68) |
| # hours gambled | 438.2a (53.55) | 225.2b (49.78) | 54.6c (42.88) |
| % of yearly income spent on gambling | 64.0a (15.80) | 54.7a (14.62) | 6.8b (12.43) |
Note: Estimates are adjusted for the effects of sex and lifetime DSM-IV PG symptoms. PG = pathological gambling.
estimates within rows with a different letter significantly differ at p < 0.05.
A more detailed picture of the frequency and intensity of gambling involvement in the PG recovery group is presented in Table 3. The conclusions that can be drawn from Table 3 are: (1) other than lottery and scratch tickets, the most popular forms of gambling among the recovery group were electronic gaming machines (“pokies” in Australia), betting on horse or dog races, and betting on sporting events, (2) there was evidence of past-year desistance among the recovery group for every gambling activity, and (3) the forms of gambling that appeared to decrease the most were those that were played in a casino, hall, or club, that is, casino table games, keno, and bingo. In addition, during the period when they were gambling the most, those in the recovery group gambled on 168.1 days/year and spent 455.6 hours/year on gambling, on average. A matched pairs t-test indicated that this was significantly more frequent gambling than occurred during the past year (number of days: t = 6.92, df = 43, p < .0001; number of hours: t = 5.94, df = 43, p < .0001). That is, the past-year frequency of gambling in the recovery group represented a significant reduction from their heaviest gambling period.
Table 3.
Lifetime and past year gambling involvement among 44 individuals identified from an Australian national survey who had recovereda from pathological gambling.
| Lifetime | Past year | |||
|---|---|---|---|---|
| Ever | More than 100 days | Ever | More than 10 days | |
| Gambling activity | % | % | % | % |
| lottery | 93 | 39 | 77 | 48 |
| scratch tickets | 95 | 18 | 59 | 30 |
| electronic gaming machines | 95 | 36 | 61 | 27 |
| horse or dog races | 77 | 11 | 50 | 9 |
| casino table games | 75 | 2 | 30 | 5 |
| keno at a club, hotel, or casino | 70 | 2 | 20 | 2 |
| bingo at a club or hall | 59 | 5 | 11 | 0 |
| card games (not at a casino) | 50 | 5 | 18 | 2 |
| betting on sporting event | 23 | 11 | 14 | 5 |
| betting on games of skill | 20 | 2 | 9 | 2 |
Note:
pathological gambling recovery was defined as a lifetime history of DSM-IV PG without endorsement of any PG symptoms in the past 12 months.
Not surprisingly, given that 36 of the 44 individuals in the PG recovery group represented natural recoveries, a group composed solely of those 36 individuals who attained recovery without treatment was very similar to the full PG recovery group. Eighty-nine percent of those in the natural recovery group participated in some form of gambling in the past year, and on average, had participated in 3.7 different gambling activities, gambled on 51.0 days, spent 48.1 hours and 6.7 percent of their yearly household income on gambling in the past year.
Discussion
If abstention from gambling was required to be considered successfully “recovered” from PG only 10% of the individuals defined as recovered in the present study would have met this criterion, even though they were completely asymptomatic. Nearly all of the individuals who had recovered from DSM-IV PG had participated in some gambling in the past year.
Cross-sectional comparisons of gambling involvement in the recovered group with the two other non-recovered groups (that is, the past-year PG and past-year problem gambling groups) suggests that although recovery from PG was not accompanied by a complete cessation of all gambling, it probably was accompanied by a substantial reduction in gambling involvement. Comparisons of past-year reports with retrospective reports of gambling involvement obtained from the recovered individuals for a previous period of heavier gambling were consistent with the cross-sectional group comparisons in showing that there was a substantial reduction in the frequency of gambling. These results are consistent with the possibility that some gambling may not be incompatible with recovery from PG. If reducing gambling rather than abstaining from gambling was a treatment goal available to individuals with gambling problems, it is possible that more individuals would seek treatment and the rate of treatment-seeking among individuals with PG would be higher (1).
Similar arguments have been made for the treatment of other addictive disorders. Alcoholism recovery in the absence of abstinence has been the focus of research and (sometimes heated) debate (19–20) for over half a century (21) because it was recognized that some individuals will not accept abstinence as a treatment goal (22). There are now a number of empirically supported alcoholism treatments with moderation drinking as a goal (21,23). Even when complete abstinence might be best for a given patient, Miller and Page (24) suggested that offering a “warm turkey” approach, that is, to offer a trial period of treatment with controlled drinking as the goal, might work for those who are reluctant to completely abstain from drinking. In the end, the controlled drinking goal may eventually evolve into an abstinence goal (“cold turkey”) if controlled drinking is not attainable. A similar strategy has been proposed for smoking cessation among smokers who are unmotivated to enter treatment and to abruptly quit smoking (25–26). Many of these smokers, when enrolled in a “reduced smoking” treatment program, are eventually able to achieve abstinence when they are allowed to reduce their smoking more gradually (25–26).
