Journal of American College Health

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/vach20

Perfectionism, substance use, and mental health

in college students: A longitudinal analysis

Sarah K. Nelsen, Alper Kayaalp & Kyle J. Page

To cite this article: Sarah K. Nelsen, Alper Kayaalp & Kyle J. Page (2021): Perfectionism,
substance use, and mental health in college students: A longitudinal analysis, Journal of American
College Health, DOI: 10.1080/07448481.2021.1891076

To link to this article: https://doi.org/10.1080/07448481.2021.1891076

Published online: 24 Mar 2021.

Submit your article to this journal

Article views: 93

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=vach20
JOURNAL OF AMERICAN COLLEGE HEALTH
https://doi.org/10.1080/07448481.2021.1891076

Perfectionism, substance use, and mental health in college students:

A longitudinal analysis
Sarah K. Nelsen, MSa, Alper Kayaalp, PhDb , and Kyle J. Page, PhDc
a
South Dakota State University, Brookings, South Dakota, USA; bIndustrial/Organizational Psychology, Department of Psychology, South
Dakota State University, Brookings, South Dakota, USA; cWorkforce Analytics Consultant, American Family Insurance, Madison,
Wisconsin, USA

ABSTRACT ARTICLE HISTORY

The study examined the relationship between perfectionism (categorized by adaptive, maladaptive, Received 1 December 2019
and non-perfectionist clusters), substance use (i.e., drinking and smoking), and mental health in a Revised 18 December 2020
large sample of college student (N ¼ 841; M ¼ 19.6 years of age) in a time-lagged design. Students Accepted 11 February 2021
were classified using a two-step cluster analysis where subscales are used in hierarchical and k-
KEYWORDS
means cluster analysis. Results revealed that adaptive perfectionists reported better mental health Alcohol use; college
compared to non-perfectionists and maladaptive perfectionists. Additionally, adaptive perfectionists students; mental health;
reported lower levels of alcohol use compared to non-perfectionists. Differences between mental perfectionism; smoking;
health and alcohol use by cluster were found over time, but not for smoking behaviors. Across all substance use
participants, substance use did not mediate the relationship between perfectionism and mental
health. Future research directions as well as practical implications are discussed.

Introduction contributing to student mental health and substance use is

required to identify students at risk. As an individual per-
The mental health of college students is a growing concern
sonality characteristic, perfectionism, defined as “the ten-
in recent years.1–5 A recent report found that 40% of stu-
dency to hold and pursue unrealistically high goals,”13 is
dents suffered from either moderate or serious psychological
one promising construct for investigation in college students
distress in the last 12 months.6 Moreover, the mental well-
who live in an environment that supports and appreciates
being of college students has been aggravated by the
high-achieving tendencies.14–20 A recent meta-analysis that
ongoing COVID-19 pandemic and its social and economic
consequences. A recent study by the Healthy Minds examined college students in the United States, Canada, and
Network7 found that the mental well-being of college stu- the United Kingdom from 1989 to 2016 indicates that per-
dents has deteriorated since the start of the pandemic, sug- fectionism has increased over the past three decades, citing
gesting that the pandemic and its accompanying effects will “more competitive environments, more unrealistic expecta-
continue impacting collegiate mental health.8 tions, and more anxious and controlling parents than gener-
Underage college substance use is also a growing public ations before” as the main reasons.14 Although the construct
health concern as it negatively impacts the intellectual lives has been recognized as a significant predictor of mental
of college students.9,10 According to a recent survey,11 over health,21–25 it has not received much attention in the college
fifty percent of college students have consumed alcohol in student mental health and substance use literature.
the past month, and more than thirty-three percent have
engaged in binge drinking during that same timeframe. Perfectionism and mental health
Similarly, sixty-three percent of college students reported
using tobacco or other nicotine delivery products in the past Perfectionism has received increasing attention from both
3 months.6 Increasing co-occurring of alcohol use and smok- scholars and practitioners during the last two decades.
ing during college increases the risk of lifelong patterns of While early studies defined perfectionism as a unidimen-
use and, ultimately, alcohol and nicotine dependence.12 sional trait with negative consequences,24,26–28 subsequent
Thus, understanding factors that increase the risk of collegi- studies identified its positive effects as well as differentiated
ate mental problems and substance use (i.e., drinking and the construct into two subtypes as adaptive and maladaptive
smoking) is a pressing issue. perfectionism.25,29,30 Further, Parker31 proposed a tripartite
Given the aforementioned concerns on college campuses, model of perfectionism adding non-perfectionists as the
a more thorough understanding of the risk factors in 3rd group.

