660 JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / JULY 2011
Six-Month Changes in Spirituality and Religiousness in
Alcoholics Predict Drinking Outcomes at Nine Months*
ELIZABETH A. R. ROBINSON, PH.D.,† AMY R. KRENTZMAN, PH.D., JON R. WEBB, PH.D.,† AND KIRK J. BROWER, M.D.
University of Michigan Addiction Research Center, 4250 Plymouth Road, Ann Arbor, Michigan 48109
ABSTRACT. Objective: Although spiritual change is hypothesized to
contribute to recovery from alcohol dependence, few studies have used
prospective data to investigate this hypothesis. Prior studies have also
been limited to treatment-seeking and Alcoholics Anonymous (AA)
samples. This study included alcohol-dependent individuals, both in
treatment and not, to investigate the effect of spiritual and religious (SR)
change on subsequent drinking outcomes, independent of AA involve-
ment. Method: Alcoholics (N = 364) were recruited for a panel study
from two abstinence-based treatment centers, a moderation drinking
program, and untreated individuals from the local community. Quantita-
tive measures of SR change between baseline and 6 months were used
to predict 9-month drinking outcomes, controlling for baseline drinking
and AA involvement. Results: Signifi cant 6-month changes in 8 of 12
SR measures were found, which included private SR practices, beliefs,
daily spiritual experiences, three measures of forgiveness, negative
religious coping, and purpose in life. Increases in private SR practices
and forgiveness of self were the strongest predictors of improvements
in drinking outcomes. Changes in daily spiritual experiences, purpose
in life, a general measure of forgiveness, and negative religious coping
also predicted favorable drinking outcomes. Conclusions: SR change
predicted good drinking outcomes in alcoholics, even when controlling
for AA involvement. SR variables, broadly defi ned, deserve attention
in fostering change even among those who do not affi liate with AA or
religious institutions. Last, future research should include SR variables,
particularly various types of forgiveness, given the strong effects found
for forgiveness of self. (J. Stud. Alcohol Drugs, 72, 660–668, 2011)
and Alcoholism Grants R01 AA014442 and T32 AA007477-21. Portions of
this research were presented at the 2008 Research Society on Alcoholism
meeting, June 27–July 2, Washington, DC.
dress or via email at: email@example.com. Jon R. Webb is with East
Tennessee State University, Department of Psychology, Rogers-Stout Hall,
Johnson City, TN.
Received: February 12, 2010. Revised: February 28, 2011.
*This research was supported by National Institute on Alcohol Abuse
†Correspondence may be sent to Elizabeth A. Robinson at the above ad-
treatment were previously described (Robinson et al., 2007).
Signifi cant changes were found in 5 of 10 SR dimensions,
and positive change in two (daily spiritual experiences and
purpose in life) were associated with no heavy drinking
at 6 months, controlling for Alcoholics Anonymous (AA)
involvement and gender. However, the sample in this earlier
study was drawn from one site; included individuals with
abuse as well as dependence; and did not include untreated
individuals, who make up the bulk of those with alcohol use
disorders in this country (Cohen et al., 2007; Dawson et al.,
2005, 2007). In addition, the analyses in our 2007 article
investigated only the relationship of 6-month SR change to
6-month drinking, not to subsequent drinking, which is a
more powerful test of the impact of SR change. The longitu-
dinal survey described here obtained SR data at baseline and
6 months as well as drinking data at baseline and 9 months,
allowing us to test whether 6-month SR change predicts
subsequent drinking. The sample of 364 alcohol-dependent
individuals was drawn from four sources—two outpatient
IX-MONTH CHANGES IN spirituality and religious-
ness (SR) in patients with alcohol use disorders entering
treatment programs, a moderated drinking program, and
untreated alcoholics recruited from the larger community.
Since our earlier article (Robinson et al., 2007), other
studies have investigated this issue, all recruiting from
treatment programs or AA. Three described SR change in
alcoholics in treatment (Piderman et al., 2007, 2008; Sterling
et al., 2007). Piderman and colleagues’ longitudinal survey
of 74 alcoholics found increases in spiritual well-being, pri-
vate religious practices, and positive religious coping from
intake to discharge (2007). At 1-year follow-up, those who
had achieved 1 year of abstinence were compared with those
who had not, combined with those lost to follow-up (2008).
Increases from intake to discharge in private SR practices
(e.g., prayer, reading) and existential well-being were asso-
ciated with sobriety. Sterling and colleagues (2007) divided
a treatment sample at 3-month follow-up into two matched
groups: those who had achieved 1 month of sobriety and
those who had not. SR was measured at intake, discharge,
and 3-month follow-up. Although the SR of both groups in-
creased from baseline to discharge, at the 3-month follow-up
individuals who relapsed had signifi cantly lower scores on
two SR measures than those who did not.
Two studies investigated spiritual change in cross-section-
al, retrospective surveys of AA members. Poage et al. (2004)
found that length of sobriety correlated with a spirituality
measure but not with contentment or stress. Sandoz (1999)
asked 56 members of AA about spiritual experiences. Those
who had such experiences (n = 46) were older and reported
longer sobriety, working more AA steps, and providing AA-
ROBINSON ET AL. 661
alcohol-dependence treatment and found similar results
(Carrico et al., 2007; Flynn et al., 2003; Heinz et al., 2007;
Jarusiewicz, 2000; Sherman and Fischer, 2002; Stewart and
The studies described above provide additional evidence
that SR change may be a factor in recovery. Although longi-
tudinal evidence has confi rmed cross-sectional fi ndings, the
length of follow-up and the size of the samples are often lim-
ited. Some investigators have measured SR only at baseline,
although there is clear evidence that it changes over time. In
addition, many investigators used a nonspecifi c measure of
spirituality, making it diffi cult to determine its meaning and
which SR dimensions are most likely to change and support
recovery. A variety of dimensions may be involved, including
SR practices, forgiveness, existential concerns, and day-to-
day experiences of SR.
