Skip to main content
PMC home page
PMC Canada Author Manuscripts logoLink to PMC Canada Author Manuscripts
. Author manuscript; available in PMC: 2016 Jun 13.
Published in final edited form as: J Stud Alcohol Drugs. 2015 Jan;76(1):95–105.

The Relationship of Higher Education to Substance Use Trajectories: Variations as a Function of Timing of Enrollment

KARA THOMPSON a,*, JACQUELINE HOMEL b, BONNIE LEADBEATER b
PMCID: PMC4905754  CAMSID: CAMS5730  PMID: 25486398

Abstract

Objective

This study examined the association between time to enrollment into postsecondary education and trajectories of heavy episodic drinking (HED) and marijuana use using a prospective longitudinal study.

Method

Participants included 391 postsecondary students (55% female) drawn from the Victoria Healthy Youth Survey, a five-wave, multi-cohort sample interviewed biennially between 2003 and 2011. Using piecewise latent growth modeling, we compared changes in the trajectories of HED and marijuana use before and after postsecondary enrollment across three groups of young adults: (a) direct entrants (enrolled directly out of high school), (b) gap entrants (took a year off), and (c) delayed entrants (took longer than a year off).

Results

Heavy drinking increased after enrollment for direct entrants and gap entrants and decreased for delayed entrants. Marijuana use increased after enrollment for direct entrants, and decreased for gap entrants and delayed entrants. Yet, overall levels of marijuana use were significantly higher among the gap and delay entrants over time compared with direct entrants. Group differences in heavy drinking appeared to reflect age-related changes in drinking patterns. However, differences in marijuana use may reflect pre-existing inequities in access to higher education across groups.

Conclusions

The association between postsecondary education and increased substance use may be limited to students who enroll at a postsecondary institution directly out of high school. However, students who delay enrollment have higher levels of substance use before enrollment, as well as lower high school grades and socioeconomic status compared with direct entrants, and may be particularly vulnerable to long-term substance use problems and degree noncompletion.

Considerable research has shown that going to college is associated with significant increases in alcohol and marijuana use during the emerging adult years (Bachman et al., 2002; Carter et al., 2010; Fleming et al., 2012; Schulenberg & Patrick, 2011). However, increased substance use is not uniform across all postsecondary students; rather, heterogeneity in contexts such as living situation, type of postsecondary institution, and part-time versus full-time status moderate this association (Carter et al., 2010). We argue that student differences in the time of entry into postsecondary education (PSE) may also affect the association between postsecondary enrollment and increased substance use. The purpose of the current study was to examine whether the expected increases in heavy drinking and marijuana use after enrollment in postsecondary institutions are consistent across students who vary in the time taken to enroll in PSE after high school graduation.

Timing of postsecondary education

At the population level, the timing of the transition to PSE is well defined because it is contingent on the completion of high school and is often a prerequisite for entrance into professional careers. Time to enrollment in PSE is also strongly influenced by expectations of normative adult timetables. For most young adults, the pursuit of PSE precedes other life course transitions such as employment, marriage, and parenthood (Crockett & Beal, 2012; Oesterle et al., 2010). Consistent with these age expectations, more than 75% of postsecondary students in Canada and the United States are between 17 and 24 years old (American College Health Association, 2013a, 2013b; Statistics Canada, 2010). However, not all students enroll in PSE directly following high school completion.

Today, approximately one third of postsecondary students delay their entry into PSE following high school graduation by at least 1 year (Hango, 2011; National Centre for Education Statistics, 2005), and this trend of taking time off between leaving high school and enrolling in PSE has increased internationally (Bozick & DeLuca, 2005; Cammelli et al., 2011; Hango, 2011; Holmlund et al., 2008). In Canada, approximately 50% of youth who enrolled in PSE did so within 3 months, 73% within 15 months, and 81% within 28 months (Hango, 2011). Similar rates are reported in studies in the United Kingdom (Crawford & Cribb, 2012), the United States (Bozick & DeLuca, 2005), and Australia (Curtis et al., 2012).

Young people delay enrollment for a variety of reasons. Delayed enrollment may reflect practical considerations, such as the need to accumulate savings to offset the increasing costs of PSE (Goldrick-Rab & Han, 2011). Others take time off because they want to “take a break,” do not feel developmentally mature enough or ready for PSE, or are not academically prepared (Curtis et al., 2012; Hango, 2011; O’Shea, 2013). However, research to date also suggests that delayed enrollment may reflect individual differences in the characteristics of students who delay entry and those who enroll on time, and these characteristics may close off opportunities for direct enrollment. For example, compared with students who enter on time, students who delay tend to have lower high school grades, parents with no PSE, and lower educational aspirations and are more likely to transition into adult roles at an early age (i.e., parenthood, marriage) and enroll at a 2-year rather than a 4-year institution (Bozick & DeLuca, 2005; Hango, 2011; Roksa & Velez, 2012).

Despite the growing popularity of delaying PSE enrollment among young adults, there is mounting debate regarding the value or impact of these experiences. Proponents of delayed enrollment suggest that taking time off after high school can help create better prepared and more highly motivated students because it offers opportunity for self-reflection and career clarification (Martin, 2010; O’Shea, 2013). Some research has found that taking a year off (a “gap-year”) between high school and PSE has a positive impact on student academic performance in college (Birch & Miller, 2007; Crawford & Cribb, 2012). However, delaying enrollment by 6 or more months has also been associated with a higher risk of degree noncompletion (Bozick & Deluca, 2005; Roksa & Velez, 2012) and lower lifetime earnings (Crawford & Cribb, 2012; Holmlund et al., 2008), even after accounting for pre-existing differences in sociodemographic and academic characteristics between students who delay enrollment and those who do not.

