Adolescents’ Future Orientation and Nonmedical Use of Prescription Drugs

Abstract

Introduction

How adolescents think about their future (i.e., future orientation) impacts their risk-taking behavior. The purpose of the present analysis was to explore whether future orientation (future planning, perceived risk to future goals, and positive future expectations) was associated with nonmedical use of stimulants and analgesics in a sample of high school students.

Methods

Information on future orientation and nonmedical use of prescription drugs (NMUPD) were collected using a paper-and-pencil survey from a sample of 9^th-12^th grade students in a Midwestern school.

Results

Higher perceived risk to future goals and positive future expectations were associated with a lower likelihood of self-reported nonmedical use of stimulants (n = 250; OR = 0.46, 95% CI: 0.26, 0.83; OR = 0.15, 95% CI: 0.05, 0.47, respectively). Only higher perceived risk to future goals was associated with a lower likelihood of self-reported nonmedical use of analgesics (n = 250; OR = 0.40, 95% CI: 0.24, 0.68). In a follow-up analysis limited to students who endorsed alcohol or marijuana use, perceived risk to future goals remained associated with a lower likelihood of nonmedical use of stimulants and analgesics.

Conclusions

Results suggest that risk perception might be a salient protective factor against both nonmedical use of stimulants and analgesics. Overall, the differential impact of conceptualizations of future orientation might depend on the class of prescription drug used, demonstrating a need to consider prescription drugs individually in the development of future studies and interventions.

1. Introduction

Nonmedical use of prescription drugs (NMUPD) is a serious problem facing American adolescents. In 2015, approximately one out of twenty high school seniors reported nonmedical use of stimulants or analgesics (National Institute on Drug Abuse [NIDA], 2016). Research has shown that adolescents who delay the onset of NMUPD are less likely to be diagnosed with prescription drug abuse or dependence in their lifetime (McCabe, West, Morales, Cranford, & Boyd, 2007). These findings suggest that adolescence is a crucial time to avert NMUPD; thus, identification of factors that protect against such behaviors is a priority (Egan, Van Horn, Monahan, Arthur, & Hawkins, 2011; Piko & Kovács, 2010). One potential protective factor is future orientation, generally described as “individuals’ tendency to engage in future thinking” (Seginer, 2009, p. 3). Previous research suggests that adolescents’ planning for the future and sense of internal control over the future increases from early to middle adolescence, supporting the notion that future orientation is a growing concern during this developmental period (Nurmi, 1989).

Future orientation, however, is a broad term associated with many conceptualizations (Johnson, Blum, & Cheng, 2014). The present study focuses on three different conceptualizations: future planning, perceived risk to future goals, and positive future expectations. Future planning is the extent to which an individual plans for their future and strives toward future goals (Zimbardo & Boyd, 1999). This might focus individuals’ resources on goal-oriented positive activities and, in turn, might lessen the appeal of engaging in risk behaviors. Alternatively, considering future consequences of behavior helps individuals link the present with the future (Strathman et al., 1994). Understanding how potentially enjoyable risk behavior in the present can damage future goals might focus attention on consciously avoiding these negative behaviors. This might be particularly critical during the period of adolescence when contextual circumstances (e.g., encouraging peers) often make risk behavior appealing. Finally, positive future expectations are the perceived likelihood of attaining specific objectives in life (e.g., have a happy life; Dubow, Arnett, Smith, & Ippolito, 2001). Positive future expectations provide a thematic vision of what an individual’s future could be and might deter risk behaviors that put this vision in jeopardy.

Indeed, previous research suggests that these conceptualizations of future orientation might be protective against substance use in adolescence. For example, future planning and positive future expectations are associated with lower levels of tobacco, alcohol, and marijuana use during adolescence and young adulthood (Apostolidis, Fieulaine, & Soulé, 2006; Barnett et al., 2013; Henson, Carey, Carey, & Maisto, 2006; Dunn, Kitts, Lewis, Goodrow, & Scherzer, 2011). Similarly, perceived risk to future goals has been linked to a lower likelihood of engaging in dangerous alcohol use in adolescents (McKay, Percy, & Cole, 2013). To our knowledge, however, researchers have not yet explored how NMUPD is related to future orientation and, in particular, its various conceptualizations.