All of these samples were drawn from treatment centers
or AA, where one would expect to see SR changes, given the
spiritual emphasis of 12-step approaches. We do not know
whether SR changes in non-treatment-seeking alcoholics,
even though most alcoholics are not in treatment (Cohen et
al., 2007)—and many alcoholics, whether treated or not—
reduce their drinking over the course of a year (Dawson et
al., 2005, 2007). Therefore, it would be useful to examine
whether SR changes predict drinking outcomes, without
regard to AA involvement and treatment.
Based on these earlier studies, the hypotheses investigated
here are that in a large sample of alcoholics (a) increases
would be found from baseline to 6 months in day-to-day
spiritual experiences, private SR practices, sense of meaning/
purpose in life, use of positive SR coping, and forgiveness;
(b) 6-month changes would not be found in beliefs, values,
use of negative SR coping, and perceptions of God; and (c)
6-month changes in daily spiritual experiences and meaning
in life would be associated with less drinking at 9-month
follow-up, after controlling for AA involvement.
Another group of studies used samples from drug- and
alcohol dependence was conducted, with data on SR, drink-
ing, AA involvement, treatment, and other variables collected
every 6 months. At the intervening 3-month point, data on
drinking, AA, and treatment were obtained.
A naturalistic longitudinal survey of 364 individuals with
Design and procedure
treatment program (UOT; n = 157), a Veterans Affairs out-
patient treatment clinic (VA; n = 80), a moderated drinking
program (MOD; n = 34), and the larger community sample
(CS; n = 93). The CS participants were not in treatment.
The design was a descriptive 3-year panel study follow-
Recruitment sites were a university-affi liated outpatient
ing individuals with alcohol dependence to determine how
changes in drinking, dimensions of SR, and recovery efforts
related to each other. As approved by appropriate institu-
tional review boards, all subjects provided written informed
consent and were provided compensation for each in-person
Recruitment criteria were (a) being more than 18 years
of age; (b) having been diagnosed by the Structured Clinical
Interview for DSM-IV (SCID; First et al., 1997) with a life-
time diagnosis of alcohol dependence; (c) having consumed
alcohol in the last 90 days; (d) not suicidal, homicidal, or
psychotic; and (e) being literate in English. Those recruited
from treatment settings were approached after 1 week but
not after 4 weeks in treatment (i.e., after detoxifi cation and
identifi ed by clinical record review and then approached by
research staff. The MOD program coordinator identifi ed
potential respondents who were willing to be contacted.
We then screened potential participants by phone with the
Rapid Alcohol Problems Screen (RAPS; Cherpitel, 1995), a
fi ve-item alcoholism assessment tool. The RAPS questions
address drinking on awakening, blackouts, guilt/remorse
after drinking, failing to meet expectations, and losing
friends because of drinking. Our fi nal subsample (the CS
group) was recruited from untreated individuals in the com-
munity through advertisements in local print media. When
participants called us, we described the study to them, and,
if they were interested, they were screened with the RAPS
(Cherpitel, 1995). All baseline interviews began with the
alcohol-dependence portion of the SCID (First et al., 1997)
to confi rm diagnoses. Across all sites, 469 individuals were
approached by us or contacted us, of whom 364 (77.6%)
met criteria and agreed to participate. Response rates and
demographic, clinical, and drinking variables varied by site.
Potential study participants from UOT and VA sites were
acteristics. Two thirds were male (66%), the mean age was
44 years, and the mean amount of education was more than
14 years. More than a third were married or cohabitating
with a partner, almost 30% had never married, and about
one third were no longer married. The sample’s ethnicity
refl ected the local community, with 82% of the subjects of
White ethnicity and 10% of Black ethnicity; the remainder
were of other ethnicities, including self-identifi ed multi-
ethnicities. More than half were employed full time or part
time. Of the 44% unemployed, 24% were retirees, 18% were
disabled, and 11% were students. Almost 30% of the sample
had an annual income less than $15,000, although 22%
Table 1 presents baseline demographic and clinical char-
662 JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / JULY 2011
made more than $85,000. Comparing this sample with that
of our previous study (Robinson et al., 2007), this one was
somewhat older and more educated, and it included fewer
African Americans. Compared with the National Epidemio-
logic Survey on Alcohol and Related Conditions (NESARC)
nationally representative sample of alcoholics (Dawson et
al., 2005), this sample appeared similar in age, male/female
ratio, and education, but fewer subjects were married.
Regarding clinical characteristics, 57% of participants
had severe dependence (six to seven DSM criteria; SCID;
First et al., 1997). More than half had been in treatment be-
fore study entry. Age at onset was mid-20s; more than 85%
reported a family history of alcohol problems. Scores on a
measure of consequences of alcohol problems, the Short
Inventory of Problems (SIP; Forcehimes et al., 2007; Miller
et al., 1995), indicated a moderate level of alcohol problems
(M = 21.0). Almost three quarters (72%) stated that they
wanted to be abstinent. Three quarters reported having AA
experience, and a smaller percentage (67%) had attended a
meeting. Again, informally comparing this sample with the
NESARC sample (Dawson et al., 2005), this sample had a
slightly older age at onset and higher rates of family history
of alcohol problems.
Retention and attrition analysis
terview, 316 (86.8%) completed the 6-month interview. At
9 months, valid drinking data were collected on 283 respon-
dents (77.7%). A 6-month attrition analysis (Menard, 1991)
showed no signifi cant differences on demographic variables
in Table 1 among those retained and not retained. Of the
clinical variables, only SIP scores differed signifi cantly, with
nonparticipants at 6 months having higher SIP scores. A sec-
ond attrition analysis with 9-month data found that, among
demographics, only education varied signifi cantly (14.5 years
retained vs. 13.7 not retained). We also found differences on
four clinical variables: age at fi rst symptoms (29.2 years vs.
