Reducing alcohol-exposed pregnancy risk in college women: Initial outcomes of a clinical trial of a motivational intervention

Abstract

A significant number of college women are at risk for alcohol-exposed pregnancy (AEP) owing to binge drinking paired with using contraception ineffectively. This article describes a randomized controlled trial of a one-session motivational interviewing-based intervention to reduce AEP risk among college women and presents 1-month outcomes demonstrating the early impact of this intervention. There were 228 female students from a mid-Atlantic urban university enrolled in the trial. Eligibility criteria were being in the age range of 18–24 years and being at risk for AEP. Risk for AEP was defined as having sexual intercourse with a man in the past 90 days while using contraception ineffectively (no use, incorrect use of an effective method, or use of an ineffective method only); drinking at risky levels was defined as engaging in at least one binge in the past 90 days or consuming an average of eight standard drinks per week. One-month outcome data were available for 212 of the 228 enrolled women (a follow-up rate of 93%), with complete data available for 105 women assigned to the control condition and 94 assigned to the intervention condition. At 1-month follow-up, 15% of the control subjects and 25% of the intervention women reported no risk drinking, a significant difference favoring the intervention group. Significantly fewer control subjects (48%) used effective contraception at 1-month follow-up as compared with intervention women (64%), χ²(1) = 5.1, p < .03. Significantly more intervention women (74%) were no longer at risk for AEP at 1 month as compared with control subjects (54%), χ²(1) = 8.15, p < .005. Factors that were associated with continued AEP risk at 1-month follow-up were a higher number of standard drinks per day consumed in the month prior to baseline (odds ratio, 1.1) and assignment to the control condition (odds ratio, 2.9). The risks of unintended pregnancy and AEP among drinking women in college merit greater prevention efforts. The results of this study show the promise of one preventive intervention that warrants additional study.

1. Introduction

The college population is replete with high-risk drinkers. The 2001 survey of 10,904 students at 119 colleges and universities conducted by the Harvard School of Public Health found that 40% of female students drank four or more drinks in a row in the past 2 weeks, meeting the criteria for binge drinking (Wechsler et al., 2002). Four of five sorority and fraternity members surveyed met the standard for binge drinking. One in five students was a frequent binge drinker, defined as having three binges in the past 2 weeks, a rate that has not changed since 1993. Women’s drinking, in particular, appears to be increasing with more than a doubling of frequent binge drinking among women enrolled in all-women’s colleges (from 5.3% in 1993 to 11.9% in 2001) and a lesser, but still significant, increase for women in coeducational schools. Reifman and Watson (2003) found that women had a higher probability of adopting binge drinking during their first year of college than men. College risky drinking, and especially engaging in binging, is an entrenched behavior with numerous problematic outcomes such as higher total volume of alcohol consumed, higher consumption of illicit drugs and cigarettes, unplanned sex, rape, accidents, driving while intoxicated, and deaths (Howard & Wang, 2004; Jones, Oeltmann, Wilson, Brener, & Hill, 2001; Wechsler & Isaac, 1992; Wechsler, Davenport, Dowdall, Moeykens, & Castillo, 1994). A cross-sectional and prospective study by O’Neill, Parra, and Sher (2001) concluded that heavy drinking during the college years strongly predicts alcohol use disorders up to 10 years later.

Another health risk to college women is inadequate use of contraception. Sawyer, Pinciaro, and Anderson-Sawyer (1998) investigated the reasons for pregnancy testing at a student health clinic by college women. They found that 37% of the women failed to use any contraception method at the time of possible pregnancy. Pregnancy testing data were examined over a 5-year period and revealed that nearly 60% of women had a method of birth control but did not use it on the occasion of sexual intercourse. Their findings suggest that unintended pregnancy in college students is an ongoing problem. Their results also indicated that even when students adopt a contraceptive method, they fail to use it consistently. Similarly, Kusseling, Wenger, and Shapiro (1995) found that female college students failed to use contraception effectively over time, typically using them reliably at baseline but demonstrating subsequent failure to use them effectively at follow-up. Their study confirmed that college women are at high risk for unintended pregnancy and suggested that interventions should address the consistency of contraceptive use. Ni-Riain (1998) found that knowledge and access to contraceptives are not sufficient to ensure correct, consistent use of contraceptives. All of the participants in the study were students who presented for emergency contraception in a 3.5-year period at a large student health center. These women became more effective contraceptive users after an interactive counseling session following the pregnancy scare. These results could indicate that women’s motivation to use contraception changed following feedback that they could become pregnant if they did not improve their contraception use.