The present study sought to explore whether future planning, perceived risk to future goals, and positive future expectations were differentially associated with the nonmedical use of stimulants and analgesics in a sample of high school students. Although future planning might focus individuals’ activities on behavior that will advance them toward future goals, it does not necessarily preclude risky behavior. Indeed, highly-motivated adolescents engage in NMUPD (Veliz, Boyd, & McCabe, 2013). Alternatively, perceived risk to future goals focuses on risk perception and links current, specific behavior with future consequences. Previous research suggests that risk perception is a powerful key to lower tobacco and alcohol use among adolescents (McKay, Percy, & Cole, 2013; Virgili, Owen, & Severson, 1991). Furthermore, although future planning might focus on positive motivation, perceived risk to future goals highlights negative potential outcomes. For adolescents, recognizing negative effects of certain behaviors might especially motivate avoidance of those actions (Reynolds et al., 2015). Finally, positive future expectations might be only weakly associated with current behavior because they are merely anticipated events rather than outcomes that individuals have to strive toward.

The motivation for using analgesics and stimulants differs, potentially resulting in opposing directions of association. Previous research suggests that one of the primary motivations for the nonmedical use of analgesics is sensation-seeking (McCabe, Boyd, Cranford, & Teter, 2009). Alternatively, stimulants might be used by highly motivated adolescents to excel in various activities, including school (King, Jennings, & Fletcher, 2014; Veliz et al., 2013). Thus, although future orientation might be associated with lower nonmedical use of analgesics, it might instead be associated with higher nonmedical use of stimulants. This possibility was explored in the present study.

Previous studies have shown that adolescents who engage in NMUPD are also significantly more likely to use other substances, most often alcohol and marijuana (McCabe, Boyd, & Teter, 2005; McCabe, West, Schepis, & Teter, 2015). For example, McCabe, Teter and Boyd (2009) found that approximately 86% of middle and high school students who reported nonmedical use of stimulants also reported alcohol use in the past year and 69% reported marijuana use in the past year. On the contrary, only 37% and 14% of adolescents who reported no stimulant use reported alcohol use in the past year and marijuana use in the past year, respectively. We, therefore, explored the relationship between future orientation and NMUPD in the context of polysubstance use. Specifically, we explored whether adolescents reporting lifetime use of alcohol or marijuana in addition to NMUPD report differential levels of future planning, perceived risk to goals, and positive future expectations compared to adolescents who report lifetime use of alcohol or marijuana alone.

2. Methods

2.1 Participants and Procedure

This secondary data analysis is based on data collected in one Midwestern high school in Fall 2014. The parent study was aimed at developing school curricula to improve positive future orientation and decrease substance use. All students were invited to complete a paper-and-pencil survey administered by trained school staff during students’ homeroom class. Electronically mailed letters with information about the study were sent to parents by the PI and School Principal. Parents were instructed to contact the PI or School Principal to exclude their student; student assent was obtained prior to survey administration. The study was reviewed by the University of Michigan Institutional Review Board and received exemption (Exemption number: HUM00090000).

Approximately 86% of 9^th through 12^th grade students participated in the survey (n = 408, M_age = 15.36, SD = 1.21; 50% female; 72% White). Nonparticipation was due to lack of parental consent or student assent, or school absence at the time of survey administration. Students were given instructions on how to develop a personal identification code for the survey to maintain anonymity. The survey was completed by most students in approximately 33 minutes.

For the present analysis, we excluded students who did not respond to all measures of interest. Table 1 shows the sample demographics for both the stimulant and analgesic models (n = 250). Participants that were included and participants that were excluded from analyses did not significantly differ according to gender, grade, ethnicity, or parent education.

Table 1.

Sample characteristics and differences between adolescents who did and did not endorse lifetime nonmedical use of prescription stimulants or analgesics.