26.0, respectively), SIP scores (20.2 vs. 23.5), experience
with AA (72% vs. 84%), and prior AA meeting attendance
(64% vs. 78%). Those retained had fewer drinks per drink-
ing day (DDD; 9.01 drinks vs. 11.4). This pattern of attrition
suggests that the 9-month data may include the less severe
alcoholics. However, because there were no differences in
SCID severity, physiological dependence, and wanting to be
abstinent, we are fairly confi dent that this analysis provides
useful information on changes in SR and drinking among
Of the 364 participants who completed the baseline in-
baseline and at 6 months. Internal reliability estimates from
our sample are presented where appropriate.
Alcohol use and consequences. Every 3 months, data on
alcohol use during the prior 90 days were obtained with the
Timeline Followback (TLFB) interview (Sobell and Sobell,
1992; Sobell et al., 1996). From the TLFB, we calculated
percentage of days abstinent (PDA), percentage of heavy
drinking days (PHDD), average number of DDD, and num-
ber of days since last drink (DSLD). Heavy drinking was
defi ned as fi ve or more standard drinks per day for men or
four or more for women.
Negative consequences of alcohol use were assessed with
the SIP (Miller et al., 1995), which had Cronbach’s α coef-
fi cients of .93 and .95 at baseline and 6 months, respectively.
Spirituality and religiousness. SR instruments were
drawn from the psychology of religion literature and prior
research (Robinson et al., 2007). Using the same defi nitions
of SR previously developed (Robinson et al., 2007), we op-
All measures, except drinking data, were obtained at
TABLE 1. Demographics and clinical characteristics at baseline (N = 364)
Variable % or M (SD)
Age, in years
Education, in years
Separated, divorced, widowed
Prior alcohol treatment? (% yes)
Mild, 3–4 symptoms
Moderate, 5 symptoms
Severe, 6–7 symptoms
Age at fi rst alcohol problems
Family history of alcohol problems
Want to be abstinent? (% yes)
Prior AA experience?
Ever attended an AA meeting?
Attended AA meeting in last year?
Notes: Percentages on income may not add up to 100% because of miss-
ing data. All respondents have SCID-verifi ed lifetime alcohol-dependence
diagnoses. SCID = Structured Clinical Interview for DSM-IV; SIP = Short
Inventory of Problems; AA = Alcoholics Anonymous.
ROBINSON ET AL. 663
erationalized SR on several dimensions, drawing heavily on
the monograph from the Fetzer Institute/National Institute
on Aging (1999). The SR dimensions from that report used
here are the following: private practices; day-to-day spiritual
experiences; meaning, values, and beliefs; forgiveness; and
religious coping. We also added perceptions of God, beliefs,
sense of purpose/meaning in life, and a more robust measure
of forgiveness. Higher scores on all SR variables indicate
Perceptions of God were assessed with the Loving and
Controlling God scales (Benson and Spilka, 1973), two
fi ve-item semantic differential scales (0–6) of perceptions of
God. The Loving God scale had Cronbach’s α’s of .79 and
.78 at baseline and 6 months, respectively. The reliability of
the Controlling God scale was more marginal, with α’s of
.69 and .70.
Beliefs were measured with the fi rst item of the Religious
Background and Behaviors scale (Connors et al., 1996)
used in Project MATCH (Matching Alcoholism Treatments
to Client Heterogeneity), which assesses belief in God and
practice of religion. The baseline mean on this 5-point scale
was 3.8 (SD = 1.2) (3 = I don’t know what to believe about
God and 4 = I believe in God, but I’m not religious), which
was congruent with other indicators (e.g., religious prefer-
ence, congregational involvement, self-ranking of one’s
religiousness and spirituality) that respondents are generally
not religious, although most believe in God.
Private SR practices were assessed with a fi ve-item scale
from the Fetzer Institute/National Institute on Aging mono-
graph (1999), measuring the frequency of prayer, meditation,
scripture reading, or other private SR behaviors. The α’s for
this scale were .77 at both baseline and 6 months.
Daily Spiritual Experiences (Underwood and Teresi,
2002) measures such day-to-day experiences as a sense of
connection with God; receiving strength, comfort, and love
from God; experiences of peacefulness and awe; and a long-
ing for closeness with God. Participants responded to 16
items on a 6-point scale ranging from never or almost never
to many times a day. Baseline and 6-month α’s were .94 and
Six items from the Fetzer Institute/National Institute on
Aging (1999) measured values and beliefs, such as “I have
a sense of mission or calling in my own life,” “I feel a deep
sense of responsibility for reducing pain and suffering in
the world,” and “I believe in a God who watches over me.”
Baseline and 6-month α’s were .83 and .81, respectively.
Forgiveness was measured with three scales to ensure ro-
bust measurement of a factor that fi gures largely in writings
on recovery (AA, 1976; Kurtz and Ketcham, 1992; Webb
and Trautman, 2010). The three-item Forgiveness Scale
from the Fetzer Institute/National Institute on Aging (1999)
measures forgiving others, forgiving one’s self, and feeling
forgiven by God on a four-point scale ranging from never
to almost always. Baseline and 6-month α’s were both low
(.46)—which is not surprising, given the small number of
items and the disparate domains measured.
In addition, the Mauger scales (Mauger et al., 1992)
assess forgiveness of self and of others. These two 15-
item scales use dichotomous (true/false) response options.
Forgiveness of self had α’s of .83 and .77 at baseline and 6
months, respectively, whereas forgiveness of others had α’s
of .77 and .75.
The use of positive and negative religious coping strate-
gies in stressful situations was measured with an adaptation
of Pargament et al.’s (1998) Brief RCOPE. Positive religious
coping refl ects a secure relationship with God, belief that
life has meaning, sense of spiritual connectedness to oth-
ers, benevolent reappraisals, collaborative religious coping,
seeking spiritual support, and connection to God and oth-
ers. Negative religious coping refl ects feeling punished or
abandoned by God. Response options on 16 items ranged
on a 4-point scale from not at all to a great deal. Positive
religious coping had baseline and 6-month α’s of .93 and
.95, and negative religious coping had baseline and 6-month
α’s of .74 and .72, respectively.
