Exploring the motives and mental health correlates of intentional food restriction prior to alcohol use in university students

Abstract

This study explored the prevalence of and motivations behind ‘drunkorexia’ – restricting food intake prior to drinking alcohol. For both male and female university students (N = 3409), intentionally changing eating behaviour prior to drinking alcohol was common practice (46%). Analyses performed on a targeted sample of women (n = 226) revealed that food restriction prior to alcohol use was associated with greater symptomology than eating more food. Those who restrict eating prior to drinking to avoid weight gain scored higher on measures of disordered eating, whereas those who restrict to get intoxicated faster scored higher on measures of alcohol abuse.

Keywords

drinking behaviour drunkorexia eating behaviour eating disorders females health behaviour risk factors students

The high comorbidity between alcohol abuse and eating disorders (EDs) has been well documented (National Center on Addiction and Substance Abuse at Columbia University (NCASU), 2003). University years are a high-risk period for both ED symptoms and problematic drinking (Substance Abuse and Mental Health Services Administration (SAMHSA), 2010; Tylka and Subich, 2002). While the prevalence of clinical EDs is relatively low (American Psychiatric Association (APA), 2000), prevalence of subclinical ED symptoms (i.e. chronic dieting, intermittent binging/purging) is much higher among young women, with one study estimating the rate at 60 per cent (Tylka and Subich, 2002). Female dieters who are engaged in alcohol abuse are at increased risk for negative alcohol-related consequences including physical injury and blackouts, compared to those who engage in alcohol use with no history of dieting (Krahn et al., 2005).

Given the high prevalence of binge drinking and disordered eating in Canadian college-aged women (Piran and Robinson, 2011), recent attention has been paid to understanding how female dieters negotiate shape/weight concerns while engaging in alcohol use. ‘Drunkorexia’ is a colloquial term that has been used by the media in the last 5 years (e.g. Kershaw, 2008). Despite debate on a definition and the appropriateness of the term (Barry et al., 2013), it is generally used to describe young women who restrict their food intake prior to drinking alcohol in order to offset the calories in alcohol and also, in some cases, so that they may become intoxicated faster (Barry and Piazza-Gardner, 2012). Despite limited empirical research, this phenomenon has become a popular topic of discussion among media outlets and Internet groups (Chambers, 2008).

Although not a recognized symptom of clinical EDs, caloric restriction prior to alcohol use poses significant health risks, especially for young women (Barry et al., 2013). The concentration of alcohol within the body and the sudden intoxicating effects are amplified by food deprivation (Watkins and Adler, 1993). Repeated episodes of alcohol abuse in combination with food restriction can result in sudden blackouts, alcohol poisoning, increased chances of engaging in risky behaviours and injuries and, over time, cirrhosis of the liver and cognitive decline (National Institute on Alcohol Abuse and Alcoholism (NIAAA), 2004). Women are at greater risk than men for the negative effects of caloric restriction prior to alcohol use because they generally have less water in their body (Baraona, 1998) and less gastric alcohol dehydrogenase activity as compared to men (Frezza et al., 1990), resulting in higher blood ethanol concentrations after consuming the same amount of alcohol.

Little research exists on planned caloric restriction prior to alcohol use. Peralta (2002) studied the phenomenon qualitatively through semi-structured interviews of 78 students and found that 18 per cent of participants endorsed restricting calories primarily to avoid weight gain and, in some cases, to enhance the intoxicating effects of alcohol. Other studies found that approximately 14 per cent of their student samples engage in caloric restriction prior to alcohol use (Burke et al., 2010; Osborne et al., 2011). Overall, these studies have shown that approximately 64–69 per cent of students who engage in caloric restriction prior to drinking do so in order to avoid weight gain and 25 per cent do so in order to get intoxicated faster (Giles et al., 2009; Osborne et al., 2011).

The main objectives of this study were to investigate the prevalence, mental health correlates and motivations behind intentional eating changes prior to alcohol use. Two studies with independent samples were used to explore these objectives.

Study 1

Study 1 investigated the prevalence of dietary changes prior to planned alcohol consumption in a large sample of Canadian university students. Unlike previous studies, we examined all self-reported methods of changing eating behaviour prior to drinking in order to provide a clearer picture of students’ eating behaviour prior to drinking. Study 1 also identified the motivations for changing one’s eating behaviour prior to drinking alcohol. We hypothesized that food restriction prior to drinking would be more common in women (Piazza-Gardner and Barry, 2013), whereas eating more food to avoid sickness or a hangover would be more common in men.