Substance use and the timing of developmental transitions

Research shows that going to college is associated with a significant increase in heavy episodic drinking (HED) and marijuana use between ages 18 and 21 years, and then a decline in use between ages 21 and 25 years (Schulenberg & Patrick, 2011). However, it is unclear whether enrollment in college similarly affects students who enroll directly out of high school at age 18 and students who delay their enrollment. Research has investigated associations between alcohol use and the timing of other role transitions (Chassin et al., 2013), such as marriage (Bogart et al., 2005) and parenthood (Little et al., 2009), but not PSE. Findings suggest that marriage tends to be associated with declines in substance use regardless of its timing (Bachman et al., 1997; Bogart et al., 2005; Chassin et al., 2013). However, Little et al. (2009) showed that the timing of parenthood had an important effect on discontinuity in alcohol use. Young adult parents experienced expected declines in alcohol use following parenthood, but adolescent parents experienced increases in alcohol use following parenthood (Little et al., 2009). However, little is known about patterns of substance use in relation to the timing of postsecondary enrollment—for example, whether enrollment will be associated with increases in substance use for students who take a year off after high school, as well as those who enroll directly.

The current study builds on our previous research that investigated how type of postsecondary institution and age at the time of enrollment influenced alcohol use trajectories (Thompson et al., in press). We found no differences in the trajectories of alcohol use between students who attended 2-year colleges, 4-year universities, or transfer programs (start at a 2-year college and transfer to a 4-year university) after accounting for student age at the time of enrollment. This study showed that students who were younger when they enrolled in PSE had steeper increases in heavy drinking after enrollment than older students. Students who are older than 18 years when they enroll in PSE have, by definition, taken time off between high school graduation and PSE. This developmental difference between students who enroll directly out of high school and students who delay enrollment may reduce the impact of college enrollment on increases in substance use. In this study, we take the next step by specifically investigating the differences in developmental pathways of both heavy drinking and marijuana use between students who enroll directly after high school, students who delay enrollment for 1 year, and students who delay longer than a year. This work will provide an understanding of the impact of college enrollment for groups of students who vary in timing of enrollment in college.

Despite limited research on the link between time to PSE enrollment and substance use, from a theoretical perspective, there are a number of reasons to expect that young adults who delay PSE enrollment are less likely to experience the typical increases in substance use than students who enroll immediately after high school. By definition, students who delay enrollment are older than students who enroll directly after high school. Thus, delayed entrants’ transition to PSE is more likely to be intertwined with other life course transitions. Recent studies have shown that students who delay enrollment longer than 6 months after leaving high school are more likely to have children and be cohabiting or married while in school (Bozick & DeLuca, 2005; Roksa & Velez, 2012). Research shows that substance use begins to decline with the onset of parenthood and marriage during emerging adulthood (Bachman et al., 1997). For students who delay enrollment into PSE, social roles may compete with leisure time and offset the effect of PSE on increased substance use.

Also, the timing of transitions may influence a person’s behavior because of age-related social norms across the life course (Elder, 1998). Delayed entrants may not experience increases in substance use after enrollment because the subjective meaning of both substance use and PSE differs for these students compared with those who enroll directly out of high school. For substance use, the average trajectory of alcohol and marijuana use escalates in adolescence, peaks in the early 20s, and then begins to decline (Chassin et al., 2013; Homel et al., 2014; Maggs & Schulenberg, 2005; Thompson et al., 2014). Moreover, social and recreational motives for substance use (i.e., to have a good time with friends) decrease between ages 18 and 29 years (Patrick et al., 2011). For delayed entrants, substance use may decline after enrollment in PSE because their age at the time of enrollment coincides with the normative age-graded declines in substance use.

The subjective meaning of enrollment in PSE may also differ between delayed entrants and direct entrants because of pre-existing inequities in access to higher education. Research consistently shows that delayed entrants have earlier difficulties that may hinder a direct pursuit of education following high school, such as low socioeconomic status (Hango, 2011; Goldrick-Rab & Han, 2011; National Centre for Educational Statistics, 2005), poor academic achievement (Birch & Miller, 2007; Curtis et al., 2012; Hango, 2011), and earlier substance use (Crawford & Cribb, 2012). Delayed entrants may represent a resilient group of young people who pursue PSE despite earlier disadvantage, for whom the transition into PSE may reflect a turning point, fostering declines in substance use (Roche et al., 2006; Teruya & Hser, 2010).

Current study

The literature suggests that the level of substance use among young people differs depending on variation in the nature and timing of developmental transitions (Maggs et al., 2012; Oesterle et al., 2010; White et al., 2006). The purpose of the current study was to consider whether increases in heavy drinking and marijuana use after enrollment in PSE are conditional on the developmental timing of enrollment. Does the trajectory of substance use differ between students who enroll directly into PSE and students who delay enrollment? We hypothesized that students who enroll directly out of high school would experience greater increases in substance use after PSE enrollment compared with students who delay enrollment and that increases in substance use after PSE enrollment would be attenuated for youth who take more time off between high school graduation and PSE enrollment.