Existential meaning/purpose was measured with Crum-
baugh and Maholick’s (1964) Purpose in Life scale, used in
previous studies (Robinson et al., 2007; Waisberg and Por-
ter, 1994). Based on Frankl’s existential perspective (1969,
1992), this scale assesses the degree to which an individual
has a sense of meaning or signifi cance in his or her life. This
20-item measure has 7-point Likert response scales. In this
sample, α’s were .88 at baseline and .91 at 6 months.
Involvement in Alcoholics Anonymous. The AA Involve-
ment (AAI) scale (Tonigan et al., 1996), which includes
attendance data (lifetime and past year) and an involve-
ment subscale, was adapted for this study. Two items that
overlapped with our follow-up periods were excluded. The
modifi ed AAI scale used in this study had six items, with
yes/no responses on AA activities such as having a spon-
sor, providing service, sense of being a member of AA, and
celebrating a sobriety birthday. Baseline and 6-month α’s for
the modifi ed AAI subscale were .81 and .74. At baseline, the
mean score on this measure was 1.83 (SD = 2.0), and at 6
months it was 1.81 (SD = 1.9). A paired sample t test indi-
cated that this change was not signifi cant.
determine whether assumptions of normality in parametric
statistical tests would be violated. Two drinking variables at
9 months, PHDD and DDD, were suffi ciently skewed that
we transformed them into dichotomous variables (e.g., no
drinking vs. any). Our analyses next proceeded in two stages:
(a) testing for 6-month changes in SR dimensions and (b)
whether those SR changes predicted 9-month drinking. We
examined the signifi cance of changes in SR variables from
The distributions of outcome variables were examined to
664 JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / JULY 2011
baseline to 6 months using paired sample t tests. To obtain
an estimate of effect size, Cohen’s d was calculated (Cohen,
1988, 1992) on each variable changing signifi cantly over 6
months (p ≤ .05) or nearly signifi cantly (p < .10). These SR
variables were then individually tested as predictors of each
drinking variable at 9 months in a series of multiple linear
regression or multiple logistic regression analyses, control-
ling for that baseline drinking variable and AAI. Baseline
AAI was a covariate because of strong relationships with
both SR and outcome variables and the lack of signifi cant
change in AAI from baseline to 6 months. Given the de-
scriptive nature of the study, Bonferonni corrections were
not used. Our analytic emphasis instead was on identifying
patterns of signifi cant relationships across drinking and SR
Drinking data at baseline and 9 months
(SD = 31.3), mean PHDD was 32.7% (SD = 29.8), and mean
DDD was 9.5 drinks (SD = 8.2). Mean DSLD was 25.4 days
(SD = 27.1). Drinking data at 9 months indicated signifi -
cant improvement (p ≤ .05) on all drinking variables. At 9
months, PDA in the last 90 days was 80.7 (SD = 31.0), with
45.6% of the sample abstinent during that period. PHDD
was 8.8% (SD = 21.0), and DDD was 3.1 (SD = 4.3). Mean
DSLD was 130.6 days (SD = 143.6).
At baseline, mean PDA in the last 90 days was 56.1%
Six-month changes in spirituality and religiousness
deviations; ns; p values; and, for those variables that changed
signifi cantly or approached signifi cance, Cohen’s d values
(Cohen, 1988, 1992) to indicate effect size. Variables that
changed signifi cantly from baseline to 6 months are bolded.
Eight of 12 SR variables changed signifi cantly. They were
beliefs, private SR practices, daily spiritual experiences, the
three-item Fetzer Institute/National Institute on Aging for-
giveness measure, both Mauger measures (forgiveness of self
and of others), negative religious coping, and purpose in life.
As Cohen’s d indicates, these effects were small. Changes
in forgiveness of self and negative religious coping had the
largest effect sizes, which foreshadowed the following fi nd-
ings on their relationship to drinking outcomes.
Table 2 presents baseline and 6-month means; standard
Linear regressions predicting 9-month drinking
linear regressions predicting PDA and DSLD, from 6-month
change in an SR variable, controlling for baseline AAI and
drinking (PDA and DSLD, respectively). Both PDA and
Table 3 presents the results of the individual multiple
DSLD were predicted by increases in private SR practices,
daily spiritual experiences, forgiveness of self, and purpose
in life. Changes in three SR variables (Fetzer forgiveness,
forgiveness of others, and negative religious coping) pre-
dicted PDA or DSLD. Changes in these SR variables were
associated with favorable change in drinking outcomes, with
the notable exception of forgiveness of others, which was
negatively associated with PDA. Note that a drop in nega-
tive religious coping was favorable. Predictive relationships
were not found for the SR variables of beliefs and meaning,
values, and beliefs.
Logistic regressions predicting 9-month drinking
logistic regressions predicting PHDD and DDD. The table
includes the signifi cant and marginally signifi cant odds ra-
tios. Both PHDD and DDD were predicted by increases in
private SR practices and forgiveness of self and decreases
in negative religious coping. The odds of any PHDD were
also lower with increases in the Fetzer forgiveness scale
and purpose in life. Increases in daily spiritual experiences
signifi cantly predicted reduced odds of any drinking (i.e.,
DDD). Again, no signifi cant relationships were found in
these logistic regressions for beliefs and for meaning, values,
and beliefs, as well as for forgiveness of others.
Table 4 presents the results of the individual multiple
TABLE 2. SR variables at baseline and 6 months: Means, standard devia-
tions, p values from p a i red samples t test, and Cohen’s d
Scale and range
of possible scores
M (SD) M (SD)
n p d
Loving God scale,
Controlling God scale,
Private SR practices,
Daily spiritual experiences,
Meaning, values, beliefs,
Mauger forgiveness of self,
Mauger forgiveness of others, 10.26
Positive religious coping,
Negative religious coping,
Purpose in life,
269 .406 . .–
270 .696 . .–
316 .032 0.075
316 .000 0.130
316 .000 0.126
316 .056 0.081
316 .010 0.105
316 .000 0.263
316 .000 0.166
316 .000 -0.256
313 .000 0.178
Notes: Differences between time points analyzed with paired-sample t tests.
d = Cohen’s d values (1988, 1992) provided for SR variables with sig-
nifi cant change. SR = spirituality and religiousness. Variables that changed
signifi cantly from baseline to 6 months are highlighted in bold.