Combining binge drinking or regular drinking with using contraception ineffectively results in the risk of alcohol-exposed pregnancy (AEP). The consequences of unintended pregnancy affected by alcohol may include increased abortion rates or a range of neurobehavioral effects if the fetus is carried to term, which are considered to be fetal alcohol spectrum disorders, including the most severe form, fetal alcohol syndrome (American Academy of Pediatrics, 2000; Eustace, Kang, & Coombs, 2003; NOFAS, 2004; Sihvo, Bajos, Ducot, & Kaminski, 2003). Among 2,672 community women surveyed in six settings across three communities, rates of risk for AEP were high—between 10% and 26% depending on setting (Project CHOICES Research Group, 2002). Given that high rates of risky-level drinking are consistently reported in college women (Wechsler et al., 2002), these women may also be at risk for AEP if they are sexually active and fail to use effective contraception. A recent survey of 2,012 university women found that most sexually active women (74%) were risky drinkers and that a significant minority (21%) were ineffective contraceptive users, with a combined risk for AEP evidenced in 17% of those who were sexually active (Ceperich, Ingersoll, Nettleman, & Johnson, 2004). College women were at risk primarily because of binge drinking paired with using contraceptives ineffectively.

To address the risk of AEP among college women, interventions could target drinking, contraception, or both. There have been many efforts to reduce problematic drinking among college students, with no published reports of interventions to improve contraception among college women. Heavy-drinking high school seniors who received a motivational intervention showed decreased drinking and lower levels of negative consequences of drinking 2 years later (Marlatt et al., 1998). Baer et al. (1992) found that a single session of individualized feedback and professional advice using a motivational counseling style produced 40% reductions in drinking for 2 years among young adults. These results were equivalent to those achieved by a 6-week skills building intervention.

Several studies have attempted to reduce risky drinking in college samples using assessment and personalized feedback either delivered in a motivational interviewing (MI) counseling style or through mailed feedback without interpersonal interaction. Nye, Agostinelli, and Smith (1999) found that providing either self-focusing information or norms-based feedback information increased problem recognition and discrepancy about current drinking and other goals. Murphy et al. (2001) randomized heavier-drinking college students to BASICS, which is an individual treatment including cognitive–behavioral skills training, motivational enhancement, and feedback about drinking patterns, to an education condition in which participants watched two alcohol-related videos and talked individually with a counselor afterward, or to a no-treatment control. Heaviest-drinking students who were in the BASICS condition showed the most reduction at 3- and 9-month follow-up. Agostinelli, Brown, and Miller (1995) found that mailed personalized feedback resulted in reductions in drinking and peak intoxication levels among heavy-drinking college students. Similarly, Walters (2000) found that mailed feedback designed to influence drinking among drinking college students was superior to group or classroom education plus feedback and to controls at a 6-week follow-up in two studies conducted at different college campuses. In a similar study, Walters, Bennett, and Miller (2000) found that mailed feedback produced a 53% decrease in drinking quantity at 6-week follow-up. Mailed feedback was again superior to participation in a psychoeducational group that included feedback and to control. Taken together, these studies suggest that intervention models using personalized risk feedback or a single session focusing on an individual at risk can be effective to reduce drinking among the college-age population. Additional sessions have shown no enhanced benefit despite their increased cost.