CONTACT Alper Kayaalp Alper.kayaalp@sdstate.edu Industrial/Organizational Psychology, Department of Psychology, South Dakota State University, 857
11th St., 029B Hansen Hall, Brookings, South Dakota 57007, USA.
ß 2021 Taylor & Francis Group, LLC
2 S. K. NELSEN ET AL.

Empirical studies have tested and supported this tripartite hypothesis: Adaptive perfectionists will have better mental
model in the general population and student samples.31–37 health than maladaptive perfectionists and non-perfectionists
Overall, this stream of research indicates that perfectionistic (Hypothesis 1).
individuals hold very high standards and strive diligently to
meet those standards. However, within the tripartite model,
Perfectionism, mental health, and substance use
maladaptive perfectionism has been associated with “high
personal standards, excessive self-criticism, worries about The research on perfectionism has primarily focused on
perceived mistakes and failures, and dissatisfaction with main effects; as such, there is little understanding of the
one’s performance.”17 In contrast, in adaptive perfectionism, mechanisms through which perfectionism influences import-
one’s failure to meet high personal standards can be ant outcomes,49 such as mental well-being. It is of the
endured “without excessive self-criticism or feelings of dis- utmost importance for colleges to identify the factors that
satisfaction.”17,38–40 In comparison, the third group (non- can help to explain the relationship between perfectionism
perfectionists) has been distinguished from the others by and mental health.17 Given the prevalence of co-occurrence
their low levels of high personal standards.20,37 of alcohol and tobacco use among college students, a con-
This categorization of perfectionism has allowed research- struct of particular interest regarding the relationship
ers and practitioners to investigate perfectionism in relation between perfectionism and mental health is substance use.
to a variety of psychological outcomes. For example, mal- According to the stress-coping model,54 substances such
adaptive perfectionism is positively associated with mental as alcohol, tobacco, and opiates may be used as a coping
health issues in college students17,41–45 while adaptive perfec- mechanism to reduce stress, negative emotions, or to
tionism is negatively associated with mental health enhance positive affect.55 The model indicates that individu-
issues.21,42 Although informative, these prior studies are lim- als will use substances to cope instead of using more adap-
ited in several ways. Most are cross-sectional studies and tive coping strategies due to a lack of effective coping skills
focus on only adaptive and maladaptive perfectionists with- such as the ability to make decisions and problem solve.
out considering non-perfectionists. Furthermore, most of However, using substances to cope results in negative conse-
these studies have been characterized by small sample sizes quences such as increased unresolved stress, which then
with limited generalizability. contributes to decreases in mental well-being.55
In line with previous perfectionism research,46–49 we use Although there has been research suggesting a link
the Conservation of Resources (COR) theory50 as the theor- between perfectionism and substance use,25,56 there remain
etical framework for our study. Briefly, the theory posits inconsistencies in the literature.57–59 Furthermore, while
that individuals will experience psychological strain when coping strategies has gained increasing attention in perfec-
their resources become exhausted, which may affect their tionism research,19,60,61 substance use, specifically in the
general health.51 When applying COR to collegiate samples, form of alcohol use and smoking, has not been examined in
we argue that college students often perceive having insuffi- college students as a potential mechanism between perfec-
cient resources to cope with the challenges of college life tionism and mental health. This relationship may also be
given their intellectual, emotional, and time-intensive explained by the COR theory. Specifically, maladaptive per-
responsibilities.34 Within this framework, we posit that per- fectionists may be more prone to a loss spiral (refers to a
fectionistic students may strive to excel and seek high per- process of the expense of resources that makes them
sonal standards considerably more to maintain resources unavailable to cope with future loss, thus potentially leading
(e.g., accomplishment, achievement, time, energy, self-effi- to further resource loss62) and consequently use maladaptive
cacy, social support) than non-perfectionists.48,51 Further, coping strategies (such as drinking and smoking) to deal
when valuable resources become depleted over time, mal- with their loss. Although avoiding a stressor may provide
adaptive perfectionists may be more heavily impacted by short-term relief, this strategy may feed into a larger loss
this loss and face greater mental health deteriorations in spiral; ultimately increasing the psychological strain an indi-
comparison to non-perfectionist students. Additionally, vidual experiences, the dysfunctional coping strategies they
adaptive perfectionists’ practice of setting high personal use, and reducing their mental well-being. Conversely, gain
goals for themselves, achieving high standards, and main- spiral (“gain begets further gain”62) may occur when adap-
taining resources, may have greater benefits to one’s mental tive perfectionists attain more resources by using adaptive
health52 compared to maladaptive perfectionists, whose coping strategies.63 In an attempt to gain and maintain
health may be jeopardized by feeling displeased with their resources, adaptive perfectionists may experience slow, but
inability to live up to their high standards and their inability salient resource gains that may increase their motivation to
to retain resources.14 Similarly, maladaptive perfectionist continue to use adaptive coping strategies. Research has sup-
students can be expected to be more preoccupied with their ported this as maladaptive perfectionists were found to have
mistakes, thus draining their resources (e.g., time and atten- higher levels of stress and were more likely to use drugs
tion) to successfully accomplish tasks.49 These assertions than adaptive perfectionists.64,65
support COR theory in that accomplishment and achieve- To advance the theoretical understanding of how and
ment are seen as personal resources that aid an individual, why perfectionism influences mental health, we respond to
ultimately reducing stress, and increasing one’s mental well- calls of previous scholar for longitudinal designs in perfec-
being.53 The above arguments give rise to the following tionism research.49,60,61 As such, we hypothesize that;
 JOURNAL OF AMERICAN COLLEGE HEALTH 3