ROBINSON ET AL. 665
baseline to 6 months) or desire to be abstinent as control
variables did not alter the basic nature of these results, al-
though the signifi cance of some dropped to the trend level,
particularly with the logistic analyses.
Adding treatment exposure (i.e., treatment days from
many SR dimensions changing in alcoholics over the course
of 6 months, although the specifi c SR dimensions differed
somewhat from prior fi ndings (Robinson et al., 2007). We
found signifi cant change in 8 of 12 quantitative measures of
SR, specifi cally beliefs, daily spiritual experiences, private
SR practices, forgiveness (overall, of self, and of others),
negative religious coping, and purpose in life or sense of
meaning. Signifi cant changes were not found in perceptions
of God and positive religious coping, although margin-
ally signifi cant changes were found in meaning, values, and
Furthermore, 6-month changes in several of these SR
variables were predictive of 9-month drinking outcomes,
controlling for baseline AA involvement and drinking. In-
creases in forgiveness of self and private SR practices were
the most consistent predictors, predicting all four drinking
variables at 9 months. Increases in daily spiritual experi-
ences and purpose in life and decreases in negative reli-
gious coping signifi cantly predicted three of four drinking
outcomes. Increases in the brief global measure of forgive-
ness (Fetzer Institute/National Institute on Aging, 1999)
predicted two outcomes; however, increases in forgiveness
of others negatively predicted only one outcome. Increases
in beliefs and meaning, values, and beliefs did not predict
In a larger and more diverse sample, we again found
TABLE 3. Signifi cant multiple linear regressions predicting 9-month per-
centage of days abstinent (PDA) and mean days since last drink (DSLD)
from 6-month change in each spirituality and religiousness (SR) variable,
controlling for baseline drinking and baseline Alcoholics Anonymous In-
in last 90 days DSLD
b p b p
Private SR practices
Daily spiritual experiences
Meaning, values, beliefs
Mauger forgiveness of self
Mauger forgiveness of others
Negative religious coping
Purpose in life
Notes: Analyses were conducted only for those SR variables that changed
signifi cantly at p < .10. Data for multivariable models signifi cant at p ≤ .05
are in bold. b = unstandardized b.
TABLE 4. Signifi cant multiple logistic regressions predicting 9-month
dichotomized percentage heavy drinking days (PHDD) and mean drinks
per drinking day (DDD) from 6-month change in each spirituality and
religiousness (SR) variable
in last 90 days in last 90 days
Private SR practices
Daily spiritual experiences
Meaning, values, beliefs
Mauger forgiveness of self
Mauger forgiveness of others
Negative religious coping
Purpose in life
– – – –
Notes: Controlling for baseline drinking variable and baseline Alcoholics
Anonymous Involvement scale. Analyses were conducted only for those SR
variables that changed signifi cantly at p < .10. Data for multivariable models
signifi cant at p ≤ .05 are in bold. Data for models that approach signifi cance
are included but are not in bold. OR = odds ratio.
reinforces and supports subsequent reductions in drinking.
Specifi cally, increases in forgiveness of self and private SR
practices were consistent predictors of 9-month outcomes.
For each unit increase in forgiveness of self (note the average
change was 1 point or unit), PDA increased by 1.3 percent-
age points, DSLD increased by 11.2 days, and the odds of
a heavy drinking day or any drinking were 14% less likely
at 9 months. Similarly for private SR practices, which also
changed 1 point on average, PDA increased by 1.2 percent-
age points, DSLD increased by 5.8 days, and participants
were 6% and 7% less likely to engage in heavy drinking and
any drinking, respectively.
The current analyses provide evidence that SR change
global measure of forgiveness from the Fetzer Institute/
National Institute on Aging (1999) were not associated with
outcomes, this study found increases in forgiveness of self
and overall forgiveness to be predictive of drinking out-
comes. The fi ndings on the importance of forgiveness of self
are congruent with other studies. Several studies have docu-
mented that, among those with alcohol problems (Webb and
Brewer, 2010; Webb et al., 2006), forgiveness of self versus
forgiveness of others and feeling forgiven by God were the
least endorsed dimensions of forgiveness. Two other studies
have found evidence of the relative importance of forgive-
ness of self in alcohol outcomes (Webb et al., in press, sub-
mitted for publication-b). Webb et al. (in press) found that,
at baseline, 6-month follow-up, and longitudinally (baseline
to 6 months), forgiveness was associated with lower alcohol
use and problems, indirectly through mental health. Base-
line forgiveness of self and forgiveness of others were both
Although our earlier study found that increases in the
666 JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / JULY 2011
individually associated with multiple drinking outcomes. At
6-month follow-up and longitudinally, relationships were
found only for the outcome of alcohol problems and for
forgiveness of self or forgiveness of others. Similarly, Webb
et al. (submitted for publication-b) describe cross-sectional
fi ndings of salutary associations between alcohol outcomes
and forgiveness of self (largely through mental health) and
feeling forgiven by God (only directly). Webb and colleagues
(in press) argue that forgiveness of one’s self may be the
most important dimension of forgiveness in substance use
disorders and recovery, as compared with forgiveness of
others and by God. Our results are congruent with that
Our fi ndings on forgiveness of others are surprising, par-
ticularly given that clinicians and many in recovery advocate
efforts to forgive others. AA’s central text, Alcoholics Anony-
mous, states that “resentment is the ‘number one’ offender”
(AA, 1976, p. 64). However, if AA involvement is the pri-
mary contributor toward forgiveness of others, this analysis
would obscure this relationship, because we controlled for
AA involvement. To check, we ran the analyses without
controlling for AA involvement and found no signifi cant
relationships between drinking outcomes and forgiveness
of others. Our results are also incompatible with the work
of Lin et al. (2004). They contrasted two therapies designed
to augment the usual treatment received by individuals in
a residential alcohol/drug treatment center: forgiveness (of
others) therapy and treatment as usual. The forgiveness (of
others) therapy was found to more effectively reduce a vari-
ety of negative emotions and vulnerability to substance use;
no data were collected on actual behaviors. Our post hoc
analysis without controlling for AA involvement and the Lin
et al. study (2004) make us doubt that this fi nding of a nega-
tive relationship between increases in forgiveness of others
and one drinking variable will hold up in future research.