Method

Participants

Participants were undergraduate students from a large, multicultural university located in Toronto, Canada, where the legal drinking age is 19. A total of 3859 students registered in the Fall 2013 semester of Introductory Psychology took part in the study. Of those, 450 participants were excluded because of excessive missing data (>50%) or they were over the age cutoff. The final sample consisted of 3409 students (70% female) aged 16–26 (M = 19.60; standard deviation (SD) = 2.08) years. The ethnic distribution of the sample was 37.1 per cent White, 14.8 per cent East Asian, 21.8 per cent South Asian, 8.4 per cent Black, 2.5 per cent Hispanic, 0.4 per cent Aboriginal and 15 per cent other ethnic identification.

Procedure

This study received ethics approval from the university. Participants were asked to complete an online survey for which they received partial course credit.

Measures

In the absence of any standardized measures of our behaviours of interest, we developed Likert scale items assessing eating habits prior to planned alcohol use based on previous studies (e.g. Burke et al., 2010). First, participants were asked how often they change their eating prior to an occasion where they would be consuming alcohol (‘never’, ‘rarely’, ‘sometimes’, ‘usually’ or ‘always’). A total of 1193 participants were exempt from further questions because they stated that they ‘never’ engage in such behaviour. Second, if applicable, participants were asked how they change their eating prior to drinking alcohol (‘I eat more food’, ‘I eat less food’, ‘I skip a meal’, ‘I skip more than one meal’ or ‘I eat no food’) in order to categorize participants by eating behaviour. Participants who endorsed eating less food, no food or skipping meals were all considered in the ‘eating less’ category. Third, participants were asked to select the primary reason why they change their eating prior to drinking alcohol. Possible reasons included: ‘to avoid becoming sick’, ‘to prevent a hangover’, ‘to avoid gaining weight’, ‘to save calories’, ‘to save money’ and ‘to become intoxicated faster’. Finally, participants were asked to what extent they consider the caloric content in their alcohol (‘not at all’, ‘slightly’, ‘somewhat’ or ‘very much’).

Data analysis

Frequency analyses were used to examine the participants’ self-reported eating behaviours and motivations prior to alcohol use. Chi-Square tests with Cramer’s V as the indicator of effect size (Cramer, 1999) were conducted to compare groups across (1) gender, (2) method (i.e. eating more or less), (3) motivation (i.e. primary reason for changing eating prior to drinking) and (4) level of concern regarding calories in alcohol. Nonparametric Chi-Square tests were used for post-hoc analyses of each group in significant Chi-Square tests.

Results

Eating changes prior to alcohol use in first-year university students

A total of 46 per cent of the sample selected at least ‘sometimes’ intentionally changing their food intake before an occasion where they knew that they would be drinking alcohol (5% always, 14% usually, 27% sometimes, 19% rarely and 35% never). Of those participants who reported that they changed their eating prior to drinking (n = 1599), 63 per cent reported eating more and 37 per cent reported eating less food. Of those who reportedly eat less food (n = 591), a small number of participants endorsed skipping meals (6%) or eating no food all day (1%) prior to an occasion where they will be drinking alcohol (Figure 1(a)). The most common reasons given for changing one’s eating prior to drinking were to avoid becoming sick (49%) and to prevent a hangover (31%). A total of 10 per cent of participants reported that they changed their eating prior to alcohol consumption in order to avoid weight gain/save calories, 8 per cent endorsed doing so to get intoxicated faster, and 2 per cent reported that they wanted to save money (Figure 1(b)).

Figure 1.

(a) Eating changes prior to alcohol use in university students by gender, (b) primary motivation for eating changes prior to alcohol use in university students by gender and (c) primary motivation as a function of eating changes prior to alcohol use (n = 1519).

Pearson Chi-Square analyses results revealed that participants who endorsed eating more food prior to drinking alcohol reported different primary motivations for changing their eating than did those who endorsed eating less food (Figure 1(c)), yielding a large effect size (χ²(4, 1519) = 506.08, p < .001, V = 0.60). Specifically, participants who reported eating more food prior to drinking alcohol were significantly more likely to endorse their primary motivations as either ‘to avoid becoming sick’ (χ²(1, 764) = 154.96, p < .001) and ‘to prevent a hangover’ (χ²(1, 485) = 262.78, p < .001), whereas those who reported eating less food prior to drinking alcohol were significantly more likely to choose their primary motivations as ‘to avoid gaining weight’ (χ²(1, 139) = 112.41, p < .001) and ‘to become intoxicated faster’ (χ²(1, 122) = 114.13, p < .001).