Figure 1. The proposed mediation model.

adaptive perfectionists will have lower levels of substance For Time 1, 69.7% were female and the average age was
use (i.e., smoking and drinking) than maladaptive perfec- 19.14 years old (SD ¼ 3.20). Approximately 86.1% identified
tionists and non-perfectionists (Hypothesis 2). Additionally, as White, 3.4% as Hispanic/Latino, 3.5% as Black/African
substance use will mediate the relationship between mal- American, 3.5% as Asian, 1.5% as Native American or
adaptive perfectionism and mental health such that mal- Alaskan, and the remaining 2.0% chose not to respond. For
adaptive perfectionism will be related to higher levels of Time 2, 78.0% were female and the average age was
substance use that will related to greater mental problems 19.18 years old (SD ¼ 4.81). Approximately 90.1% identified
(Hypothesis 3). as White, 2.8% as Hispanic/Latino, 1.2% as Black/African
In summary, the mental health and substance abuse con- American, 3.7% as Asian, 0.6% as Native American or
sequences of perfectionism type warrant careful investigation Alaskan, and the remaining 1.6% chose not to respond.
in the collegiate population, who specifically constitute a
vulnerable group due to their developmental stage in life,66
and are especially susceptible to multiple stressors in college
life. As such, building upon the conceptualization of perfec- Measures
tionism as a multidimensional trait, the current study aims Perfectionism
to understand the role that perfectionism plays in the use of Perfectionism was measured using the Almost Perfect Scale-
dysfunctional coping habits of substance use (i.e., smoking Revised scale68 (APS-R) which consists of 23 Likert-type
and drinking) and mental health in a time-lag design among items, ranging from 1 (strongly disagree) to 7 (strongly
college students (Figure 1). agree). Previous studies using this scale have demonstrated
good internal consistency among college students.20,69 The
Method scale consists of three facets of perfectionism: discrepancy
(12 items; a ¼ .94), high standards (7 items, a ¼ .87), and
Participants and procedure order (4 items, a ¼ .84). These items were modified to relate
Participants were recruited through SONA, a human subject to school. Example items include: “I often feel frustrated
pool system, at a public, Midwestern university, and received because I can’t meet my goals,” “I have high standards for
course credit that fulfilled the research requirement for my performance at work or school,” and, “I am an orderly
introductory-level courses. An explanation of the procedures person,” for each scale, respectively.
of the study was given and all subjects signed a consent
form. All study activities were approved by the Institutional
Review Board (IRB) of the participating university. Mental health
Participants completed the first survey at Time 1 (before The mental health of participants was measured using the