For some individuals faced with overcoming PG, “warm turkey” may be the best approach for achieving recovery, with or without abstinence. For example, over the course of a 12-week controlled-gambling PG treatment program, Ladouceur et al (1) found that only a minority of patients (34%) maintained the goal of controlled gambling throughout the entire course of treatment and that the majority of participants shifted to an abstinence goal at least once; by the end of the 12 weeks of treatment, 39% of the patients were pursuing an abstinence goal. Ladouceur et al (1) cited evidence supporting the idea that some individuals choose an initial goal of controlled gambling with the intention of eventually transitioning into an abstinence goal (27).
Limitations
There are limits to studying recovery from PG within the context of a cross-sectional epidemiologic survey. Recovery could be studied more effectively within the context of a longitudinal study in which participants were followed before, during, and after developing PG disorder, and the assessments obtained concurrently, rather than based on retrospective reports. Such a longitudinal study could collect more fine-grained information about the precipitating factors that prompted recovery, the methods that were used to achieve recovery, and the factors underlying the maintenance of recovery. Long-term follow-ups could confirm that individuals classified as “recovered” had achieved stable long-term recoveries. For example, although most of the recovered individuals in this study had been symptom-free for a number of years, it is still possible that individuals who were classified as “recovered” may actually be in remission and eventually relapse to PG.
Studying PG recovery in a cross-sectional epidemiologic survey of the general population also has a number of strengths: it yields results that are generalizable (28), allows one to estimate the proportion of affected individuals who recover with and without abstaining from gambling, and is probably the method that makes the fewest assumptions about who recovers and how recovery occurs. In a study of natural recovery from alcohol dependence based on two samples of individuals that were either identified from a community-based survey or from media solicitations of individuals who had recovered from alcohol dependence without treatment, Rumpf et al (29) obtained estimates of past-year abstinence of 15.6% versus 83.5%, suggesting that media-recruitment of individuals who have recovered from alcohol dependence may lead to inaccurate estimates of the likelihood of recovery in the absence of abstinence from alcohol in the community. The same may be true for estimating the likelihood of recovering from PG in the absence of abstinence from gambling in the community.
Despite limitations the results of this study were relatively clear. In a general population survey, nearly all of the PG recoveries were achieved in the absence of abstinence. Those who had recovered from PG were still gambling but were gambling less than those who were currently symptomatic. In sum, controlled gambling appears to be a popular road to recovery from PG in the community.
Acknowledgments
This work was supported by National Institutes of Health Grant MH66206.
References
- 1.Ladouceur R, Lachance S, Fournier P. Is control a viable goal in the treatment of pathological gambling? Behav Res Ther. 2009;47:189–197. doi: 10.1016/j.brat.2008.11.004. [DOI] [PubMed] [Google Scholar]
- 2.Ferentzy P, Skinner W, Antze P. Recovery in Gamblers Anonymous. Journal of Gambling Issues. 2006:17. doi: 10.1007/s10899-009-9167-3. available at http://www.camh.net/egambling/issue17/ferentzy1.html. [DOI] [PubMed]
- 3.Blaszczynski A, McConaghy N, Frankova A. Control versus abstinence in the treatment of pathological gambling: A two to nine year follow-up. Br J Addict. 1991;86:299–306. doi: 10.1111/j.1360-0443.1991.tb01782.x. [DOI] [PubMed] [Google Scholar]
- 4.Ladouceur R. Controlled gambling for pathological gamblers. J Gambl Stud. 2005;21:51–8. doi: 10.1007/s10899-004-1923-9. [DOI] [PubMed] [Google Scholar]
- 5.Petry NM. Pathological gambling: etiology, comorbidity, and treatment. Washington, DC: American Psychological Association; 2005. [Google Scholar]
- 6.Weinstock J, Ledgerwood DM, Petry NM. Association between posttreatment gambling behavior and harm in pathological gamblers. Psychol Addict Behav. 2007;21:185–193. doi: 10.1037/0893-164X.21.2.185. [DOI] [PubMed] [Google Scholar]
- 7.Slutske WS. Natural recovery and treatment-seeking in pathological gambling: Results of two US national surveys. Am J Psychiatry. 2006;163:297–302. doi: 10.1176/appi.ajp.163.2.297. [DOI] [PubMed] [Google Scholar]
- 8.Slutske WS, Blaszczynski A, Martin NG. Sex differences in the rates of recovery, treatment-seeking, and natural recovery in pathological gambling: Results from an Australian community-based twin survey. Twin Res Hum Genet. 2009;12:425–432. doi: 10.1375/twin.12.5.425. [DOI] [PubMed] [Google Scholar]
- 9.Slutske WS, Meier MH, Zhu G, Statham DJ, Blaszczynski A, Martin NG. The Australian twin study of gambling (OZ-GAM): Rationale, sample description, predictors of participation, and a first look at sources of individual differences in gambling involvement. Twin Res Hum Genet. 2009;12:63–78. doi: 10.1375/twin.12.1.63. [DOI] [PubMed] [Google Scholar]
- 10.Gerstein D, Volberg R, Murphy S, Toce M, et al. Gambling Impact and Behavior Study: Report to the National Gambling Impact Study Commission. New York: Christiansen/Cummings Associates; 1999. [Google Scholar]
- 11.American Psychiatric Association. Diagnostic and statistical manual of mental disorders: DSM-IV. 4. Washington, D.C.: American Psychiatric Association; 1994. [Google Scholar]
- 12.Wickwire EM, Burke RS, Brown SA, Parker JD, Ryan KM. Psychometric evaluation of the National Opinion Research Center DSM-IV Screen for Gambling Problems (NODS) Am J Addict. 2008;17:392–395. doi: 10.1080/10550490802268934. [DOI] [PubMed] [Google Scholar]
- 13.Hodgins D. Using the NORC DSM Screen for Gambling Problems as an outcome measure for pathological gambling: Psychometric evaluation. Addict Behav. 2004;29:1685–1690. doi: 10.1016/j.addbeh.2004.03.017. [DOI] [PubMed] [Google Scholar]
- 14.Lesieur HR, Blume SB. The South Oaks Gambling Screen (SOGS): A new instrument for the identification of pathological gamblers. Am J Psychiatry. 1987;144:1184–1188. doi: 10.1176/ajp.144.9.1184. [DOI] [PubMed] [Google Scholar]
- 15.Kessler RC, Hwang I, LaBrie R, Petukhova M, Sampson NA, Winters KC, et al. DSM-IV pathological gambling in the National Comorbidity Survey Replication. Psychol Med. 2008;38:1351–1360. doi: 10.1017/S0033291708002900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Petry NM, Stinson FS, Grant BF. Comorbidity of DSM-IV pathological gambling and other psychiatric disorders: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry. 2005;66:564–574. doi: 10.4088/jcp.v66n0504. [DOI] [PubMed] [Google Scholar]
- 17.Welte J, Barnes G, Wieczorek W, Tidwell M, Parker J. Alcohol and gambling pathology among U.S. adults: Prevalence, demographic patterns, and comorbidity. J Stud Alcohol. 2001;62:706–12. doi: 10.15288/jsa.2001.62.706. [DOI] [PubMed] [Google Scholar]
- 18.Walker M, Toneatto T, Potenza MN, Petry N, Ladouceur R, Hodgins DC, et al. A framework for reporting outcomes in problem gambling treatment research: The Banff, Alberta Consensus. Addiction. 2006;101:504–11. doi: 10.1111/j.1360-0443.2005.01341.x. [DOI] [PubMed] [Google Scholar]
- 19.Marlatt GA. The controlled-drinking controversy: a commentary. Am Psychol. 1993;38:1097–110. doi: 10.1037//0003-066x.38.10.1097. [DOI] [PubMed] [Google Scholar]
- 20.Sobell MB, Sobell LC. Controlled drinking after 25 years: How important was the great debate? Addiction. 1995;90:1149–53. [PubMed] [Google Scholar]
- 21.Saladin ME, Santa Ana EJ. Controlled drinking: More than just a controversy. Curr Opin Psychiatry. 2004;17:175–87. [Google Scholar]
- 22.Wilbourne PL, Miller WR. Treatment for alcoholism: Older and Wiser? Alcohol Treat Q. 2002;20:41–59. [Google Scholar]
- 23.Hester RK, Miller WR. Handbook of alcoholism treatment approaches. 3. Boston, MA: Allyn and Bacon; 2003. [Google Scholar]
- 24.Miller WR, Page AC. Warm turkey: other routes to abstinence. J Subst Abuse Treat. 1991;8:227–334. doi: 10.1016/0740-5472(91)90043-a. [DOI] [PubMed] [Google Scholar]
- 25.Fagerstrom KO. Can reduced smoking be a way for smokers not interested in quitting to actually quit? Respiration. 2005;72:216–20. doi: 10.1159/000084057. [DOI] [PubMed] [Google Scholar]
- 26.Hughes JR, Carpenter MJ. Does smoking reduction increase future cessation and decrease disease risk? A qualitative review. Nicotine Tob Res. 2006;8:739–49. doi: 10.1080/14622200600789726. [DOI] [PubMed] [Google Scholar]
- 27.Dowling N, Smith D. Treatment goal selection for female pathological gambling: A comparison of abstinence and controlled gambling. J Gambl Stud. 2007;23:335–345. doi: 10.1007/s10899-007-9064-6. [DOI] [PubMed] [Google Scholar]
- 28.Rumpf H, Bischof G, John U. Remission without formal help: new directions in studies using survey data. In: Klingemann H, Sobell LC, editors. Promoting self-change from addictive behaviors: practical implications for policy, prevention, and treatment. New York: Springer; 2007. pp. 73–85. [Google Scholar]
- 29.Rumpf H, Bischof G, Hapke U, Meyer C, John U. Studies on natural recovery from alcohol dependence: sample selection bias by media solicitation? Addiction. 2000;95:765–75. doi: 10.1046/j.1360-0443.2000.95576512.x. [DOI] [PubMed] [Google Scholar]