We might speculate that effective change in forgiveness of
others may develop more slowly during recovery, particularly
without AA involvement. Forgiveness of others may be more
effective after one has begun to forgive one’s self, later in the
recovery process, or with AA involvement. We may also fi nd,
as have others, that mental health status may play a role in
forgiving others (Webb et al., 2009, in press, submitted for
In sum, although research supports multiple dimensions
of forgiveness as factors in substance use disorders and
recovery, it appears that forgiveness of self may be most
important, particularly in the fi rst 6 months of recovery. In-
deed, whereas the Big Book (AA, 1976) discusses the central
role of resentments and the resolution of grudges, which are
commonly conceptualized as interpersonal concerns, the Big
Book also discusses the value of forgiveness of self. This
evidence of the importance of forgiveness of self, although
not overlooking the importance of other dimensions, may
require a shift in attention in treatment and prevention.
Private spiritual and religious practices
SR practices predict drinking outcomes suggests that SR-
related behaviors, such as prayer, meditation, and reading,
may be relevant to improved outcome. Note that this is a
measure of private religious practices, not involvement with
a religious congregation or worship with others. These fi nd-
ings provide support for the anecdotal reports of recover-
ing alcoholics that prayer and other practices support their
sobriety, in this case without involvement in AA. This is
the most behavioral of our SR measures. In our previous
cross-sectional analysis, we had found that private practices
were not associated with drinking outcomes, which led us
to speculate that early changes in SR dimensions relevant to
drinking may be primarily cognitive. However, these results
suggest that behaviors, specifi cally private SR practices, may
also support initial sobriety.
The consistency of our fi nding that increases in private
Negative religious coping
judgmental, condemning, or abandoning deity are associated
with improved drinking outcomes at 9 months. Although at
baseline few respondents strongly endorsed feeling aban-
doned and/or punished by God, those whose scores improved
(i.e., endorsing fewer negative coping strategies 6 months
later) were drinking less at 9 months. Decreases in this di-
mension rest on shifting to a more benevolent perception of
and relationship to a deity.
Decreases in negative SR coping strategies that assume a
Daily spiritual experiences
2007), it appears that changes in experiences of day-to-day
spirituality may be important in recovery. Increases in per-
ceptions of connection to God and of receiving strength,
comfort, and love from God, as well as experiences of
peacefulness and awe, are associated with less subsequent
drinking. Again, because this fi nding is based on control-
ling for AA involvement, the catalyst for this increase is
unclear. It may be a consequence of increases in private SR
As we reported in our previous article (Robinson et al.,
Purpose in life
to be an important issue for alcoholics as they recover,
because increases remain a predictor of subsequent drink-
ing outcomes (for at least three of four outcomes). It may
be intuitively obvious that, as alcoholics become sober,
an increase in the sense that one’s life has meaning might
provide motivation to sustain sobriety. However, the causal
connection and the mechanisms behind this relationship are
Sense of meaning/purpose in life was again confi rmed
ROBINSON ET AL. 667
not empirically obvious. Interventions specifi cally targeted at
increasing recovering alcoholics’ sense of meaning/purpose
in their lives, such as logotherapy (Frankl, 1992), might help
to determine how this factor affects recovery.
Alcoholics Anonymous involvement
association with SR and sobriety, which allowed us to iden-
tify the impact of non-AA-related SR change. However,
the relationship between SR change and AA involvement
was thereby obscured, making it harder to answer questions
about the nature of AA involvement that leads to SR change
and which dimensions of SR are most affected by such in-
volvement. Subsequent analyses should investigate how AA
involvement affects various dimensions of SR, particularly
because AA-mediated SR change is probably the most com-
mon type of SR change among alcoholics.
A panel study cannot defi nitively clarify causal rela-
tionships; however, we found that SR change appears to
reinforce and support reductions in drinking frequency and
quantity, regardless of AA involvement. Given the pervasive-
ness of the 12-step model in the context of contemporary
recovery, it is striking that SR change has an independent
relationship with subsequent drinking patterns. In many
ways, we set a fairly conservative test of the impact of SR
change, because we controlled for the factor most commonly
thought to alter SR among alcoholics. Our fi ndings suggest
that alcoholics who are decreasing their drinking without AA
fi nd ways of supporting and nurturing their SR. From a clini-
cal perspective, this is a reminder that non-AA-mediated SR
change occurs and that supporting positive shifts in clients’
SR may be helpful, regardless of whether the context for that
shift is within AA. Further work may show that SR change
is a component of natural recovery (Sobell et al., 2000).
We controlled for AA involvement because of its strong
Limitations and future directions
one geographic area, the Midwest of the United States, and
it is known that there are striking geographic differences
in religiousness (Pew Forum on Religion and Public Life,
2008). In addition, this was hardly a representative sample of
alcoholics, treated or untreated, because the NESARC data
(Cohen et al., 2007; Dawson et al., 2007) have indicated that
the fi eld’s treatment samples do not represent the vast major-
ity of alcoholics. Even with the inclusion of some untreated
alcoholics, the generalizability of this study to all alcoholics
is therefore limited, particularly in applying its fi ndings to
ethnic minorities and other specifi c populations. In terms
of the measures used in this study of our SR constructs, the
majority of the measures are robust. Nevertheless, there are
many questions that may be answered only through qualita-
tive data collection and analysis, such as the idiographic
This study was limited in drawing its sample from only
nature of alcohol-dependent individuals’ spiritual and reli-
gious cognitions and life experiences. Lastly, the time span
between SR change and subsequent outcomes may be short
by some standards, suggesting the need to confi rm this rela-
tionship using a longer time span.