The current study used a piecewise latent growth modeling framework to assess the direction and magnitude of change in substance use before and after PSE enrollment. We compared trajectories of alcohol and marijuana use across three groups of young adults who enrolled in PSE at different times: (a) direct entrants (enrolled directly out of high school), (b) gap entrants (took a year off), and (c) delayed entrants (took longer than a year off). We also investigated group differences in sex, socioeconomic status, high school grades, type of PSE, and cohabitation or marriage at the time of enrollment.

Method

Participants

The Victoria Healthy Youth Survey is a five-wave, multi-cohort study of young people between ages 12 and 27 years (see Leadbeater et al., 2012, for details). Participants were recruited from a medium-sized Canadian city using random-digit dialing of 9,500 private telephone listings. Of the 1,036 eligible households with an adolescent age 12–18 years, 662 adolescents (Mage = 15.52, SD = 1.93; 51% female) agreed to participate (64%) (Wave 1). Participants were assessed biennially between 2003 and 2011. Response rates were 87% (n = 578) at Wave 2, 81% (n = 539) at Wave 3, 70% (n = 459) at Wave 4, and 70% (n = 464) at Wave 5. Seventy percent of respondents participated in all five waves. The sample was 85% White, 4% Asian, 4% mixed/biracial, 3% Aboriginal, and 4% other (e.g., Black, Hispanic, or other). Forty-three percent of fathers and 49% of mothers completed college or university training. The living situation, parental education, and ethnicity reported by participating youth were almost identical to that of the population from which the sample was drawn (Albrecht et al., 2007). Our sample was restricted to 391 participants who had enrolled in PSE by the final wave of assessment.

Informed consent was obtained from parents or guardians if youth were under 18 and from the youth at each wave. Two-part questionnaires were administered by trained interviewers in participants’ homes or an alternate location that provided privacy, and respondents received a $35 honorarium per interview. In Part 1, the interviewer recorded responses; Part 2 responses were recorded by the youth to enhance confidentiality for potentially sensitive issues such as use of illegal substances and sexual behavior.

Measures

Substance use

Alcohol was measured using participants’ reported frequency of HED at each wave (Thompson et al., 2014): “How often in the past 12 months have you had five or more drinks on one occasion?” Marijuana use was measured using participants’ reported frequency of use: “How often have you used marijuana in the past 12 months?” For both measures, response options were 0 (never), 1 (a few times per year), 2 (a few times per month), 3 (once a week), and 4 (more than once a week).

Time to enrollment

Time to PSE enrollment was calculated using participants’ reported date of leaving high school and reported date of enrollment in a postsecondary institution. PSE was defined as enrollment in either a 2-year college or a 4-year university. Participants were then categorized into three enrollment groups based on their time to enrollment. The “direct” group (n = 218) included students who enrolled in PSE within 6 months of leaving high school. The average number of months between high school graduation and enrollment for the direct entrants was 2.03 months (SD = 0.41), and their average age at time of enrollment was 17.63 years (SD = 0.54). The “gap” group (n = 104) included students who enrolled in PSE between 6 months and 18 months after leaving high school. On average, the gap group enrolled after 12.48 months (SD = 3.80) and were 18.6 years old (SD = 0.57) at the time of enrollment. The “delay” group (n = 69) included students who enrolled in PSE at least 19 months after leaving high school. On average, the delay group enrolled after 39.99 months (SD = 16.00) and were 20.83 years old (SD = 1.41) at the time of enrollment.

Covariates

Socioeconomic status was assessed as maternal education measured on a 5-point scale that ranged from 1 (less than a high school diploma) to 5 (finished college/ university). To assess high school grades, participants were asked, “In general, what are your grades right now?” with 5-point response options that ranged from 1 (mostly F’s) to 5 (mostly A’s). The type of PSE was coded as 1 (2-year college program) or 0 (4-year university). To assess cohabitation and marriage, participants were asked to retrospectively provide data about who they were living with at each age. The living situation data were matched to the participant’s reported age at the time of enrollment, and responses were categorized as 0 (no cohabitation) or 1 (cohabitation) at the time of enrollment. Marriage and cohabitation were combined because of the age of the respondents.

Plan of analyses

The five waves of data were restructured to center the time metric on the event of enrolling in PSE, allowing us to pinpoint the specific effect of postsecondary enrollment on increases or decreases in the trajectory of heavy drinking and marijuana use for each enrollment group. Piecewise latent growth curve models, each with a single intercept and two growth factors, were used to model the trajectory of alcohol and marijuana use over time (Li et al., 2001). Preliminary analyses examined unconditional piecewise growth models to determine the nature of the developmental trajectories of each substance. Slope 1 described linear changes in substance use pre-enrollment, and Slope 2 described linear and quadratic changes in substance use post-enrollment. The intercept (set to zero) was centered at year of enrollment. Heavy drinking and marijuana use were treated as ordered categorical outcomes with thresholds constrained to equality across all time points (Feldman et al., 2009; Mehta et al., 2004). The mean of the intercept at time of enrollment was set to zero as a reference for estimation of the means of the other growth parameters.