In terms of gender differences, although there were no differences in the frequency in which men and women changed their eating prior to drinking, results revealed significant differences in whether they tended to eat more or less food prior to planned alcohol use, with a small effect size (χ²(1, 1599) = 7.97, p = .005, V = 0.07). As shown in Figure 1(a), men reported eating more food prior to alcohol use at rates greater than expected (χ²(1, 485) = 69.05, p < .001), whereas women were more likely to report eating less food (χ²(1, 1114) = 63.52, p < .001). Women and men also differed in their primary reasons for changing their food intake prior to alcohol use, yielding a moderate effect size (χ²(4, 1592) = 41.71, p < .001, V = 0.15). Women were more likely to endorse their primary motivations as ‘to avoid gaining weight’ (χ²(1, 151) = 70.26, p < .001), ‘to get intoxicated faster’ (χ²(1, 122) = 29.51, p < .001) and ‘to avoid becoming sick’ (χ²(1, 780) = 160.67, p < .001) when drinking. Men were more likely to select ‘to prevent a hangover’ (χ²(1, 504) = 28.57, p < .001) as their primary motivation (Figure 1(b)).

Participants’ level of concern for calories in alcohol

A total of 42 per cent of participants were at least ‘slightly’ concerned with the calories in their alcohol, with 7 per cent being ‘very much’ concerned. Pearson Chi-Square analyses revealed that women reported significantly greater concern for the calories in their alcohol than men, with a small effect size (χ²(3, 3187) = 10.40, p = .015, V = 0.06). Specifically, women were more likely to select that they were ‘rarely’ (χ²(1, 574) = 119.59, p < .001), ‘slightly’ (χ²(1, 424) = 69.78, p < .001) or ‘very much’ concerned (χ²(1, 786) = 47.51, p < .001), whereas men were more likely to select being ‘not at all concerned’ (χ²(1, 1673) = 196.25, p < .001). Concern for the amount of calories in alcohol also varied as a function of primary motivation for changing one’s eating prior to alcohol use, with a medium effect size (χ²(12, 1786) = 163.19, p < .001, V = 0.18). In addition, 58 per cent of those whose primary motivation was to ‘avoid gaining weight’ (χ²(3, 169) = 10.91, p < .001) were ‘somewhat’ to ‘very much’ concerned with the caloric intake in their alcohol, whereas 77 per cent of those whose primary motivation was to ‘avoid becoming sick’ (χ²(3, 868) = 506.65, p < .001) or ‘avoid a hangover’ (χ²(3, 564) = 231.09, p < .001) were ‘not at all’ to ‘slightly’ concerned with the calories in their alcohol. In addition, 71 per cent of those whose primary motivation was to ‘become intoxicated faster’ were ‘not at all’ to ‘slightly’ concerned with the calories in their alcohol (χ²(3, 138) = 61.30, p < .001).

Study 2

The objectives of Study 2 were to focus more closely on the phenomenon of ‘drunkorexia’ and to understand the mental health correlates of and motivations behind food restriction prior to alcohol use among women. While not aiming to uncover cause-and-effect, we explored how mood is associated with disordered eating, alcohol-related problems and women’s motivations to eat differently prior to alcohol use. Given the association between disordered eating and alcohol abuse (Piran and Robinson, 2011), we predicted that young women who eat less prior to alcohol use would report higher levels of disordered eating, alcohol problems, depression and anxiety than those who eat more food. Another objective of Study 2 was to compare individuals across their primary motivation to change their eating prior to alcohol use on measures of disordered eating, alcohol problems and mood (see supplemental material). Based on Study 1’s findings, we predicted that food restriction for the purpose of avoiding weight gain would be associated with higher levels of disordered eating, and that the motive of becoming intoxicated faster would be associated with higher levels of alcohol problems as compared to a control group (i.e. those who eat more food).

Method

Participants

In the Fall of 2012, we recruited a targeted sample of women, independent from Study 1, who reported regularly changing their eating prior to drinking alcohol. A total of 3484 students enrolled in Introductory Psychology at York University were eligible to participate and were prescreened. Of those, women who endorsed changing their eating prior to drinking alcohol were made aware of the study (1105 eligible) and a total of 270 women volunteered to participate. Participants were excluded for excessive missing data (>50%) and who were over the age cutoff of 26. Women who indicated that they change their eating in order to ‘save money’ were excluded from analyses due to very small numbers (n = 2). In addition, participants who indicated unusual or idiosyncratic responses to their primary motivation were excluded (e.g. eat less in order to prevent a hangover). The final sample consisted of 226 women with a mean age of 19.69 years (range = 17–26; SD = 2.07 years). The ethnic distribution of the sample was 50.9 per cent White, 24.9 per cent Asian, 9.7 per cent Middle Eastern, 6.2 per cent Black, 3 per cent Hispanic, 1.3 per cent Aboriginal and 4 per cent other ethnic identification. The majority of participants lived off-campus with their parents or guardians (74%). The remaining participants lived either off-campus with roommates/by themselves (12%) or on-campus in residence (14%). In terms of overall alcohol use, approximately 16 per cent of the sample met criteria for the Rutgers’ Alcohol Problem Index’s (RAPI) suggested cutoff for possible alcohol abuse (Danielson et al., 2003). About 8 per cent of the sample reported engaging in binge drinking (i.e. drinking four or more drinks on one occasion at least once per week), whereas 40 per cent of participants endorsed binge drinking one to two times per month.