To date, there are no published reports on preventing the combined risk of AEP owing to risky drinking and inadequate contraception among sexually active college women. One model of an individual-focused intervention used a platform of MI (Miller & Rollnick, 2002) to deliver a five-session personalized feedback and counseling plus contraceptive consultation intervention to community women at risk for AEP (Project CHOICES Intervention Research Group, 2003). This intervention was found to be feasible and promising with community women; 68.5% of participants were no longer at risk for AEP at the 6-month follow-up. More women achieved this by changing both drinking and contraception (32.9%) than by improving contraception only (23.1%) or by reducing drinking only (12.6%). Therefore, interventions to reduce AEP risk among college women should include an emphasis on improving contraception and/or reducing risky drinking. In this article, we describe a novel one-session intervention to reduce AEP risk among college women and present 1-month outcomes of a randomized trial designed to test the efficacy of this intervention.

2. Materials and methods

2.1. Participants

Two hundred twenty-eight female students from a mid-Atlantic urban university enrolled in the trial. Eligibility criteria were being 18 to 24 years old and being at risk for AEP. The status of being at risk for AEP included having sexual intercourse with a man in the past 90 days, using contraception ineffectively (no use, incorrect use of an effective method, or use of an ineffective method only), and drinking at risk levels, defined as consuming five or more standard drinks per occasion (binge) at least once in the past 90 days or consuming eight or more standard drinks per week on average.

2.2. Measures

The assessment battery required 1–1.5 hours and covered demographic characteristics, sexual behavior history, contraception history, drinking history, and health behaviors and attitudes including multivitamin use, knowledge of folic acid functions, and belief that drinking daily has health benefits, among others. In addition, we measured illicit drug use, personality variables, and psychiatric distress, all areas hypothesized to relate to the two behaviors constituting risk for AEP, risky drinking and ineffective contraception. Personality variables were measured with the Five-Factor Inventory (FFI) and psychiatric distress was measured with the Brief Symptom Inventory (BSI) and the Outcomes Questionnaire 45 (OQ-45). The FFI (Costa & McCrae, 1992) is a 60-item form of the 240-item NEO-Personality Inventory (PI), a widely used instrument for assessing normal adult personality on five major dimensions of normal personality: neuroticism, extraversion, openness, conscientiousness, and agreeableness. Responses are on a five-point scale from strongly disagree to strongly agree. The NEO-PI is based on decades of factor analytic research and its five major domain scales have high internal consistency (.86–.95) and can be considered personality traits. The FFI takes approximately 10–15 minutes to take and approximately 5 minutes to score.

The BSI (Derogatis, 1993) assesses psychological symptoms and state distress experienced in the past 2 weeks using self-report for 53 questions on a five-point scale. Because it measures state psychiatric distress, it is expected to change over time. The BSI requires 4–5 minutes to complete and 5 minutes to score using its computer scoring program. Similarly, the OQ-45.1 (Lambert, Lunnen, Umphress, Hansen, & Burlingame, 1994) is a 45-item self-report outcome/tracking instrument designed to assess change over time in three domains: symptom distress, interpersonal functioning, and social role. Because the BSI is more typically used with clinical samples than with college students, the OQ-45.1 was added because its test–retest reliability in college samples has ranged from .66 to .86 and its internal consistency reliability is excellent, generally above .90 in college samples for the total and symptom distress scales and ranging from .70 to .90 for the interpersonal functioning and social role scales. It takes 5 minutes to administer and score.

Study-specific questions were adapted from recent studies assessing AEP risk, including the Project CHOICES feasibility study (Project CHOICES Intervention Research Group, 2003).

2.3. Procedures

Women were recruited for the study through mailings or flyers posted on campus and in the student health center and were screened over the telephone or in person. A research counselor administered the assessment battery to eligible and consenting women. Following the assessment, the counselor opened a randomization envelope and proceeded to conduct either the control condition or the intervention condition. Follow-up assessments were mailed to participants at 1 month postbaseline.

2.4. Group conditions

The intervention was entitled BALANCE (Birth Control and Alcohol Awareness: Negotiating Choices Effectively) and included a single session of personalized feedback and MI counseling style. The BALANCE intervention, based on MI, was a 60- to 75-minute counseling session following a semistructured counseling manual. The intervention used the MI counseling style throughout a series of structured activities. Activities included recording 90 days of timeline follow-back (TLFB; Sobell & Sobell, 1992) data on drinking and contraception, providing personalized feedback of risk, and using exercises including decisional balance, temptation, and confidence charts and development of goal statements and change plans for drinking and contraception. After a participant had recorded TLFB data and engaged in several of the activities, she was given a 10-minute break while the counselor computed her feedback information. Information was based on the TLFB and temptation and confidence scales related to both drinking and contraception behaviors. Following the break, the counselor provided feedback using an elicit–provide–elicit strategy commonly used in MI. The counselor then guided the participant in filling out importance, confidence, and readiness rulers for changing drinking and contraception behaviors. The counselor and participant then collaboratively created a change plan for both behaviors. If a participant indicated that she did not want to change a behavior, she was guided to create a plan to maintain current behaviors (rather than to increase drinking or lack of contraception). All sessions were audiotaped, and tapes were used in weekly individual and group supervision sessions conducted by the senior authors. A complete description of the BALANCE intervention was presented elsewhere (Ingersoll, Ceperich, & Nettleman, 2004).

The control condition involved an informational pamphlet about women’s health. A minimal treatment control is consistent with recommendations for initial testing of a novel intervention (Rounsaville, Carroll, & Onken, 2001).

2.5. Analyses

Descriptive statistics characterized the sample demographics, contraceptive behaviors, drinking behaviors, and AEP risk. Change scores were derived for continuous measures by subtracting baseline values from 1-month outcome values. t Tests and χ² analyses were used to assess whether intervention and control groups differed at baseline and at 1 month. Finally, a hierarchical logistical regression analysis was used to determine multivariate predictors of the positive outcome of reduced AEP risk.

3. Results

3.1. Baseline characteristics of the BALANCE sample

Participants were mostly Caucasian (70%) and African American (16%), with an average age of 20 years. Most (88%) were single/never married, living with a roommate (56%), and attending university full time (90%). Most worked, either full time (22%) or part time (71%). Demographic characteristics of the sample by group are shown in Table 1.

Table 1.

Demographic characteristics of the BALANCE sample by group

Variable	Control (n = 114)	Intervention (n = 114)
Age [years, M (SD)]	20.5 (1.8)	20.4 (1.8)
Race [n (%)]
Caucasian	83 (73)	76 (67)
African American	17 (15)	20 (17)
Asian	3 (3)	10 (9)
Latina	3 (3)	2 (2)
Other	4 (4)	5 (4)
Pacific Islander	2 (2)	1 (1)
Single [n (%)]	99 (88)	99 (88)
Married [n (%)]	8 (7)	9 (8)
Living with [n (%)]
Roommate	64 (56)	64 (56)
Alone	18 (16)	22 (19)
Parent	10 (9)	13 (11)
Relative	12 (11)	8 (7)
Male partner	10 (9)	7 (6)
Student statusa [n (%)]
Full-time student	109 (96)	97 (95)
Part-time student	5 (4)	17 (15)
Full-time work [n (%)]	17 (23)	16 (21)
Part-time work [n (%)]	52 (70)	54 (72)

^aχ²(1) = 7.24, p < .008.

Regarding drinking and sexual behavior at baseline among the full sample (intervention and control groups combined), most women were binge drinking, having sex with one partner, and using withdrawal as a method of birth control. Most women (73%) reported that they drank 3–6 drinks on an average drinking day, and 42% reported drinking an average of eight or more standard drinks per week.

Most women (73%) had one male sexual partner in the past 3 months whereas 17.2% had two, 5.3% had three, and 4.3% had four or more male sexual partners in the past 3 months. Natural (49%) or hormonal (23%) contraception methods were the most prevalent, with withdrawal, condoms, and birth control pills being the most common methods used.

Table 2 depicts these and other drinking and contraception characteristics of the sample by group and provides tests of baseline differences between the two groups. No difference existed between the control and intervention groups on baseline sexual and contraceptive behaviors including age at initiating sex, age at initiating contraception, type of contraception, number of sexual partners, and reasons for not using contraception. Groups were also similar with regard to baseline drinking variables including age at having the first drink, average standard drinks per day, average drinks per week, most standard drinks in 1 day, number of binges in the past 1- and 3-month periods, and drinking risk category: binge only (46% control, 44% intervention), binge with eight or more drinks per week (44% control, 41% intervention), or binge with unknown number of drinks per week (11% control, 15% intervention). Randomization failed to balance the groups only on student status, with fewer full-time students in the intervention group (85%) than in the control group (96%).