Parameters were estimated using a full-information maximum likelihood estimator with robust standard errors, MLR in MPlus 7.1 (Muthén & Muthén, 1998–2012). This approach uses data from all available time points for a given case under the assumption that data are missing at random, which allows for missingness to be related to variables included in the analyses (Little & Rubin, 2002). MLR also offers protection against inflated alpha values as a result of incomplete data, nonnormality, and small sample sizes (Rhemtulla et al., 2012; Savalei, 2010). With categorical outcomes, chi-square and related fit statistics (e.g., root mean square error of approximation) are not available, but nested models can be compared using −2 times the log likelihood difference, which is distributed as chi-square. To test our research question, each of the growth parameters was regressed on dummy-coded variables for each enrollment group. Direct entrants were the reference group, but analyses were repeated using the gap group as the reference to test contrasts between all pairs of groups.

In preliminary models, sex was included to adjust for the association between enrollment group and substance use trajectories. However, sex was related only to levels of use, not rates of change, and therefore was not included as a predictor of rates of change in the final models. Sex × Timing Group interaction effects (sex × each timing dummy) on the growth parameters were also tested but were not significant and did not improve model fit (HED: Δχ2 = 5.76, Δdf = 8, p = .67; marijuana: Δχ2 = 8.42, Δdf = 8, p = .39). Type of PSE was also included as a predictor of levels of use and rates of change in use post-entry.

Results

Characteristics of the enrollment groups

Table 1 shows the characteristics for each group of post-secondary students. The direct enrollment group had the highest grades during high school, followed by the gap-year group and the delayed enrollment group. The direct enrollment group also had significantly higher socioeconomic status (measured by maternal education) than the delayed enrollment group and was more likely to enroll in a university program than a 2-year college program compared with both the gap and delay groups. Last, members of the delayed enrollment group were most likely to be cohabiting with a partner at the time of enrollment, followed by the gap-year group and the direct enrollment group.

Table 1.

Characteristics of each enrollment group

Variable Full sample (N = 391) % Direct enrollment (n = 218) % Gap year (n = 104) % Delayed enrollment (n = 69) % χ2 (2)
Female 54.0 55.5a 57.7a 43.5a 3.84
Cohabiting at time of enrollment 9.7 3.2a 12.5b 26.1c 32.51***
Enrollment in a 4-year university 68.5 83.5a 55.8b 40.6c 55.47***
M (SD) M (SD) M (SD) M (SD) F(2, 388)
Maternal education 4.13 (1.25) 4.26 (1.17)a 4.04 (1.27)a,b 3.86 (1.42)b 3.11*
High school grades 4.30 (0.70) 4.54 (0.59)a 4.20 (0.70)b 3.68 (0.63)c 49.94***
Age at time of enrollment 17.63 (0.54) 18.6 (0.57) 20.83 (1.41)
Open in a new tab

Note: Estimates that do not share a subscript are significantly different at p < .05.

*

p < .05;

***

p < .001.

Changes in heavy drinking over time

For the total sample, on average, heavy drinking increased the fastest before enrollment in PSE (Slope 1 = 12.27, p < .001), continued to increase after enrollment (Slope 2 = 6.35, p < .001), and then began to decline approximately 3 years after enrollment (quadratic = −10.08, p < .001) (Table 2, Model 1). There was also significant between-individual variation in the growth factors.

Table 2.

Enrollment group differences in heavy drinking over time

Variable Model 1 Est. (SE) Model 2 Est. (SE)
Linear slope pre-entry (Slope 1)
 Mean 12.27 (1.18)*** 16.39 (1.89)***
  Direct 0a
  Gap −2.99a(2.10)
  Delay −11.57b(2.37)***
Level at entry (intercept)
 Mean 0 0
  Direct 0a
  Gap 1.71b(0.47)***
  Delay 1.78b(0.52)***
 Sex (1 = female) −1.79 (0.32)***
 Postsecondary type (0 = university) 0.41 (0.38)
Linear slope post-entry (Slope 2)
 Mean 6.35 (1.18)*** 10.70 (1.59)***
  Direct 0a
  Gap −5.90a,b(2.10)
  Delay −11.57b(2.37)***
 Postsecondary type (0 = university) −3.99 (2.57)
Quadratic change post-entry
 Mean −10.08 (2.09)*** −15.49 (2.72)***
  Direct 0a
  Gap 7.25b(4.28)†
  Delay 10.62b(6.16)†
 Postsecondary type (0 = university) 6.03 (4.92)
Variance
 Slope 1 86.54 (26.27)** 58.32 (1.42)**
 Intercept 9.54 (1.54)*** 8.39 (1.42)***
 Slope 2 11.53 (4.98)* 10.01 (4.62)*
Covariances
 Slope 1 with
  Intercept 1.71 (4.25) 5.53 (3.67)
  Slope 2 20.75 (8.82)* 8.27 (8.76)
 Intercept with
  Slope 2 −4.42 (2.15)* −3.74 (1.94)*
Model fit
 df 13 25
 AIC 3,959.94 3,834.52
 BIC 4,011.53 3,933.74
Open in a new tab

Notes: Parameter estimates are from models using the direct enrollment group as the reference group. Models were re-estimated using the gap enrollment group as the reference group to test pairwise differences in the effects of enrollment group on the growth factors. For a given growth factor, pairs of enrollment groups that do not share a subscript are significantly different from one another (p < .05). Est. = estimate; AIC = Akaike Information Criterion; BIC = Bayesian Information Criterion.