mid-term), which measured demographics and variables of General Health Questionnaire (GHQ-12).53 GHQ-12 is the
interest: perfectionism, mental health, and substance use. most extensively used screening instrument for common
Then, in Time 2 (after mid-term), approximately one month mental disorders, in addition to being a more general meas-
later (M ¼ 30.88 days), participants completed the second ure of psychiatric well-being.53 The questionnaire consists of
survey which measured mental health. Two separate data 12 Likert-type items (Time 1: a ¼ .88; Time 2: a ¼ .90),
collections occurred during two semesters: Fall 2018 and ranging from 1 (strongly disagree) to 5 (strongly agree).
Spring 2019.
Example items include: “I have been able to concentrate on
For Time 1, 951 students completed the survey. Attention
whatever I’m doing,” and “I have lost much sleep
check items (i.e., “What color is the sky?” and “What year is
over worry.”
it?”) were included in Time 1 and Time 2. A third attention
check item (i.e., “The state [name of] University is located
in?”) was included in Time 2. Attention check items have
been noted to help remove careless respondents and reduce Alcohol use
Type 1 error.67 Thirty students were removed for not pass- Alcohol use was measured using the Alcohol Use AUDIT-C
ing attention check items and 80 students were removed scale.63 The questionnaire consists of three questions per-
from analyses for duplicate responses (retention rate of taining to one’s alcohol consumption within the past year
88.5%). Of this, 336 students (40.0%) completed the second (Time 1: a ¼ .80; Time 2: a ¼ .78) each with separate
survey, as time allowed. As such, the final samples were 841 anchors ranging from 1 to 5. A sample item is: “How often
for Time 1 and 336 for Time 2. do you have a drink containing alcohol?”
4 S. K. NELSEN ET AL.

Table 1. Means, standard deviations, and ANOVA results by perfectionism clusters.

Perfectionism clusters
NP (n ¼ 273) MP (n ¼ 240) AP (n ¼ 295)
Variable Mean SD Mean SD Mean SD ANOVA
1. Personal standards .96BC .89 .47A .64 .50A .63 354.95
2. Order .93BC .84 .46A .71 .51A .66 335.88
3. Discrepancy .00BC .73 1.06AC .62 .86AB .54 606.33
4. General health (T1) 3.22BC .64 3.08AC .66 3.91AB .51 147.95
5. Smoking (T1) 1.58 4.68 1.13 4.84 1.41 5.41 .52
6. Alcohol (T1) 2.07C .88 1.92 .84 1.89A .82 3.70
7. General health (T2) 3.17C .65 3.07C 68 3.73AB .55 38.87
8. Smoking (T2) 1.00 3.03 .76 2.58 .58 2.24 .70
9. Alcohol (T2) 1.81 .75 1.77 .76 1.83 .80 .20
Note. For numbers 1–3, values are standardized z-scores. NP is non-perfectionists; MP is maladaptive perfectionists; AP is adaptive perfectionists, and T is time.
Significant at p ¼ .05.
Significant at p ¼ .01.
Significant at p ¼ .001. Different letters indicate which clusters have significantly different means by perfectionism cluster using Tukey post-hoc analysis. For
Time 1, n ¼ 779–808. For Time 2, n ¼ 311–314.