	Stimulants			Analgesics
	Full Sample (N = 250) % (n)a	Use (N = 24) % (n)b	χ2	Full Sample (N = 250) % (n)a	Use (N = 26) % (n)b	χ2
Gender			9.4**			6.3*
Male	46.4 (116)	3.5 (4)		46.4 (116)	5.2 (6)
Female	53.6 (134)	14.9 (20)		53.6 (134)	14.9 (20)
Ethnicity			1.8			0.83
White	76.4 (191)	11.0 (21)		76.4 (191)	9.4 (18)
Non-white	23.6 (59)	5.1 (3)		23.6 (59)	13.6 (8)
Grade			8.6*			0.98
9th	25.2 (63)	4.8 (3)		25.2 (63)	7.9 (5)
10th	25.2 (63)	4.8 (3)		25.2 (63)	9.5 (6)
11th	24.4 (61)	18.0 (11)		24.4 (61)	13.1 (8)
12th	25.2 (63)	11.1 (7)		25.2 (63)	11.1 (7)
Parent education			0.04			8.8**
College or less	35.2 (88)	9.1 (8)		35.2 (88)	18.2 (16)
Graduate or professional	64.8 (162)	9.9 (16)		64.8 (162)	6.2 (10)
Lifetime use of alcohol or marijuana	56.6 (141)	16.3 (23)	16.6***	56.6 (141)	17.7 (25)	18.5***

^aPercentages are calculated within column.

^bPercentages are calculated within row.

^cN = 249

^*p < .05.

^**p < .01.

^***p < .001. Note. Significant omnibus χ² tests are boldfaced.

2.2 Measures

Nonmedical use of prescription drugs

Nonmedical use of stimulants and analgesics was assessed using modified items from the Monitoring the Future survey (Bachman, Johnston, & O’Malley, 2014). Lifetime nonmedical use of stimulants was assessed with a single item: “In your lifetime, on how many occasions (if any) have you taken a prescription drug to stay awake/alert without a doctor’s prescription (i.e., a stimulant)?” Lifetime nonmedical use of analgesics was also assessed with a single item: “In your lifetime, on how many occasions (if any) have you taken prescription painkillers without a doctor’s prescription?” Response options ranged from 0 to 40+ occasions. Each item was dichotomized as 0 (Never) or 1 (1 or more occasions).

Future orientation

We assessed three conceptualizations of future orientation. Future planning was assessed with five items adapted from the Zimbardo Time Perspective Inventory (e.g., “I finish work that is due tomorrow before playing today”; Zimbardo & Boyd, 1999; Kruger et al., 2015). Responses ranged from 1 (Agree a lot) to 4 (Disagree a lot). Scores were averaged and reverse scored so that higher scores indicated higher future planning (α = 0.72).

Perceived risk to future goals was assessed using two items (Arthur, Hawkins, Pollard, Catalano, & Baglioni, 2002; Michigan Department of Education [MDE], 2014). Participants were asked how much they think people risk not accomplishing their future goals if they: “Take a prescription drug to stay awake/alert (i.e., a stimulant not prescribed to them)” and “Take pain medication not prescribed to them.” Response options ranged from 1 (No risk) to 4 (Great risk). Higher scores indicated greater perceived risk. Each item was matched to its corresponding outcome variable.

Positive future expectations were assessed with six items (e.g., “I will be able to handle the problems that come up in my life”; Wyman, Cowen, Work, & Kerley, 1993). Responses ranged from 1 (Agree a lot) to 4 (Disagree a lot). Scores were averaged and reversed scored so that higher scores indicated higher positive future expectations (α = 0.82).

Alcohol or marijuana use

Lifetime use of alcohol or marijuana was assessed with three items: “In your lifetime, on how many occasions (if any) have you: 1) had at least one drink of alcohol (more than just a few sips)? 2) had five or more drinks of alcohol in a row, that is, within a couple hours? and 3) used marijuana?” Response options ranged from 0 to 40+ occasions. Each item was dichotomized as 0 (Never) or 1 (1 or more occasions). The three dichotomous items were then combined, such that an affirmative response to any of the three items was coded as the presence of lifetime use of alcohol or marijuana (0 = No lifetime use of alcohol or marijuana; 1 = Lifetime use of alcohol or marijuana) for analysis.

Demographic variables

Participants were asked to indicate their age, gender, grade, and ethnicity. Participants were also asked about each parent: “How far did your [mom or dad] or the person that is like a [mom or dad] to you get in school?” The highest level of education reported across parents was used in analyses. If information was not available on both parents, the available parent’s educational information was used.

2.3 Data Analysis

Data were analyzed using SAS 9.4® (Cary, NC). Bivariate relationships between outcome and demographic variables were assessed with χ² tests of association. Logistic regression was used to assess the relationship between future orientation and lifetime nonmedical use of stimulants, and future orientation and lifetime nonmedical use of analgesics. We first examined unadjusted models for each outcome. Unadjusted models were also used to evaluate multicollinearity of the three independent variables. The variance inflation factor of each independent variable was examined and no evidence of multicollinearity was found. We then examined two adjusted models that included the three future orientation variables and four demographic variables predicting 1) lifetime nonmedical use of stimulants and 2) lifetime nonmedical use of analgesics. Finally, follow-up analyses were performed in which we examined the same models in a sample restricted to students who endorsed lifetime use of alcohol or marijuana.

3. Results

3.1 Descriptive Statistics

Table 1 shows the association between lifetime nonmedical use of stimulants and analgesics and each demographic variable.

3.2 Stimulants

Twenty-four participants (9.6%) reported nonmedical use of stimulants in their lifetime. As seen in Table 2, both higher perceived risk to future goals (OR = 0.46; 95% CI: 0.26, 0.83) and higher positive future expectations (OR = 0.15; 95% CI: 0.05, 0.47) were associated with a lower likelihood of nonmedical use of stimulants.

Table 2.

Logistic regression models predicting lifetime nonmedical use of stimulants and analgesics.

	Stimulants				Analgesics
	Unadjusted Estimatesa		Adjusted Estimatesb		Unadjusted Estimatesa		Adjusted Estimatesb
	B (S.E.)	Odds ratio (95% CI)	B (S.E.)	Odds ratio (95% CI)	B (S.E.)	Odds ratio (95% CI)	B (S.E.)	Odds ratio (95% CI)
Future Planning	0.20 (0.50)	1.22 (0.46, 3.22)	−0.07 (0.56)	0.93 (0.31, 2.78)	0.13 (0.46)	1.14 (0.46, 2.79)	0.01 (0.49)	1.01 (0.38, 2.65)
Perceived Risk to Future Goals	−0.73 (0.27)	0.48 (0.28, 0.82)	−0.78 (0.30)	0.46 (0.26, 0.83)	−0.70 (0.23)	0.50 (0.32, 0.79)	−0.91 (0.27)	0.40 (0.24, 0.68)
Positive Future Expectations	−1.96 (0.52)	0.14 (0.05, 0.39)	−1.91 (0.58)	0.15 (0.05, 0.47)	−0.37 (0.49)	0.69 (0.27, 1.80)	−0.17 (0.52)	0.85 (0.30, 2.37)
Gender	--	--	1.65 (0.63)	5.22 (1.52, 17.92)	--	--	1.41 (0.54)	4.11 (1.42, 11.91)
Ethnicity	--	--	0.25 (0.73)	1.29 (0.31, 5.41)	--	--	0.74 (0.52)	2.09 (0.76, 5.76)
Grade
10th	--	--	−0.01 (0.94)	0.99 (0.16, 6.28)	--	--	0.49 (0.71)	1.64 (0.41, 6.59)
11th	--	--	1.55 (0.76)	4.69 (1.06, 20.78)	--	--	0.55 (0.66)	1.74 (0.47, 6.37)
12th	--	--	1.03 (0.80)	2.81 (0.58, 13.53)	--	--	0.72 (0.68)	2.06 (0.55, 7.77)
Parent education	--	--	0.13 (0.54)	1.14 (0.40, 3.26)	--	--	−1.27 (0.46)	0.28 (0.11, 0.69)

^aN = 250. Results significant at the 0.05 significance level are indicated in bold.

^bN = 250. Adjusted for participants’ gender, ethnicity, grade level, and parents’ highest level of education. Results significant at the 0.05 significance level are indicated in bold. Reference groups are ‘Male’ for gender variable, ‘White’ for the ethnicity variable, ‘9^th’ for the grade variable, and ‘4 year college or less’ for parent education variable.

3.3 Analgesics

Twenty-six participants (10.4%) reported nonmedical use of analgesics in their lifetime. As seen in Table 2, higher perceived risk to future goals (OR = 0.40; 95% CI: 0.24, 0.68) was associated with a lower likelihood of nonmedical use of analgesics.

3.4 Follow-up Analysis

In our sample, 95.8% (n=23) of adolescents who endorsed lifetime nonmedical use of stimulants and 96.2% (n=25) of those who endorsed lifetime nonmedical use of analgesics reported also having used alcohol or marijuana in their lifetime (p<0.0001; Table 1). Additionally, when regression models were limited to students who endorsed alcohol or marijuana use (n = 141), perceived risk to future goals remained associated with a lower likelihood of nonmedical use of stimulants (OR=0.41; 95% CI: 0.21, 0.78; Appendix) and nonmedical use of analgesics (OR=0.37; 95% CI: 0.20, 0.67; Appendix).

4. Discussion

The present study suggests that, in general, higher levels of future orientation were associated with a lower likelihood of NMUPD, yet these relationships depend on the conceptualization of future orientation and the class of prescription drug examined. Specifically, perceived risk to future goals was associated with a lower likelihood of both nonmedical use of stimulants and analgesics. This indicates that risk perception might be particularly important in focusing adolescents’ attention on specific behavior that impedes goal attainment. Past research suggests that recognizing negative consequences might be especially helpful in reducing problem behavior during adolescence (Reynolds et al., 2015). Furthermore, previous research suggests that adolescents can, indeed, recognize behavior as risky (Beyth-Marom, Austin, Fischhoff, Palmgren, & Jacobs-Quadrel, 1993). Thus, educating adolescents about risk and the consequences of their actions might be a fruitful avenue for prevention. Brief motivational interventions have shown promise for reducing alcohol and tobacco use among adolescents (Patnode, O’Connor, Rowland, et al., 2014). These interventions often incorporate the discussion of consequences of substance use (Cunningham et al., 2015). Therefore, applying these interventions to NMUPD prevention may be beneficial.

Interestingly, higher positive future expectations were associated with lower nonmedical use of stimulants, but not nonmedical use of analgesics. While this relationship was unexpected, it highlights the possible difference between stimulant and analgesic users. For individuals high in positive future expectations, who are ultimately optimistic about their future, stimulants might be viewed as unnecessary. Stimulant use, although often motivated by a desire to achieve high academic grades (Garnier-Dykstra et al., 2012), has been shown to be more likely in participants with low self-efficacy about their academic abilities (Looby, Beyer, & Zimmerman, 2015).

Finally, future planning was not associated with the nonmedical use of stimulants or analgesics. Although future planning can help focus attention on positive behavior and motivate future-oriented action, this might correlate poorly with negative behavior. Previous research suggests that adolescents are less likely than adults to simultaneously consider the risks, benefits, and options of a scenario (Halpern-Felsher & Cauffman, 2001). Thus, perceptions of risk and reward might not coalesce into overarching purposeful behavior until later in life. Instead, as indicated above, focusing specifically on perceiving the risks associated with behavior may be particularly important for preventing NMUPD, rather than simply encouraging future-oriented goal-setting broadly.

It is noteworthy that nearly all of the adolescents who endorsed lifetime nonmedical use of stimulants and lifetime nonmedical use of analgesics reported also having used alcohol or marijuana in their lifetime. These results demonstrate the pervasiveness of polysubstance use (McCabe et al., 2005; McCabe et al., 2015) and the difficulty of examining a single form of substance use. Interestingly, results of the follow-up analysis suggest that perceived risk to future goals remained associated with a lower likelihood of nonmedical use of stimulants and nonmedical use of analgesics when the sample was restricted to only individuals reporting lifetime use of alcohol or marijuana. It appears that perceived risk to future goals related to stimulant and analgesic use specifically remains an important deterrent for use. Thus, even when the use of other substances is not perceived as risky or might even be appealing, recognizing why one should not engage in use of a different substance can be important for prevention. Indeed, perceived risk to future goals associated with the nonmedical use of stimulants and analgesics was only moderately correlated with the perceived risk to future goals associated with occasional alcohol and marijuana use (r = .32 to r = .40). Therefore, though perceptions of risk related to different substances are associated, they are not synonymous. This suggests that future intervention work should carefully address specific risks associated with the use of different substances. Future research should continue to explore common factors between various types of substances and consider approaches that might be effective for polysubstance use.

4.1 Limitations

The sample included students from a single high school. Only 5% of the student population was considered economically disadvantaged (MDE, 2015) and most had highly-educated parents. Therefore, the findings are not generalizable to all youth and should be explored in more diverse samples. The cross-sectional design of the study precludes making statements about causality. Additionally, our outcome measures present limitations. First, the scope of our self-report measure does not allow us to differentiate between use and abuse or determine the impact that the substance use or abuse had on an adolescent’s functioning. Although the items assess the use of a prescription drug not prescribed to the adolescent, we are unable to ascertain if it was provided by a trusted individual in the adolescent’s life, such as a parent or friend (Schepis & Krishnan-Sarin, 2009). We are also unable to determine the motive for NMUPD use (McCabe et al., 2009); an adolescent who uses a non-prescribed analgesic provided by a parent for pain might exhibit different characteristics compared to one who purchases it from a friend and uses it for sensation-seeking motives. Future research should explore these factors in relation to future orientation. Additionally, NMUPD and lifetime use of alcohol or marijuana were self-reported, which might be influenced by recall or social desirability bias. Students were also not provided with examples of medications that would fall into the specific NMUPD classes (i.e., brand names of medications), which might have influenced their ability to answer the questions accurately.

4.2 Conclusions

Notwithstanding study limitations, our findings suggest that leveraging adolescents’ future thinking might be important in preventing NMUPD. Educating adolescents about the risks that specific substances pose to their future goals might be especially effective in curtailing NMUPD. Results of the present study suggest that the relationship between future orientation and NMUPD depends on both the conceptualization of future orientation and the class of prescription drug. Additionally, our work suggests a need to consider intervention strategies effective in the context of polysubstance use.

Highlights.

• Future orientation might play a role in adolescents’ NMUPD

• Perceived risk to future goals was associated with a lower likelihood of NMUPD

• Education about the risks of NMUPD shows potential for future intervention work

Appendix

Logistic regression models predicting lifetime nonmedical use of stimulants and analgesics among adolescents who reported lifetime use of alcohol or marijuana.

	Stimulants				Analgesics
	Unadjusted Estimatesa		Adjusted Estimatesb		Unadjusted Estimatesa		Adjusted Estimatesb
	B (S.E.)	Odds ratio (95% CI)	B (S.E.)	Odds ratio (95% CI)	B (S.E.)	Odds ratio (95% CI)	B (S.E.)	Odds ratio (95% CI)
Future Planning	0.12 (0.54)	1.13 (0.39, 3.26)	−0.37 (0.62)	0.69 (0.20, 2.34)	0.30 (0.52)	1.35 (0.49, 3.74)	−0.11 (0.59)	0.90 (0.28, 2.83)
Perceived Risk to Future Goals	−0.80 (0.30)	0.45 (0.25, 0.80)	−0.90 (0.33)	0.41 (0.21, 0.78)	−0.82 (0.27)	0.44 (0.26, 0.74)	−1.01 (0.31)	0.37 (0.20, 0.67)
Positive Future Expectations	−1.41 (0.57)	0.24 (0.08, 0.75)	−1.19 (0.66)	0.31 (0.08, 1.12)	0.07 (0.57)	1.07 (0.35, 3.25)	0.51 (0.65)	1.66 (0.47, 5.94)
Gender	--	--	1.88 (0.71)	6.54 (1.64, 26.07)	--	--	1.72 (0.63)	5.61 (1.63, 19.26)
Ethnicity	--	--	−0.17 (0.74)	0.84 (0.20, 3.62)	--	--	0.30 (0.57)	1.35 (0.44, 4.15)
Grade
10th	--	--	0.12 (1.09)	1.13 (0.13, 9.52)	--	--	−0.16 (0.84)	0.86 (0.17, 4.45)
11th	--	--	1.54 (0.93)	4.66 (0.76, 28.65)	--	--	−0.39 (0.81)	0.68 (0.14, 3.33)
12th	--	--	1.38 (0.98)	3.96 (0.58, 27.03)	--	--	0.12 (0.81)	1.13 (0.23, 5.58)
Parent education	--	--	−0.24 (0.61)	0.78 (0.24, 2.58)	--	--	−1.67 (0.53)	0.19 (0.07, 0.54)

^aN = 141. Results significant at the 0.05 significance level are indicated in bold.

^bN = 141. Adjusted for participants’ gender, ethnicity, grade level, and parents’ highest level of education. Results significant at the 0.05 significance level are indicated in bold. Reference groups are ‘Male’ for gender variable, ‘White’ for the ethnicity variable, ‘9^th’ for the grade variable, and ‘4 year college or less’ for parent education variable.