*

p < .05;

**

p < .01;

***

p < .001;

†

p < .10.

Do patterns of heavy drinking differ among enrollment groups?

As shown in Table 2, there were significant differences in the trajectories of heavy drinking over time between enrollment groups (Model 2). Before enrollment, the delayed entrants had significantly higher levels of heavy drinking and increased more slowly in heavy drinking than direct entrants and gap-year takers. At the time of the enrollment, both the gap and delay groups had significantly higher levels of heavy drinking compared with direct entrants, possibly reflecting higher adolescent use. After enrollment, direct entrants and gap entrants increased similarly in their frequency of HED. Direct entrants experienced a significantly greater increase in heavy drinking compared with the delayed entrants and had the fastest declines in heavy drinking approximately 4 years after enrollment. For the delayed entrants, levels of heavy drinking began to decline after enrollment. Females had lower levels of heavy drinking at the time of enrollment. Type of postsecondary institution enrolled in did not predict levels or rates of change in heavy drinking over time. These enrollment group differences in heavy drinking are plotted in Figure 1a. Figure 1b plots heavy drinking by age and shows age-related changes in heavy drinking by enrollment group.

Figure 1.

Figure 1

Open in a new tab

Estimated growth trajectories of heavy drinking for each enrollment group by (a) time and (b) age. Note: Dotted lines indicate average age at enrollment for each group.

Changes in marijuana use over time

On average, marijuana use increased fastest before enrollment (Slope 1 = 6.74, p < .001), was stable after enrollment (Slope 2 = 1.91, p = .08), and began to decline approximately 3 years after enrollment (quadratic = −5.75, p = .01) (Table 2, Model 1). There was also significant between-individual variation in the growth factors.

Do patterns of marijuana use differ among enrollment groups?

As shown in Table 3, there were significant differences in the trajectories of marijuana use over time between enrollment groups (Model 2). Before enrollment, the delayed entrants had higher levels of marijuana use and increased more slowly in use than direct entrants and gap-year takers. At the time of enrollment, both the gap and delay groups had significantly higher levels of marijuana use compared with direct entrants. After enrollment, the direct entrants had a significantly greater increase in marijuana use compared with the gap group. The gap and delay group did not differ significantly in levels and rates of change in marijuana use after enrollment. Their levels of marijuana use remained stable immediately following enrollment. All groups declined at a similar rate approximately 4 years after enrollment. Females had lower levels of marijuana use at the time of enrollment. Students who enrolled in 2-year colleges had higher levels of marijuana use at the time of enrollment, and faster declines in marijuana use after enrollment, than did students who enrolled in a university. Enrollment group differences in marijuana use are plotted in Figure 2a. Figure 2b plots marijuana use by age and shows age-related changes in marijuana use by enrollment group.

Table 3.

Enrollment group differences in marijuana use over time

Variable Model 1 Est. (SE) Model 2 Est. (SE)
Linear slope pre-entry (Slope 1)
 Mean 6.74 (1.10)*** 10.01 (1.66)***
  Direct 0a
  Gap −2.52a(1.96)
  Delay −7.53b(2.07)***
Level at entry (intercept)
 Mean 0 0
  Direct 0a
  Gap 2.04b(0.53)***
  Delay 1.91b(0.65)**
 Sex (1 = female) −0.73 (0.36)*
 Postsecondary type (0 = university) 1.01 (0.43)*
Linear slope post-entry (Slope 2)
 Mean 1.91 (1.33) 5.23 (1.87)**
  Direct 0a
  Gap −6.08b(1.96)*
  Delay −0.81a,b(3.89)
 Postsecondary type (0 = university) −5.50 (2.77)*
Quadratic change post-entry
 Mean −5.75 (2.44)* −10.74 (3.31)***
  Direct 0a
  Gap 7.86a(5.11)
  Delay −8.99a,b(7.88)
 Postsecondary type (0 = university) 14.31 (4.96)**
Variance
 Slope 1 60.38 (17.34)*** 47.89 (15.46)**
 Intercept 14.12 (2.66)*** 12.20 (2.33)***
 Slope 2 30.57 (9.34)** 25.53 (9.05)**
Covariances
 Slope 1 with
  Intercept 11.32 (5.41)* 13.34 (5.14)**
  Slope 2 16.32 (10.94) 17.44 (11.75)
 Intercept with
  Slope 2 5.55 (3.91) −3.52 (3.47)
Model fit
 df 13 25
 AIC 3,625.72 3,543.64
 BIC 3,677.31 3,642.53
Open in a new tab

Notes: Parameter estimates are from models using the direct enrollment group as the reference group. Models were re-estimated using the gap enrollment group as the reference group to test pairwise differences in the effects of enrollment group on the growth factors. For a given growth factor, pairs of enrollment groups that do not share a subscript are significantly different from one another (p < .05). Est. = estimate; AIC = Akaike Information Criterion; BIC = Baysian Information Criterion.

*

p < .05;

**

p < .01;

***

p < .001.

Figure 2.

Figure 2

Open in a new tab

Estimated growth trajectories of marijuana use for each enrollment group by (a) time and (b) age. Note: Dotted lines indicate average age at enrollment for each of the groups.