Procedure

For Study 2, the same procedure was followed as in Study 1 with the following differences. Selective sampling was used in order to maximize the likelihood of obtaining participants who engaged in dietary changes prior to planned alcohol use. We excluded men from this study in order to focus on those participants who were considered the highest risk for engaging in the practice of food restriction prior to alcohol use (Krahn et al., 2005). Eligible participants were selected based on their responses to a prescreen question that was administered to all first-year psychology students: ‘How often do you change your eating before an occasion where you will be drinking alcohol’? Participants who endorsed at least ‘sometimes’ changing their eating prior to drinking alcohol were invited to participate in the study and those who volunteered were included in the sample. This criterion was used to capture individuals who would be able to accurately identify their motives for doing so.

Measures

Eating behaviour

Participants were given the same set of questions with regard to eating changes prior to alcohol use as in Study 1 with a few exceptions. Participants in Study 2 were invited to provide alternate reasons, besides those in our questionnaire, as to how and why they change their eating prior to engaging in planned alcohol use.

Dietary restraint

The Restraint Scale–Revised (RRS) (Polivy et al., 1988) was used to measure participant’s level of restrained eating. It consists of 10 scored Likert-scale items that measure a participant’s tendency to restrict food and worry about body size/shape. The RRS has been shown to be reliable and valid (e.g. Allison et al., 1992). A total score was used in order to capture the full-range of behaviours and attitudes. The measure demonstrated good internal consistency in the current sample (α = .86).

Eating disorder symptoms

The Eating Disorder Examination–Questionnaire (EDE-Q) (Fairburn and Beglin, 1994) is a widely used 36-item self-report measure that assesses the presence and severity of eating pathology over the past 28 days. In this study, we used the global scale only as a measure of disordered eating. The scale has shown high internal consistency and criterion validity (Hilbert et al., 2007). In the current sample, the internal consistency of the global EDE-Q score was very good (α = .90).

Depression

The Beck Depression Inventory-II (BDI-II) (Beck et al., 1996) was used as a measure of depressive symptoms over the past 2 weeks. Higher scores are indicative of more severe depression as suggested by the BDI-II’s use of Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV) criteria for depression. The high internal consistency, reliability and discriminant validity has been confirmed (e.g. Dozois et al., 1998). The internal consistency in this study was found to be very good (α = .92).

Anxiety

The Beck Anxiety Inventory (BAI) (Beck and Steer, 1990) is a 21-item self-report measure of anxiety symptoms over the past month. The BAI has shown good reliability and validity (Beck and Steer, 1990). The internal consistency in our sample was 0.92.

Alcohol-related problems

Based on the guidelines for females outlined in the NIAAA (2004), binge drinking frequency in females was assessed by asking participants how often they drink four or more standard alcoholic drinks (14 g) in a row on one occasion. The RAPI (White and Labouvie, 1989) was chosen as a measure of alcohol problems because it assesses the frequency that students experience alcohol-related problems (e.g. ‘Not able to do your homework’). A score above 15 is normally used as a cutoff to identify problematic drinking (Danielson et al., 2003). The internal consistency in this sample was very good (α = .93).

Data analysis

Analysis of variance (ANOVA) tested the hypothesis that restricting food prior to alcohol use is associated with higher levels of disordered eating, alcohol problems and higher symptoms of depression and anxiety, as compared to women who eat more food prior to alcohol use. Next, multivariate analysis of variance (MANOVA) was used to compare groups differing in their primary motivation to alter eating prior to alcohol use (1. women who eat more food to prevent sickness/hangover, 2. women who eat less food to avoid weight gain, and 3. women who eat less food to get intoxicated faster) across the dependent measures (disordered eating, dietary restraint, alcohol problems, depression and anxiety). Prior to analyses, one-way ANOVAs were conducted to confirm no pre-existing groups differences in terms of mean age and self-reported body mass index (BMI). Violations to the assumptions of homogeneity of variance test were corrected using the Welch statistic (F_w). A log transformation was applied to the dependent measure of alcohol-related problems (RAPI). Skewness was decreased from 2.34 to 0.04. Post-hoc analyses used the Scheffe test. There were no missing data. Effect sizes were interpreted using Cohen’s d (Cohen, 1988).

Results

Comparison of women who eat less food versus more food prior to alcohol consumption

A significant main effect of type of eating behaviour change (eating either more or less) was found on the dependent measures of interest (Table 1). Specifically, participants who reported eating less food prior to drinking scored significantly higher than did those who reported eating more on ED symptoms, dietary restraint, anxiety symptoms, depressive symptoms and alcohol problems.

Table 1.

Study 2 – main effect of type of eating behaviour change on restraint, eating disorder symptoms, anxiety, depression and alcohol abuse (n = 226).