Table 2.

Baseline contraception and drinking characteristics of the BALANCE sample by group

Variable	Control (n = 114)	Intervention (n = 114)	t or χ2
Age at first sexual intercourse     [years, M (SD)]	16.3 (2.0)	16.2 (1.7)	ns
Age at first contraception use     [years, M (SD)]	16.3 (1.9)	16.5 (1.6)	ns
Sexual partners in 3 months     [M (SD)]	1.4 (0.8)	1.5 (1.0)	ns
Type of current contraception     [n (%)]			ns
Withdrawal	53 (46)	56 (50)
Condoms	22 (19)	25 (22)
Pill	25 (22)	21 (19)
Nothing	4 (4)	5 (4)
Rhythm	3 (3)	3 (3)
Spermicide	2 (2)	2 (2)
Emergency	2 (2)	1 (1)
Depo provera	2 (2)	0 (0)
Reasons for no contraception     [n (%)]			ns
Heat of the moment	56 (49)	58 (51)
Agreed not to	8 (7)	9 (8)
Forgot	9 (8)	4 (4)
Alcohol or drug use	8 (7)	5 (4)
Inaccessible	0 (0)	4 (4)
Enjoy sex less	1 (1)	2 (2)
Trust partner	9 (8)	9 (8)
Unconcerned	3 (3)	3 (3)
Side effects	2 (2)	1 (1)
Other	18 (14)	19 (17)
Estimate of the chance of     pregnancy in a year with no     contraception [M (SD)]	64.2 (29.3)	63.9 (26.0)	ns
Ever had a pap smear? [n (%)]			ns
Yes	103 (92)	96 (84)
No	10 (8)	17 (15)
Ever treated for an STD? [n (%)]			ns
Yes	21 (18)	20 (18)
No	93 (82)	93 (82)
Ever tested for HIV? [n (%)]			ns
Yes	70 (61)	62 (54)
No	44 (39)	52 (46)
Age at having first full alcoholic     drink [years, M (SD)]	14.9 (2.4)	15.7 (8.0)	ns
Most standard drinks in 1 day     [M (SD)]	7.4 (3.8)	7.9 (4.0)	ns
No. of binges in the past month     [M (SD)]	4.3 (4.8)	4.1 (5.0)	ns
No. of binges in the past 3 months     [M (SD)]	12.4 (13.0)	13.5 (15.9)	ns
Average no. of standard drinks     per day [n (%)]			ns
None	1 (1)	4 (4)
1–2	11 (10)	16 (14)
3–4	46 (40)	43 (38)
5–6	44 (39)	33 (29)
7–9	9 (8)	13 (12)
≥10	3 (3)	4 (4)
Average no. of standard drinks     per week [n (%)]			ns
0	0 (0)	1 (1)
1–7	52 (46)	50 (44)
≥8	50 (44)	47 (41)
Risk drinking caused by [n (%)]			ns
Binge only	52 (46)	50 (44)
Binge and ≥8 per week	50 (44)	47 (41)
Binge and unknown no. of drinks     per week	12 (11)	17 (15)
Have you heard reports that one     drink a day is good for health?     [n (%)]			ns
Yes	40 (35.4)	47 (41.2)
No	64 (58.6)	58 (50.9)
Have you changed your drinking     because of those reports? [n (%)]			ns
Yes, increased it	4 (4)	6 (7)
Yes, decreased it	1 (1)	0 (0)
No change	90 (95)	84 (93)
Morning drinking [n (%)]			ns
Yes	8 (7)	8 (7)
No	106 (93)	106 (93)
Blackouts [n (%)]			ns
Yes	74 (65)	83 (73)
No	40 (35)	31 (27)
Have you thought you should cut     down on your drinking? [n (%)]			ns
Yes	43 (38)	52 (46)
No	71 (62)	62 (54)
Ever illicit drug use? [n (%)]			ns
Yes	92 (81)	94 (82)
No	22 (19)	20 (18)

Although 212 women (93%) returned 1-month follow-up questionnaires, data on one or more outcome variables were missing from 13 cases. Therefore, 1-month outcome data were available on 199 of the 228 enrolled women (see Table 3). They include 105 women assigned to the control condition and 94 assigned to the intervention condition, for a usable follow-up rate of 87.3%. Sixteen women (7%) were lost to follow-up.

Table 3.

One-month outcomes of the BALANCE trial

Variable	Control (n = 105)	Intervention (n = 94)	t or χ2
Average no. of standard drinks per week [M (SD)]	11.4 (10.7)	9.5 (14.7)	ns
Binges in the past month [M (SD)]	4.4 (4.8)	2.9 (3.8)	t = 2.34, Satterthwaite corrected for unequal variances, p < .02
Highest no. of standard drinks per day [M (SD)]	7.1 (3.6)	5.9 (3.5)	t = 2.12, p < .04
Change in no. of binges from baseline to MI [M (SD)]	0.2 (5.1)	−1.1 (4.0)	t = 1.81, Satterthwaite corrected for unequal variances, p < .07
Change in highest no. of standard drinks per day     from baseline to MI [M (SD)]	−0.4 (3.5)	−2.2 (4.1)	t = 3.08, p < .003
Drink risk [n (%)]			ns
None	16 (15.2)	27 (28.7)
Binge only	31 (29.5)	29 (30.9)
Binge and ≥8 per week	56 (53.3)	36 (38.3)
Binge and unknown no. of drinks per week	1 (1)	1 (1.1)
≥8 per week only	1 (1)	1 (1.1)
Drink risk (binary) [n (%)]			χ2(1) = 5.72, p < .02
No	16 (15.2)	27 (29.4)
Yes	89 (84.8)	67 (70.7)
Contraception [n (%)]			χ2(1) = 5.1, p < .03
Effective	50 (47.6)	58 (63.7)
Ineffective	55 (52.4)	33 (36.3)
AEP risk [n (%)]			χ2(1) = 8.15, p < .005
No	57 (54.3)	68 (73.9)
Yes	48 (45.7)	24 (26.1)

3.2. One-month drinking outcomes

All women reported risky-level drinking at baseline. In contrast, 15% of control subjects and 25% of intervention women reported no risk drinking at 1-month follow-up, a significant difference favoring the intervention group. Control subjects reduced their highest number of standard drinks per day by 0.4 whereas intervention women reduced theirs by 2.2; these significant differences also favored the intervention group (t = 3.08, p < .003). Women in the control condition increased slightly the number of binges at 1 month compared with that at baseline. In contrast, women in the intervention group reduced binges in the past month by 1.2 (SD, 4.0), a difference trending toward significance. All drinking comparisons are shown in Table 3.

3.3. One-month contraception outcomes

Significantly fewer control subjects (48%) used effective contraception at 1-month follow-up as compared with intervention women (64%), χ²(1) = 5.1, p < .03. More than half of the women in the control group remained in the ineffective contraception category whereas only approximately a third (36%) of the intervention continued to be ineffective users of contraception.

3.4. One-month primary outcome: Reduced AEP risk

Two of the women in the intervention group were unable to be categorized as to their risk classification owing to partial missing data and were removed from the analysis of reduced AEP risk. χ² Analysis revealed that 73.9% of intervention women and 54.3% of control subjects were no longer at risk for AEP at 1 month, χ²(1) = 8.15, p < .005. This difference indicates an effect size for the intervention of 0.19.

3.5. Predictors of continued risk for AEP

To determine whether group assignment was the primary factor influencing outcome or whether other baseline factors influenced response to the intervention or control condition, a logistical regression analysis was conducted to determine the factors explaining continued AEP risk using a multivariate approach. Specific baseline variables hypothesized to be potential mediators of outcome were entered, including age, drinking factors (highest number of drinks per day, number of binges in 3 months), contraception factors (age at first intercourse, age at first contraception use, number of partners in 3 months, and the woman’s estimate of the chances of pregnancy without contraception), psychiatric distress, and personality variables (Global Severity Index from the BSI and total score on the OQ-45 and neuroticism, extraversion, openness, conscientiousness, and agreeableness T scores from the NEO-FFI). Group was coded as a two-level dummy variable. Table 4 presents the results of the logistic regression analysis.