Discussion

A large body of research describes the impact of college attendance on substance use during young adulthood. The current study contributes to this research by showing that the effect of college enrollment on increases in substance use differs between students who enroll directly out of high school and students who delay enrollment. Consistent with our hypothesis, students who enrolled in PSE directly out of high school experienced the greatest increases in alcohol and marijuana use post-enrollment. However, students who delayed more than a year had much higher levels of both heavy drinking and marijuana use before enrolling in PSE and also maintained higher levels of marijuana use than the direct entrants over time. Students taking a gap year were similar to direct entrants in alcohol use but similar to delayed entrants in marijuana use.

Which young adults delay enrollment in postsecondary education?

A growing number of postsecondary students take time off between high school graduation and PSE. In the current study, 44% of students delayed their enrollment for 1 year or more. Consistent with past research, the decision to take time off between high school and PSE is embedded in a web of factors that create barriers for the pursuit of PSE (Bozick & DeLuca, 2005; Hango, 2011; Roksa & Velez, 2012). Our findings showed that students who delayed enrollment for even 1 year differed from students who enrolled directly on a number of individual characteristics. Compared with direct entrants, students in the gap and delay groups came from lower socioeconomic backgrounds and had poorer high school grades. They were more likely to enroll in 2-year college programs than in 4-year university programs and were also more likely to combine school attendance with other social roles, such as cohabitation.

Building on past findings, we also found that the enrollment groups differed in earlier levels of substance use. Delayed entrants had higher levels of heavy drinking and marijuana use than direct entrants and gap-year takers before enrollment. The combination of lower socioeconomic status, poorer high school grades, and higher levels of substance use may reflect a lack of motivation, expectation, or opportunity for pursuing PSE directly after high school. However, this group of delayers also demonstrates resiliency, pursuing PSE despite this earlier disadvantage. Their attendance at 2-year colleges rather than at 4-year universities likely reflects the lower cost, lower entrance requirements, and more applied programs characteristic of colleges. More research is needed to explicate the combination of variables and the causal relations among them that contribute to delayed entry. Research is also needed to show how these youth overcome disadvantages that could have limited their educational attainment to high school only.

Do patterns of substance use differ across enrollment groups?

The findings showed that the association between college attendance and increases in heavy drinking and marijuana use was conditional on the timing of enrollment. The stronger effect of college enrollment on heavy drinking for students who enroll directly out of high school may possibly reflect age norms rather than a causal effect of postsecondary enrollment. When plotted by age, the trajectories of heavy drinking corresponded closely to normative age-graded population trends in use, peaking at age 21 and then declining (Maggs & Schulenberg, 2005; Thompson et al., 2014). Therefore, increases in heavy drinking observed among direct entrants may simply reflect the fact that they have further to go to reach their developmental peak when they enroll in college than students who delay enrollment.

However, group differences in marijuana use did not correspond to expected age trends like that of heavy drinking. Rather, students who delayed enrollment consistently had much higher levels of marijuana use over time compared with direct entrants. Their early and persistent use may have precipitated their delayed enrollment in the first place. Therefore, differences in marijuana use across enrollment groups possibly reflect individual characteristics of the delay entrants that predict both marijuana use and timing of enrollment in PSE (i.e., high school grades) (Hango, 2011; Homel et al., 2014). Alternatively, differences in marijuana use among these groups could also reflect differences in contextual factors. For example, the effects of postsecondary attendance on marijuana use in the delay group may be buffered by contextual factors such as cohabitation (Chassin et al., 2013) and attendance at a 2-year rather than a 4-year college (Homel et al., 2014). Although there remains heterogeneity within the enrollment groups, the current findings suggest that colleges should be aware that students who enroll at an older age might enter college with academic vulnerabilities and higher levels of marijuana use that could potentially increase their likelihood of long-term problems with substance use, as well as academic problems. However, given the limited research on trajectories of marijuana use following enrollment in PSE and the relatively small sample size of the delay group in this study, more research is needed to better understand differences in marijuana use between enrollment groups.

Last, it is important to note that the steepest increases in substance use occurred before enrollment in college. Therefore, young people enter PSE with pre-existing substance use trajectories that could possibly contribute to differences in the effect of college enrollment on substance use trajectories across enrollment groups and should be more closely examined in future research.

Limitations

Several important limitations of this study should be noted. First, limiting the generalizability of findings, our participants included Canadian, primarily White youth. Second, we were unable to further subdivide participants based on other dimensions that may also be important (i.e., part-time or full-time status or dropouts vs. completers) because of the small percentage of part-time students and dropouts in each of the enrollment groups. A larger data set would allow for examining how substance use is additionally influenced by these dimensions. Third, many of the students in our sample were attending the same local community college and university. Although both the college and university are typical examples of these types of institutions within Canada, observed differences in outcomes may be specific to these institutions. Use of nationally representative data on substance use and PSE would enhance the generalizability of our findings. Last, sex-specific thresholds of heavy drinking were unavailable; thus, rates of heavy drinking among females may be underestimated.

Conclusions

Our results provide evidence that students enrolling in PSE directly from high school are at greatest risk for increased substance use, but also that students who enroll at an older age may have pre-existing issues with substances that may also require institutional support. Although enrollment in PSE per se may not be causally related to increased substance use, postsecondary institutions provide an important context for substance use prevention and intervention initiatives. Screening and attention to substance use problems in postsecondary health services may be helpful to reduce the negative consequences of prior experiences with these substances.