Measure	Dietary behaviour prior to alcohol consumption	Mean (SD)	95% CI	Significance
Restraint**	More food (n = 114)	13.75 (6.39)	12.57–14.94	F (1, 224) = 20.04, p < .001, η² = 0.09
Restraint**	Less food (n = 112)	19.25 (6.43)	18.04–20.45	F (1, 224) = 20.04, p < .001, η² = 0.09
EDE-Q**	More food (n = 114)	1.50 (1.13)	1.29–1.71	F (1, 224) = 41.53, p < .001. η² = 0.17
EDE-Q**	Less food (n = 112)	2.16 (1.09)	1.96–2.37	F (1, 224) = 41.53, p < .001. η² = 0.17
BAI*	More food (n = 114)	14.59 (9.51)	12.82–16.35	F_w (1, 212.70) = 8.03, p = .005, η² = 0.03
BAI*	Less food (n = 112)	18.63 (11.81)	16.42–20.84	F_w (1, 212.70) = 8.03, p = .005, η² = 0.03
BDI-II*	More food (n = 114)	13.36 (8.14)	11.85–14.87	F_w (1, 195.58) = 5.88, p = .016, η² = 0.03
BDI-II*	Less food (n = 112)	16.66 (11.93)	14.42–18.89	F_w (1, 195.58) = 5.88, p = .016, η² = 0.03
logRAPI**	More food (n = 114)	0.61 (0.45)	0.53–0.69	F (1, 224) = 12.76, p < .001, η² = 0.06
logRAPI**	Less food (n = 112)	0.83 (0.48)	0.74–0.92	F (1, 224) = 12.76, p < .001, η² = 0.06

Restraint: The Restraint Scale–Revised; EDE-Q: Eating Disorder Examination–Questionnaire; BAI: Beck Anxiety Inventory; BDI-II: Beck Depression Inventory-II; RAPI: Rutger’s Alcohol Problem Index.

95% CI = 95% confidence interval (lower bound–upper bound).

Significance at the <.05 level; **Significance at the <.001 level.

Comparison of group differences between primary motivations to change food intake

The MANOVA revealed a significant main effect and large effect size of primary motivation to engage in dietary changes prior to drinking alcohol (Wilks’ λ = 0.79, F(10, 438) = 5.57, p < .001, η² = 0.11). In comparison to participants whose primary motivation was ‘to avoid becoming sick or prevent a hangover’, post-hoc analyses revealed that women whose primary motivation was ‘to avoid weight gain’ had significantly higher scores on ED symptoms and dietary restraint, both yielding a large effect size. On the other hand, women whose primary motivation was ‘to become intoxicated faster’ had significantly higher scores on alcohol problems and depressive symptoms (Table 2).

Table 2.

Study 22 – main effect of primary motivation to engage in dietary changes prior to drinking Alcohol (n = 226).

Measure	Motivation	Mean (SD)	95% CI	Significance
Restraint	To avoid sickness/hangover (n = 131)	14.10 (6.61)	12.96–15.24	20.44 > 14.10**
	To avoid weight gain (n = 73)	20.44 (6.02)	19.03–21.84	F(2, 223) = 23.66, p < .001, η² = 0.18
	To get drunk faster (n = 22)	17.50 (5.82)	14.92–20.08	F(2, 223) = 23.66, p < .001, η² = 0.18
EDE-Q	To avoid sickness/hangover (n = 131)	1.53 (1.11)	1.34–1.72	2.37 > 1.53**
	To avoid weight gain (n = 73)	2.37 (1.10)	2.11–2.63	F(2, 223) = 13.73, p < .001, η² = 0.11
	To get drunk faster (n = 22)	1.80 (1.03)	1.34–2.26	F(2, 223) = 13.73, p < .001, η² = 0.11
BAI	To avoid sickness/hangover (n = 131)	15.19 (9.98)	13.47–16.92	n.s.
	To avoid weight gain (n = 73)	18.30 (11.91)	15.52–21.08
	To get drunk faster (n = 22)	19.27 (11.61)	14.13–24.42
BDI-II	To avoid sickness/hangover (n = 131)	13.52 (8.56)	12.04–15.00	19.09 > 13.52*
	To avoid weight gain (n = 73)	16.41 (12.24)	13.55–19.27	F(2, 223) = 3.87, p = .022, η² = 0.03
	To get drunk faster (n = 22)	19.09 (11.50)	13.99–24.19	F(2, 223) = 3.87, p = .022, η² = 0.03
logRAPI	To avoid sickness/hangover (n = 131)	0.64 (0.45)	0.57–0.72	0.93 > 0.64*
	To avoid weight gain (n = 73)	0.79 (0.50)	0.67–0.90	F(2, 223) = 4.75, p = .010, η² = 0.04
	To get drunk faster (n = 22)	0.93 (0.50)	0.72–1.15	F(2, 223) = 4.75, p = .010, η² = 0.04

Restraint: The Restraint Scale–Revised; EDE-Q: Eating Disorder Examination–Questionnaire; BAI: Beck Anxiety Inventory; BDI: Beck Depression Inventory-II; RAPI: Rutger’s Alcohol Problem Index.