Further work is needed to replicate these fi ndings and, in
particular, to examine the specifi c roles of the SR dimensions,
which we found to be predictive of subsequent drinking out-
comes—forgiveness of self, private religious practices, overall
forgiveness, purpose in life, daily spiritual experiences, and
negative religious coping. Although this evidence suggests
that some SR changes affect subsequent drinking, the causal
relationship may be more indirect (i.e., through other variables,
such as health behaviors, interpersonal functioning, social
support, and/or mental health; Webb and Trautman, 2010;
Webb et al., in press, submitted for publication-a, submitted
for publication-b; Worthington et al., 2001). We may also
speculate that reductions in drinking may lead to increases
in SR and that the relationship is bi-directional.
Although the effects of SR change on subsequent drink-
ing, independent of AA involvement, appear modest, the per-
sistence of these fi ndings suggests that SR does play a role
in substance use disorders and recovery. Cross-sectional and
longitudinal data suggest, at the least, an interaction between
SR and reductions in alcohol use. It is valuable to fi nd that
reductions in drinking and subsequent suffering are associ-
ated with change in several SR dimensions (forgiveness of
self, private religious practices, overall forgiveness, daily
spiritual experiences, purpose in life, and negative religious
coping). Further work is needed to examine how change
in these SR dimensions affects dependence and recovery
ing out the initial analyses, and James A. Cranford, Ph.D., for statistical
The authors acknowledge the work of Laura Pierce, Ph.D., in carry-
Alcoholics Anonymous. (1976). Alcoholics Anonymous: The story of how
many thousands of men and women have recovered from alcoholism
(3rd ed.). New York, NY: Alcoholics Anonymous World Services.
Benson, P., & Spilka, B. (1973). God image as a function of self-esteem
and locus of control. Journal for the Scientifi c Study of Religion, 12,
Carrico, A. W., Gifford, E. V ., & Moos, R. H. (2007). Spirituality/religiosity
promotes acceptance-based responding and 12-step involvement. Drug
and Alcohol Dependence, 89, 66–73.
Cherpitel, C. J. (1995). Screening for alcohol problems in the emergency
room: A rapid alcohol problems screen. Drug and Alcohol Dependence,
Cohen, E., Feinn, R., Arias, A., & Kranzler, H. R. (2007). Alcohol treat-
ment utilization: Findings from the National Epidemiologic Survey on
Alcohol and Related Conditions. Drug and Alcohol Dependence, 86,
668 JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / JULY 2011 Download full-text
Cohen, J. (1988). Statistical power analysis for the behavior sciences (2nd
ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.
Cohen, J. (1992). Statistics: A power primer. Psychology Bulletin, 112,
Connors, G. J., Tonigan, J. S., & Miller, W. R. (1996). Measure of religious
background and behavior for use in behavior change research. Psychol-
ogy of Addictive Behaviors, 10, 90–96.
Crumbaugh, J. C., & Maholick, L. T. (1964). An experimental study in exis-
tentialism: The psychometric approach to Frankl’s concept of noogenic
neurosis. Journal of Clinical Psychology, 20, 200–207.
Dawson, D. A., Goldstein, R. B., & Grant, B. F. (2007). Rates and corre-
lates of relapse among individuals in remission from DSM-IV alcohol
dependence: A 3-year follow-up. Alcoholism: Clinical and Experimental
Research, 31, 2036–2045.
Dawson, D. A., Grant, B. F., Stinson, F. S., Chou, P. S., Huang, B., & Ruan,
W. J. (2005). Recovery from DSM-IV alcohol dependence: United
States, 2001–2002. Addiction, 100, 281–292.
Fetzer Institute/National Institute on Aging. (1999). Multidimensional mea-
surement of religiousness/spirituality for use in health research: A report
of the Fetzer Institute/National Institute on Aging Working Group, with
additional psychometric data. Kalamazoo, MI: Author.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1997). Struc-
tured Clinical Interview for DSM-IV Axis I Disorders (SCID-I), Clinician
version: User’s guide. Washington, DC: American Psychiatric Press.
Flynn, P. M., Joe, G. W., Broome, K. M., Simpson, D. D., & Brown, B. S.
(2003). Recovery from opioid addiction in DATOS. Journal of Sub-
stance Abuse Treatment, 23, 177–186.
Forcehimes, A. A., Tonigan, J. S., Miller, W. R., Kenna, G. A., & Baer,
J. S. (2007). Psychometrics of the drinker inventory of consequences
(DrInC). Addictive Behaviors, 32, 1699–1704.
Frankl, V. (1969). The will to meaning: Foundations and applications of
logotherapy. New York, NY: New American Library.
Frankl, V. E. (1992). Man’s search for meaning: An introduction to logo-
therapy (I. Lasch, Trans.; 4th ed.). Boston, MA: Beacon Press.
Heinz, A., Epstein, D. H., & Preston, K. L. (2007). Spiritual/religious ex-
periences and in-treatment outcome in an inner-city program for heroin
and cocaine dependence. Journal of Psychoactive Drugs, 39, 41–49.
Jarusiewicz, B. (2000). Spirituality and addiction: Relationship to recovery
and relapse. Alcoholism Treatment Quarterly, 18, 99–109.
Kurtz, E., & Ketcham, K. (1992). The spirituality of imperfection: Storytell-
ing and the journey to wholeness. New York, NY: Bantam.
Lin, W. F., Mack, D., Enright, R. D., Krahn, D., & Baskin, T. W. (2004).
Effects of forgiveness therapy on anger, mood, and vulnerability to
substance use among inpatient substance-dependent clients. Journal of
Consulting and Clinical Psychology, 72, 1114–1121.
Mauger, P. A., Perry, J., Freeman, T., Grove, D., McBridge, A., & McKin-
ney, K. (1992). The measurement of forgiveness: Preliminary research.
Journal of Psychology and Christianity, 11, 170–180.
Menard, S. (1991). Longitudinal research. Thousand Oaks, CA: Sage.
Miller, W. R., Tonigan, J. S., & Longabaugh, R. (1995). The Drinker Inven-
tory of Consequences (DrInC): An instrument for assessing adverse con-
sequences of alcohol abuse. Project MATCH Monograph Series, Volume
4. Rockville, MD: National Institute on Alcohol Abuse and Alcoholism.
Pargament, K. I., Smith, B. W., Koenig, H. G., & Perez, L. (1998). Patterns
of positive and negative religious coping with major life stressors. Jour-
nal for the Scientifi c Study of Religion, 37, 710–724.
Pew Forum on Religion and Public Life. (2008). U.S. religious landscape
survey: Religious beliefs and practices: Diverse and politically relevant.
Washington, DC: Author.
Piderman, K. M., Schneekloth, T. D., Pankratz, V. S., Maloney, S. D., &
Altchuler, S. I. (2007). Spirituality in alcoholics during treatment.
American Journal on Addictions, 16, 232–237.
Piderman, K. M., Schneekloth, T. D., Pankratz, V. S., Stevens, S. R., &
Altchuler, S. I. (2008). Spirituality during alcoholism treatment and
continuous abstinence for one year. International Journal of Psychiatry
in Medicine, 38, 391–406.
Poage, E. D., Ketzenberger, K. E., & Olson, J. (2004). Spirituality, content-
ment, and stress in recovering alcoholics. Addictive Behaviors, 29,
Robinson, E. A. R., Cranford, J. A., Webb, J. R., & Brower, K. J. (2007).
Six-month changes in spirituality, religiousness, and heavy drinking in
a treatment-seeking sample. Journal of Studies on Alcohol and Drugs,
Sandoz, C. J. (1999). The spiritual experience in recovery: A closer look.
Journal of Ministry in Addiction and Recovery, 6, 53-59.
Sherman, J., & Fischer, J. M. (2002). Spirituality and addiction recovery for
rehabilitation counseling. Journal of Applied Rehabilitation Counseling,
Sobell, L. C., Brown, J., Leo, G. I., & Sobell, M. B. (1996). The reliability
of the Alcohol Timeline Follow-back when administered by telephone
and by computer. Drug and Alcohol Dependence, 42, 49–54.
Sobell, L. C., Ellingstad, T. P., & Sobell, M. B. (2000). Natural recovery
from alcohol and drug problems: Methodological review of the research
with suggestions for future directions. Addiction, 95, 749–764.
Sobell, L. C., & Sobell, M. B. (1992). Timeline Follow-back: A technique
for assessing self-reported alcohol consumption. In R. Litten and J. Al-
len (Eds.), Measuring alcohol consumption: Psychosocial and biochemi-
cal methods. Totowa, NJ: Humana Press.
Sterling, R. C., Weinstein, S., Losardo, D., Raively, K., Hill, P., Petrone, A.,
& Gottheil, E. (2007). A retrospective case control study of alcohol re-
lapse and spiritual growth. American Journal on Addictions, 16, 56–61.
Stewart, C., & Koeske, G. F. (2005). Spiritual change in treatment of sub-
stance dependence. Social Work and Christianity, 32, 321–340.
Tonigan, J. S., Connors, G. J., & Miller, W. R. (1996). Alcoholics Anony-
mous Involvement (AAI) scale: Reliability and norms. Psychology of
Addictive Behaviors, 10, 75–80.
Underwood, L. G., & Teresi, J. A. (2002). The Daily Spiritual Experiences
Scale: Development, theoretical description, reliability, exploratory fac-
tor analysis, and preliminary construct validity using health-related data.
Annals of Behavioral Medicine, 24, 22–33.
Waisberg, J. L., & Porter, J. E. (1994). Purpose in life and outcome of treat-
ment for alcohol dependence. British Journal of Clinical Psychology,
Webb, J. R., & Brewer, K. (2010). Forgiveness and college student drinking
in Southern Appalachia. Journal of Substance Use, 15, 417–433.
Webb, J. R., Hill, S. K., & Brewer, K. (2010). Dimensions of social support
as mediators of the forgiveness–alcohol outcome relationship. Manu-
script submitted for publication-a.
Webb, J. R., Hirsch, J. K., Conway-Williams, E., & Brewer, K. (2011).
Forgiveness, mental health, social support, and collegiate problem-
atic drinking in Southern Appalachia. Manuscript submitted for
Webb, J. R., Robinson, E. A. R., & Brower, K. J. (2009). Forgiveness and
mental health among people entering outpatient treatment with alcohol
problems. Alcoholism Treatment Quarterly, 27, 368–388.
Webb, J. R., Robinson, E. A. R., & Brower, K. J. (in press). Mental health,
not social support, mediates the forgiveness–alcohol outcome relation-
ship. Psychology of Addictive Behaviors.
Webb, J. R., Robinson, E. A. R., Brower, K. J., & Zucker, R. A. (2006).
Forgiveness and alcohol problems among people entering substance
abuse treatment. Journal of Addictive Diseases, 25, 55–67.
Webb, J. R., & Trautman, R. P. (2010). Forgiveness and alcohol use: Apply-
ing a specifi c spiritual principle to substance abuse problems. Addictive
Disorders and Their Treatment, 9, 8–17.
Worthington, E. L., Jr., Berry, J. W., & Parrott, L., III. (2001). Unforgive-
ness, forgiveness, religion, and health. In T. G. Plante and A. C. Sher-
man (Eds.), Faith and health: Psychological perspectives (pp. 107–138).
New York, NY: Guilford Press.