Acknowledgments

This research was supported by Canadian Institutes for Health Research Grants 838-20000-075 and 93533 (to Bonnie Leadbeater) and 79917 (to Tim Stockwell).

References

  1. Albrecht AK, Galambos NL, Jansson SM. Adolescents’ internalizing and aggressive behaviors and perceptions of parents’ psychological control: A panel study examining direction of effects. Journal of Youth and Adolescence. 2007;36:673–684. [Google Scholar]
  2. American College Health Association. American College Health Association-National College Health Assessment II: Canadian Reference Group Data Report Spring 2013. 2013a Retrieved from http://www.achancha.org/reports_ACHA-NCHAII.html.
  3. American College Health Association. American College Health Association-National College Health Assessment II: Reference Group Data Report Fall 2012. 2013b Retrieved from http://www.achancha.org/reports_ACHA-NCHAII.html.
  4. Bachman JG, O’Malley PM, Schulenberg JE, Johnston LD, Bryant AL, Merline AC. The decline of substance use in young adulthood: Changes in social activities, roles, and beliefs. Mahwah, NJ: Lawrence Erlbaum; 2002. [Google Scholar]
  5. Bachman JG, Wadsworth KN, O’Malley PM, Johnston LD, Schulenberg JE. Smoking, drinking, and drug use in young adulthood: The impacts of new freedoms and new responsibilities. Mahwah, NJ: Lawrence Erlbaum; 1997. [Google Scholar]
  6. Birch ER, Miller PW. The characteristics of ‘Gap-Year’ students and their tertiary academic outcomes. Economic Record. 2007;83:329–344. [Google Scholar]
  7. Bogart LM, Collins RL, Ellickson PL, Martino SC, Klein DJ. Effects of early and later marriage on women’s alcohol use in young adulthood: A prospective analysis. Journal of Studies on Alcohol. 2005;66:729–737. doi: 10.15288/jsa.2005.66.729. [DOI] [PubMed] [Google Scholar]
  8. Bozick R, DeLuca S. Better late than never? Delayed enrollment in the high school to college transition. Social Forces. 2005;84:531–554. [Google Scholar]
  9. Cammelli A, Antonelli G, Di Francia A, Gasperoni G, Sgarzi M. Mixed outcomes of the Bologna Process in Italy. In: Schomburg H, Teichlerchap U, editors. Employment and mobility of bachelor graduates in Europe. Key results of the Bologna Process. Rotterdam, The Netherlands: Sense Publishers; 2011. pp. 143–170. [Google Scholar]
  10. Carter AC, Brandon KO, Goldman MS. The college and noncollege experience: A review of the factors that influence drinking behavior in young adulthood. Journal of Studies on Alcohol and Drugs. 2010;71:742–750. doi: 10.15288/jsad.2010.71.742. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chassin L, Sher KJ, Hussong A, Curran P. The developmental psychopathology of alcohol use and alcohol disorders: Research achievements and future directions. Development and Psychopathology. 2013;25:1567–1584. doi: 10.1017/S0954579413000771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Crawford C, Cribb J. Institute for Fiscal Studies through the Centre for Analysis of Youth Transitions (CAYT) London, UK: Department for Education; 2012. Gap year takers: Uptake, trends and long term outcomes. [Google Scholar]
  13. Crockett LJ, Beal SJ. The life course in the making: Gender and the development of adolescents’ expected timing of adult role transitions. Developmental Psychology. 2012;48:1727–1738. doi: 10.1037/a0027538. [DOI] [PubMed] [Google Scholar]
  14. Curtis D, Mlotkowski P, Lumdsen M. Bridging the gap: Who takes a gap year and why? Adelaide, Australia: National Centre for Vocational Educational Research; 2012. [Google Scholar]
  15. Elder GH., Jr The life course as developmental theory. Child Development. 1998;69:1–12. [PubMed] [Google Scholar]
  16. Feldman BJ, Masyn KE, Conger RD. New approaches to studying problem behaviors: A comparison of methods for modeling longitudinal, categorical adolescent drinking data. Developmental Psychology. 2009;45:652–676. doi: 10.1037/a0014851. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fleming CB, White HR, Haggerty KP, Abbott RD, Catalano RF. Educational paths and substance use from adolescence into early adulthood. Journal of Drug Issues. 2012;42:104–126. doi: 10.1177/0022042612446590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Goldrick-Rab S, Han SW. Accounting for socioeconomic differences in delaying the transition to college. The Review of Higher Education. 2011;34:423–445. [Google Scholar]
  19. Hango D. Delaying post-secondary education: Who delays and for how long? Ottawa, Ontario, Canada: Statistics Canada, Culture, Tourism and the Centre for Education Statistics; 2011. [Google Scholar]
  20. Holmlund B, Liu Q, Skans ON. Mind the gap? Estimating the effects of postponing higher education. Oxford Economic Papers. 2008;60:683–710. [Google Scholar]
  21. Homel J, Thompson K, Leadbeater B. Trajectories of marijuana use in youth ages 15–25: Implications for postsecondary education experiences. Journal of Studies on Alcohol and Drugs. 2014;75:674–683. doi: 10.15288/jsad.2014.75.674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Leadbeater B, Thompson K, Gruppuso V. Co-occurring trajectories of symptoms of anxiety, depression, and oppositional defiance from adolescence to young adulthood. Journal of Clinical Child and Adolescent Psychology. 2012;41:719–730. doi: 10.1080/15374416.2012.694608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Li F, Duncan TE, Duncan SC, Hops H. Piecewise growth mixture modeling of adolescent alcohol use data. Structural Equation Modeling: A Multidisciplinary Journal. 2001;8:175–204. [Google Scholar]