95% CI = 95% Confidence Interval (Lower Bound–Upper Bound).

Significance at the <.05 level; **Significance at the <.001 level; n.s. = no significance.

Discussion

Alcohol abuse and disordered eating are both common among university students and a better understanding of how and why they co-occur is needed. Media reports on the popularity of the practice of ‘drunkorexia’ whereby young women intentionally restrict their food intake prior to drinking in order to save calories and avoid weight gain has, in fact, only limited empirical support. The current set of studies sought to provide some clarity about the extent of this behaviour. This is the first known study to examine how changes to eating behaviour prior to alcohol consumption and the motivations behind them relate to mental health, eating behaviours and alcohol-related problems. This study built on previous studies by examining both individuals who report eating less and those who report eating more food prior to planned alcohol use, as a comparison group.

Nearly half (46%) of the general university student sample reported that they at least sometimes modify their eating prior to a planned episode of drinking. Students tended to eat more food rather than less food prior to drinking. Overall, eating more in order to avoid getting sick or to avoid a hangover were the most common motives. Consistent with our hypothesis, this was particularly true for male students. Women were more concerned than men with the calories in their alcohol and their eating tended to reflect this in terms of a significantly higher frequency of reported food restriction prior to drinking. Our study provides confirmatory evidence that women often negotiate their desire to avoid weight gain from drinking through intentional restriction, which is consistent with the findings of Barry et al. (2013) and Eisenberg and Fitz (2014).

Our findings are also consistent with other studies, in that approximately 18 per cent of our students engage in intentional caloric restriction prior to alcohol use (Burke et al., 2010; Osborne et al., 2011). For the women in Study 2, intentionally restricting food intake prior to alcohol use was associated with greater disordered eating, alcohol problems and symptoms of depression and anxiety. Furthermore, the differences found between eating less and more food prior to alcohol use were further broken down by motivation. Women who were concerned with the caloric content in their alcohol and reported that they reduce their food intake prior to drinking in order to avoid weight gain exhibited higher levels of disordered eating. On the other hand, women who reported that they reduce their food intake in order to become intoxicated faster displayed higher levels of alcohol problems and were relatively unconcerned with the calories in their alcohol. These women also displayed higher levels of depression, although the effect size was small.

Taken together, these studies highlight the need to assess motivations for altering food intake prior to alcohol use as part of clinical intervention. For example, women whose primary motivation for eating less food prior to drinking is to avoid weight gain may benefit more from interventions focused on healthy eating and promoting positive body image. In contrast, our findings suggest that women who wish to get drunk faster, and who intentionally eat less prior to drinking to facilitate this effect, may have multiple difficulties associated with symptoms of alcohol abuse and potentially other mental health concerns (e.g. depression). Although not tested in the present study, it is possible that these difficulties may then contribute to feelings of wanting to escape or an attempt to enhance their mood through alcohol use (Hull, 1981).

A somewhat unexpected finding was the high number of participants (30%) who reported eating more food prior to drinking alcohol. Lloyd-Richardson et al. (2008) found that 13–22 per cent of students who engage in binge drinking reported eating more food prior to alcohol use. From a harm-reduction perspective, eating more food prior to drinking would be an adaptive behaviour, given that food slows the absorption of alcohol into the bloodstream and would mitigate the effects of alcohol on one’s system (Watkins and Adler, 1993). Study 2 further demonstrated that eating more food prior to drinking was negatively associated with psychological distress, alcohol problems and disordered eating. Some alcohol safety workshops offered by colleges encourage students to eat a substantial meal prior to drinking alcohol (Dimeff et al., 1999). However, little is known about the optimal amount of food that should be eaten prior to alcohol use in addition to any potential negative consequences, including health risks for obesity. The association between weight gain and overeating while intoxicated has been well documented (e.g. Lloyd-Richardson et al., 2008). Our findings suggest the need to consider overeating behaviours prior to alcohol use as well as after drinking. More research is also needed to determine how eating more food prior to alcohol consumption then impacts drinking behaviours. For example, in Lloyd-Richardson et al.’s (2008) study, college students who were at-risk for alcohol abuse were significantly more likely to report eating more food prior to drinking than were low-risk drinkers.

Interestingly, there was a sizeable minority of participants who reported in the comments section of our survey very specific practices for avoiding a hangover or becoming sick (e.g. ‘I eat more carbohydrates’); however, these self-made remedies were not based on medical recommendations and sometimes were both dangerous and counterintuitive (e.g. eating less food to avoid a hangover). Although traditional programmes for university students focused on harm-reduction are important initiatives (e.g. drinking water, buddy-system), there has been limited literature examining the intersection between eating food and drinking alcohol. Martens et al. (2004) discuss the importance of future programmes focusing on the prevention of other behaviours associated with alcohol use in young women in order to prevent alcohol-related problems (e.g. disordered eating, depression).

Limitations and future research

We used a sample of convenience and our findings may not generalize to other populations, including those with clinical eating and alcohol use disorders. Although the majority of our student sample lived off-campus with their parents or guardians, the percentage of students who eat less prior to drinking was comparable to previous studies where the majority of participants were living on-campus (e.g. Burke et al., 2010). Also, the drinking rates of the Study 2 sample are comparable to other Canadian studies of binge drinking rates (e.g. Archie et al., 2012). Future studies, should explore how drinking frequency impacts on caloric restriction, mental health status and unhealthy/risky behaviours. We used self-report measures of mental health, problematic drinking and eating. Diagnostic interviews can confirm the presence and severity of symptoms. The current findings are only correlational in nature and cannot determine cause-and-effect relationships. Future research should consider other research designs, including experimental methods and ecological momentary assessment. More research is needed to explore how caloric restriction prior to drinking impacts other behaviours, such as typical drinking patterns and eating behaviours after drinking. Finally, there is a need for standardized definitions and measures of food restriction prior to alcohol use, including agreement on language use (Piazza-Gardner and Barry, 2013) and how much food restriction and/or alcohol use should be considered significant or meaningful (Rahal et al., 2012).

Conclusion

Intentional changes to one’s eating prior to alcohol use are a common behaviour in first-year university students. Food restriction prior to drinking is a high-risk health behaviour and is especially dangerous for women due to sex differences in alcohol metabolism. Among female university students, intentionally eating less prior to drinking alcohol is associated with negative mental health symptoms, including alcohol-related problems, disordered eating, depression and anxiety. Young women whose primary motivation is to avoid weight gain scored higher on measures of disordered eating, whereas those whose primary motivation is to get drunk faster scored higher on measures of alcohol problems. Intentional food restriction prior to alcohol use is clearly a maladaptive behaviour, but the motive matters. These studies highlight the need for individualized treatment/prevention strategies as well as harm-reduction programmes exploring healthy eating strategies prior to alcohol use.

Footnotes

Acknowledgements

The authors would like to thank Iris Sijercic for her assistance with data collection and preparation, Constance Mara for her statistical advice, and Rachel Vella-Zarb and Leah Keating for their comments on an early draft of this article.

Funding

This work was supported in part by the first author’s (K.M.R.) Vanier Canada Graduate Scholarship: Canadian Institutes of Health Research Doctoral Award and the second author’s (J.S.M.) York University Faculty of Health Minor Research Grant.

References

Allison

Kalinsky

Gorman

(1992) A comparison of the psychometric properties of three measures of dietary restraint. Psychological Assessment 4(3): 391–398.

American Psychiatric Association (APA) (2000) Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) (4th edn). Text revision. Washington, DC: APA.

Archie

Zangeneg Kazemi

Akhtar-Danesh

(2012) Concurrent binge drinking and depression among Canadian youth: Prevalence, patterns, and suicidality. Alcohol 46(2): 165–172.

Baraona

(1998) Alcohol metabolism. In: Blechman

Brownell

(eds) Behavioral Medicine and Women: A Comprehensive Handbook. New York: Guilford Press, pp. 407–413.

Barry

Piazza-Gardner

(2012) Drunkorexia: Understanding the co-occurrence of alcohol consumption and eating/exercise weight management behaviors. Journal of American College Health 60(3): 236–243.

Barry

Whiteman

Piazza-Gardner

. (2013) Gender differences in the associations among body mass index, weight loss, exercise, and drinking among college students. Journal of American College Health 61(7): 407–413.

Beck

Steer

(1990) Manual for the Beck Anxiety Inventory. San Antonia, TX: Psychological Corporation.

Beck

Steer

Brown

(1996) Manual for the Beck Depression Inventory. San Antonia, TX: Psychological Corporation.

Burke

Cremeens

Vail-Smith

. (2010) Drunkorexia: Calorie restriction prior to alcohol consumption among college freshman. Journal of Alcohol and Drug Education 54(2): 17–34.

10.

Chambers

(2008) Drunkorexia. Journal of Dual Diagnosis 4(4): 414–416.

11.

Cohen

(1988) Statistical Power Analysis for the Behavioral Sciences. Hillsdale, NJ: Lawrence Erlbaum Associates.

12.

Cramer

(1999) Mathematical Methods of Statistics. Princeton, NJ: Princeton University Press.

13.

Danielson

Overholser

Butt

(2003) Association of substance abuse and depression among adolescent psychiatric inpatients. Canadian Journal of Psychiatry 48(11): 762–765.

14.

Dimeff

Baer

Kivlahan

. (1999) Brief Alcohol Screening and Intervention for College Students. New York: The Guilford Press.

15.

Dozois

DJA

Dobson

Ahnberg

(1998) A psychometric evaluation of the beck depression inventory-II. Psychological Assessment 10(2): 83–89.

16.

Eisenberg

Fitz

(2014) ‘Drunkorexia’: Exploring the who and why of a disturbing trend in college students’ eating and drinking behaviors. Journal of American College Health 62(8): 570–577.

17.

Fairburn

Beglin

(1994) Assessment of eating disorder psychopathology: Interview or self-report questionnaire? International Journal of Eating Disorders 16: 363–370.

18.

Frezza

di Padova

Pozzato

. (1990) High blood alcohol levels in women. The New England Journal of Medicine 322(2): 95–99.

19.

Giles

Champio

Sutfin

. (2009) Calorie restriction on drinking days: An examination of drinking consequences among college students. Journal of American College Health 57: 603–609.

20.

Hilbert

Tuschen-Caffier

Karwautz

. (2007) Eating Disorder Examination–Questionnaire: Evaluation der deutschsprachigen Übersetzung. Diagnostica 53(3): 144–154.

21.

Hull

(1981) A self-awareness model of the causes and effects of alcohol consumption. Journal of Abnormal Psychology 90(6): 586–600.

22.

Kershaw

(2008) Starving themselves, cocktail in hand. New York Times, 2 March. Available at: http://www.nytimes.com/2008/03/02/fashion/02drunk.html?pagewanted=all&_r=0

23.

Krahn

Kurth

Gomberg

. (2005) Pathological dieting and alcohol use in college women – A continuum of behaviors. Eating Behaviors 6: 43–52.

24.

Lloyd-Richardson

Lucero

DiBello

. (2008) The relationship between alcohol use, eating habits and weight change in college freshmen. Eating Behaviours 9: 504–508.

25.

Martens

Taylor

Damann

. (2004) Protective factors when drinking alcohol and their relationship to negative alcohol-related consequences. Psychology of Addictive Behaviors 18(4): 390–393.

26.

National Center on Addiction and Substance Abuse at Columbia University (NCASU) (2003) Food for thought: Substance abuse and eating disorders. Available at: http://www.casacolumbia.org/templates/Publications_Reports.aspx#r25 (accessed 13 June 2013).

27.

National Institute on Alcohol Abuse and Alcoholism (NIAAA) (2004) National Institute on Alcohol Abuse and Alcoholism Council approves definition of binge drinking. NIAAA Newsletter. Available at: http://pubs.niaaa.nih.gov/publications/Newsletter/winter2004/Newsletter_Number3.htm (accessed 13 June 2013).

28.

Osborne

Sher

Winograd

(2011) Disordered eating patterns and alcohol misuse in college students: Evidence for ‘drunkorexia’? Comprehensive Psychiatry 52(6): E12.

29.

Peralta

(2002) Alcohol use and the fear of weight gain in college: Reconciling two social norms. Gender Issues 20(4): 23–42.

30.

Piazza-Gardner

Barry

(2013) Appropriate terminology for the alcohol, eating, and physical activity relationship. Journal of American College Health 61(5): 311–313.

31.

Piran

Robinson

(2011) Patterns of associations between eating disordered behaviors and substance use in tow non-clinical samples: A university and a community sample. Journal of Health Psychology 16(7): 1027–1037.

32.

Polivy

Herman

Howard

(1988) The Restraint Scale: Assessment of dieting. In: Hersen

Bellack

(eds) Dictionary of Behavioral Assessment Techniques. New York: Pergamon Press, pp. 377–380.

33.

Rahal

Bryant

Darkes

. (2012) Development and validation of the Compensatory Eating and Behaviors in Response to Alcohol Consumption Scale (CEBRACS). Eating Behaviors 13: 83–87.

34.

Substance Abuse and Mental Health Services Administration (SAMHSA) (2010) Results from the 2010 National Survey on Drug Use and Health: National findings. Available at: http://www.oas.samhsa.gov/nsduhLatest.htm (accessed 13 June 2013).

35.

Tylka

Subich

(2002) Exploring young women’s perceptions of the effectiveness and safety of maladaptive weight control techniques. Journal of Counseling and Development 80(1): 101–110.

36.

Watkins

Adler

(1993) Effect of food on alcohol absorption and elimination patterns. Journal of Forensic Sciences 38(2): 285–291.

37.

White

Labouvie

(1989) Towards the assessment of adolescent problem drinking in adolescence. Journal of Studies on Alcohol and Drugs 50: 30–37.