Table 4.

Logistic regression model of continued AEP risk among the BALANCE sample at 1-month follow-up

Variable	β	SE	OR	95% CI	Wald statistic
Assignment to control group	.52	.17	2.85	1.49–5.45	10.07**
Highest no. of standard drinks in the past month at baseline	.11	.04	1.12	1.03–1.22	6.47*
NEO-FFI extraversion T score	.02	.01	1.46	0.23–1.02	1.46, ns

^*p < .01.

^**p < .002.

Two variables were retained as independent risk factors in the final model, which was significant, likelihood ratio χ²(3) = 17.91, p = .0005. Factors that increased the risk of continued AEP risk at 1-month follow-up were assignment to the control condition (odds ratio [OR], 2.9; 95% confidence interval [CI], 1.49–5.45) and the highest number of standard drinks per day consumed in the month prior to baseline (OR, 1.1; 95% CI, 1.02–1.22). Assignment to the control condition nearly tripled the odds of remaining at risk for AEP whereas number of drinks per day conferred slight additional risk. The NEO-FFI extraversion T score remained in the model but was not an independent predictor.

4. Discussion

This is the first randomized controlled trial to show reduction of AEP risk following a motivational intervention with dual targets of risky drinking and ineffective contraception use. The BALANCE intervention was efficacious in reducing the risk of AEP at the 1-month follow-up; the effect size was moderate at 19%. The results are consistent with a previous uncontrolled study showing that a dually targeted motivational intervention reduced the risk of AEP in community women (Project CHOICES Intervention Research Group, 2003). College women reduced their risk in both groups by the 1-month follow-up, with significantly greater risk reduction in the motivational intervention group. Because group assignment was the strongest predictor of outcome and the group variable had a more potent effect than any other variable, we are confident that change is at least partially caused by the intervention.

Risk reduction was related to improved rates of effective contraception use and to reductions in the number of standard drinks consumed per drinking day. Women especially responded to the contraceptive part of the intervention and showed more change in contraception behavior than in drinking behavior. It is possible that because these college women have a strong motivation to avoid pregnancy, they were more willing to make changes in their contraception than they were to drink less alcohol. Unfortunately, most women retained a pattern of binge drinking following the intervention. However, given the intractability of binge drinking in this population, it is important to note that the intervention was effective in reducing the number of binges and amount of alcohol consumed overall.

College women have more consistently positive alcohol expectancies and consequences and lower negative consequences that result in low motivation to change this behavior (Park, 2004). These results suggest that offering feedback on pregnancy risk in addition to feedback on drinking may capitalize on college women’s strong desire to avoid pregnancy. The dual focus of an intervention such as this one allows participants to choose what change they want to make while conveying risk reduction messages in both areas that could reduce the risk of AEP and unplanned pregnancy.

This study has several limitations. First, the participating students completed follow-up information using a paper form that was mailed to them and returned to the researchers, which led to some incomplete or indecipherable answers. Second, the informed consent process and baseline assessment may have exerted an intervention effect by alerting women to their risk for AEP. Follow-up data may have been subject to distortions caused by the subjects’ desire to please the researchers or social desirability. This concern is the reason for including a control group in this study and for using mailed follow-up questionnaires. Third, because this is a novel intervention being tested for feasibility and promise, this article focuses only on the early impact of the BALANCE intervention. Future work will examine longer-term outcomes and potential mediators of change including psychiatric distress, personality factors, illicit drug use, and readiness to change.

A significant number of college women are at risk for unintended pregnancy and AEP, and these risks should be considered as important targets of prevention programming similar to that provided for excessive drinking. Alcohol-exposed pregnancy prevention interventions for college women provide another avenue for talking with women about binge drinking and ineffective contraception.