  24. Little M, Handley E, Leuthe E, Chassin L. The impact of parenthood on alcohol consumption trajectories: Variations as a function of timing of parenthood, familial alcoholism, and gender. Development and Psychopathology. 2009;21:661–682. doi: 10.1017/S0954579409000352. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Little RJ, Rubin DB. Statistical analysis with missing data. Hoboken, NJ: Wiley and Sons; 2002. [Google Scholar]
  26. Maggs JL, Jager J, Patrick ME, Schulenberg J. Social role patterning in early adulthood in the USA: Adolescent predictors and concurrent wellbeing across four distinct configurations. Longitudinal and Life Course Studies. 2012;3:190–210. doi: 10.14301/llcs.v3i2.183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Maggs J, Schulenberg J. Initiation and course of alcohol consumption among adolescents and young adults. In: Galanter M, editor. Recent developments in alcoholism, Vol. 17: Alcohol problems in adolescents and young adults. New York, NY: Kluwer Academic/ Plenum Publishers; 2005. pp. 29–47. [DOI] [PubMed] [Google Scholar]
  28. Martin AJ. Should students have a gap year? Motivation and performance factors relevant to time out after completing school. Journal of Educational Psychology. 2010;102:561–576. [Google Scholar]
  29. Mehta PD, Neale MC, Flay BR. Squeezing interval change from ordinal panel data: Latent growth curves with ordinal outcomes. Psychological Methods. 2004;9:301–333. doi: 10.1037/1082-989X.9.3.301. [DOI] [PubMed] [Google Scholar]
  30. Muthén LK, Muthén BO. Mplus user’s guide. 7. Los Angeles, CA: Authors; 1998–2012. [Google Scholar]
  31. National Centre for Education Statistics. Waiting to attend college: Undergraduates who delay their postsecondary enrollment. 2005 (NCES publication No. 2005-152). Retrieved from http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2005152.
  32. Oesterle S, Hawkins JD, Hill KG, Bailey JA. Men’s and women’s pathways to adulthood and their adolescent precursors. Journal of Marriage and Family. 2010;72:1436–1453. doi: 10.1111/j.1741-3737.2010.00775.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. O’Shea J. Gap year: How delaying college changes people in ways the world needs. Baltimore, MD: Johns Hopkins University Press; 2013. [Google Scholar]
  34. Patrick ME, Schulenberg JE, O’Malley PM, Maggs JL, Kloska DD, Johnston LD, Bachman JG. Age-related changes in reasons for using alcohol and marijuana from ages 18 to 30 in a national sample. Psychology of Addictive Behaviors. 2011;25:330–339. doi: 10.1037/a0022445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Rhemtulla M, Brosseau-Liard PÉ, Savalei V. When can categorical variables be treated as continuous? A comparison of robust continuous and categorical SEM estimation methods under suboptimal conditions. Psychological Methods. 2012;17:354–373. doi: 10.1037/a0029315. [DOI] [PubMed] [Google Scholar]
  36. Roche KM, Ensminger ME, Ialongo N, Poduska JM, Kellam SG. Early entries into adult roles: Associations with aggressive behavior from early adolescence into young adulthood. Youth & Society. 2006;38:236–261. [Google Scholar]
  37. Roksa J, Velez M. A late start: Delayed entry, life course transitions and bachelor’s degree completion. Social Forces. 2012;90:769–794. [Google Scholar]
  38. Savalei V. Small sample statistics for incomplete nonnormal data: Extensions of complete data formulae and a Monte Carlo comparison. Structural Equation Modeling: A Multidisciplinary Journal. 2010;17:241–264. [Google Scholar]
  39. Schulenberg JE, Patrick ME. Historical and developmental patterns of alcohol and drug use among college students: Framing the problem. In: White HR, Rabiner D, editors. College drinking and drug use. New York, NY: Guilford; 2011. pp. 13–35. [Google Scholar]
  40. Statistics Canada. Trends in the age composition of college and university students and graduates. 2010 Retrieved from http://www.statcan.gc.ca/pub/81-004-x/2010005/article/11386-eng.htm.
  41. Teruya C, Hser YI. Turning points in the life course: Current findings and future directions in drug use research. Current Drug Abuse Reviews. 2010;3:189–195. doi: 10.2174/1874473711003030189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Thompson K, Stockwell T, Leadbeater B, Homel J. Association among different measures of alcohol use across adolescence and emerging adulthood. Addiction. 2014;109:894–903. doi: 10.1111/add.12499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Thompson K, Stockwell T, Leadbeater B, Homel J. The effect of post-secondary education choices on heavy drinking: Does age of enrollment matter? Emerging Adulthood. doi: 10.1177/2167696814561999. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. White HR, McMorris BJ, Catalano RF, Fleming CB, Haggerty KP, Abbott RD. Increases in alcohol and marijuana use during the transition out of high school into emerging adulthood: The effects of leaving home, going to college, and high school protective factors. Journal of Studies on Alcohol. 2006;67:810–822. doi: 10.15288/jsa.2006.67.810. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES