Dr. DeLuca's Addiction Website

Humphrey's Study of MM membership

The present study This study evaluates whether MM's novel features affect its accessibility. Despite their lower levels of dependence and consumption, problem drinkers are an important focus of intervention because they account for a larger share of alcohol-related harm to society (e.g., domestic violence, alcohol-related auto fatalities) than do alcohol-dependent individuals (Sobell, Cunningham, & Sobell, 1996).

The other, extremely important reason for assessing who attends MM is to evaluate whether the organization is -- despite its stated purposes --attracting severely dependent individuals for whom moderate drinking is very difficult or impossible to maintain. If this is so, MM may have little positive value and might even potentiate harm.

This latter concern was highlighted this year by two widely-discussed events. In January, MM founder Audrey Kishline, announced to the membership that after 6 years of affiliation she no longer felt able to moderate her drinking and was leaving the organization to attend AA. Six months later, while legally intoxicated, she drove her truck down the wrong side of a highway and struck another vehicle head-on, killing both occupants. During the subsequent media firestorm, many alcohol researchers and clinicians expressed strong, divergent opinions about whether this heartbreaking event demonstrated that MM is primarily, in colloquial terms, a "dangerous temptation to alcoholics".

Here, we offer an empirical assessment of this important issue based on comprehensive data on MM members. In the addition to evaluating the characteristics of MM as a whole, we also compare the characteristics of members who access the organization over the internet versus face-to-face. Determining how the internet influences the accessibility of alcohol-related interventions could have broad import for understanding how this rapidly expanding medium can be used to promote public health. Internet-based interventions are relatively inexpensive to operate. If MM internet-based groups reach significant numbers of individuals who have alcohol problems, other alcohol treatment and self-help organizations may wish to make more extensive use of this potentially cost-effective intervention. In addition, internet-based groups may be more accessible to hard-to-reach populations, such as individuals who live in areas where services are not physically available, have poor access to transportation, or are severely disabled (Kurtz, 1997). Although many self-help organizations for a variety of health problems have started internet-based groups (Madara, 1999), the proposed study will be among the first to describe the nature of these groups and evaluate who accesses them. 


Participants Participants were 177 individuals attending MM meetings for alcohol problems (i.e., treatment professionals and students attending to learn about MM were excluded). About a third of respondents (n = 63) attended face-to-face MM meetings only, and about a fourth (n= 42) attended both face-to-face and on-line meetings. The remainder (n = 73) was involved in MM only over the internet.


August of 1999, all current MM meeting attenders received a cover letter co-signed by the investigators and MM's president. The letter, which described the study and requested participation, was accompanied by a survey. MM received a $20 donation for each survey returned to the investigators, including surveys returned blank by face-to-face meeting attenders who did not wish to participate. This latter procedure was implemented to avoid social pressure on members to participate in the study. Distribution of paper and pencil version: All 12 active face-to-face MM groups were mailed a packet of surveys and accompanying letters, along with a personal letter to the group leader. The group leader was asked to distribute surveys to all attendees at the next two meetings and return them to the project team in a stamped, pre-addressed envelope (including refusals, which each member could covertly indicate in a designated space in the survey). At face-to-face groups, 89 of 99 (89%) MM attendees completed the survey.

Distribution of electronic version:

The cover letter was posted electronically to MM internet groups with a hyperlink to a web page on which members could complete the survey on-line. A reminder message was posted two weeks later. A total of 88 surveys were completed on-line. The participation rate for the electronic version is difficult to assess because group membership cannot be precisely determined on-line, and because professionals and students not eligible for the survey were enrolled on MM's electronic discussion groups. However, based on server statistics and the content of posted messages, it appeared that approximately 200 individuals were enrolled in the MM on-line group during the survey period, of whom 160 were eligible for the survey and had not already completed it at a face-to-face group meeting. This results in a participation rate of 55% for the electronic version and 68% (177 of 259) for the study overall. In interpreting the lower response rate to the electronic version, it is worth noting that on-line MM groups generate dozens of pages of text per day and therefore many members likely did not see the announcements of the survey, rather than saw it and refused to participate.

Because members participating in both face-to-face and on-line groups were likely to receive the survey and letter twice (once in each forum), they were specifically asked to complete it only once. Inspection of gathered data revealed no duplicates, indicating that members complied with this request. Measures The survey assessed all demographics (e.g., sex, age, race) with single items, except for religiosity. Belief about God was assessed using an item from the Religious Background and Behavior questionnaire (Connors et al., 1996), and participants were also asked to report their frequency of religious service attendance. A variety of items tapped alcohol use and problems in the six months prior to MM involvement. Alcohol dependence symptoms were measured using 11 items (response range from 0 = never to 4 = often, alpha = .84) from the Alcohol Dependence Scale (Skinner & Allen, 1982). Alcohol-related problems (response range from 0 = never to 4 = often, alpha = .79) such as difficulties with health, work, and family, were assessed using a 9-item scale from the Health and Daily Living Form (Moos et al., 1992). Respondents also reported how many Days per month they were drunk or intoxicated, and rated the perceived severity of their drinking problem on a scale ranging from 1 for "no problem" to 5 for "serious problem". Frequency and typical amount of alcohol consumption was assessed using two items from the Alcohol Use Disorders Identification Test (Babor, de la Fuente, Saunders, & Grant, 1992). Respondents also were asked whether their current drinking goal was abstinence or moderation. Respondents reported the length and intensity of their involvement in MM, and whether they participated in face-to-face groups only, internet-based groups only, or both. Respondents also reported their lifetime use of helping resources, such as professional alcohol treatment, AA and other self-help groups, individual psychotherapy, and psychotropic medication. Those who used internet-based MM groups were asked their reasons for using on-line MM resources. Because our pilot research with MM members had revealed a very low prevalence of use of illicit drugs, such behavior was assessed with less specificity than was alcohol consumption. Participants were asked a single, general question about the presence of any drug-related problems in the six months prior to MM participation, and then were asked single questions about any use in the past 30 days of heroin, cocaine/crack, and amphetamines / methamphetamines.

Analysis plan:

To address the central questions of the study, the demographic characteristics of all MM members will first be described. Then, the characteristics of MM members who accessed the organization in face-to-face groups only (FTF-Only), on-line only (OL-Only) and in both forums (OL&FTF) will be compared.


Demographic Findings


MM participants differed substantially from most help-seeking samples of problem drinkers.

1. First, given gender base rates for alcohol problems (Gomberg, 1982), an unusually high proportion (49.1%) of members were women.
2.   MM participants were high on indicators of socioeconomic stability and privilege: Most were white (95.5%), employed (81.4%), married (58.3%), college-educated (71.6%), and of early middle age (49.4%).
3. Relative to the U.S. general population (Gallup organization, 1999), participants were extremely secular, with almost a third describing their religious beliefs as "atheist, agnostic, or unsure" and only a sixth attending religious services on a weekly or more frequent basis.

Chi-square analyses indicated significant differences between different types of MM participants.


Specifically, a remarkable 63.0% of OL-Only members were female, compared to about 40.0% of FTF-Only members and 38.1% of OL&FTF members.

Educational differences were also identified


Higher levels of education are associated with on-line use. Most notably, almost half of OL-only members and almost two-thirds of OL&FTF members had post-graduate education.

Analysis of members' age yielded an intriguing significant difference.


OL-only members comprised a higher proportion of adults under the age of 35 (33.8%), whereas OL&FTF members comprised the highest proportion of individuals over the age of 50 (38.1%).

Significant differences on religious beliefs exist within MM


Belief in God is highest among FTF-Only members, somewhat lower among OL-Only members, and sharply lower among FTF&OL members. A similar pattern of results was observed for religious service attendance, but did not attain statistical significance (?2 = 7.35, 4 df, p =.12).


Alexander DeLuca, M.D., FASAM.
Copyright 1999. All rights reserved.                         [Top of Page]
Revised: June 9, 2001.