Table 2. Descriptive statistics and correlations.

Variable M SD 1 2 3 4 5 6 7 8 9
1. High standards 5.78 .87 (.87)
2. Discrepancy 3.95 1.25 .00 (.94)
3. Order 5.27 1.07 .51 .00 (.84)
4. GHQ (T1) 3.43 .70 .27 .64 .17 (.88)
5. Smoking (T1) 1.36 4.94 .06 .01 .10 .07 (–)
6. Alcohol (T1) 1.95 .84 .07 .02 .10 .00 .21 (.80)
7. GHQ (T2) 338 .69 .18 .54 .16 .73 .05 .02 (.90)
8. Smoking (T2) .74 2.48 .05 .03 .05 .03 .15 .09 .11 (–)
9. Alcohol (T2) 1.823 .78 .03 .05 .01 .04 .18 .83 .01 .11 (.78)
Note. GHQ is General Health Questionnaire and T is time. At Time 1, n ¼ 783–823. At Time 2, n ¼ 312–334.
Significant at p ¼.05.
Significant at p ¼.001. In bolded parentheses is the alpha score.

Smoking history Descriptive information, zero-order correlations, and

Smoking history was measured using the one-item Smoking Cronbach’s alpha reliability scores are presented in Table 2.
History Questionnaire (SHQ).70 The item is: “Think about To examine the effects of perfectionism on mental health
your smoking during the last week. How many times did (hypothesis 1), both ANOVA and repeated measures
you use cigarettes, cigars, smokeless tobacco, tobacco, or a ANOVA were used. At Time 1 (F(2, 792) ¼ 147.95, p <
vape in an average day?” .001), Tukey’s post-hoc test revealed that AP had signifi-
cantly higher mental well-being (M ¼ 3.91) than MP
(M ¼ 3.08, p < .001) and NP (M ¼ 3.22, p < .001) and that
Results NP had significantly higher levels of mental well-being than
MP (p < .05). Additionally, there were significant differen-
In line with previous research, before hypothesis testing, ces for mental well-being at Time 2 (F(2, 308) ¼ 38.87, p <
perfectionism clusters were identified using two-step cluster .001). Tukey’s post-hoc test revealed that AP (M ¼ 3.73) had
analysis.18,31,71–73 First, the standardized APS-R subscale significantly higher mental well-being than MP (M ¼ 3.07, p
scores were used in hierarchical cluster analysis using < .001) and NP (M ¼ 3.17, p < .001) while MP and NP
Ward’s linkage method with squared Euclidean distances. were not different. Repeated measures ANOVA (Figure 2)
The initial centroids were determined and then used in step revealed that there was a significant change in mental well-
two, which utilizes nonhierarchical, k-means cluster analysis being over time (F(1, 305) ¼ 11.40, p ¼ .001) and a signifi-
that specifies a set cluster solution and adds to the precision cant interaction between time and cluster (F(2, 306) ¼ 5.48,
of cluster analysis groups. The resulting clusters were com- p ¼ .005). Specifically, AP slightly decreased in mental well-
pared to determine each perfectionism cluster (Table 1). The being over time, contrary to our hypothesis, while NP and
first cluster (n ¼ 273), which had below average scores on MP remained relatively steady. As such hypothesis 1 was
the high standards facet, represented non-perfectionists partially supported.
(NP). The second (n ¼ 240) and third (n ¼ 295) clusters For hypothesis 2, the effects of perfectionism on sub-
were therefore considered perfectionists (based on above stance use were tested. There were no significant differences
average scores on the high standards facet). The second in smoking behaviors between the clusters at Time 1 or
cluster had the highest discrepancy score identifying this Time 2. In regards to alcohol use, ANOVA revealed a sig-
cluster as the maladaptive perfectionists (MP). The third nificant difference between clusters at Time 1 (F(2, 805) ¼
cluster was therefore labeled as adaptive perfectionists (AP). 3.70, p < .05) such that AP (M ¼ 1.89) had lower levels of
 JOURNAL OF AMERICAN COLLEGE HEALTH 5

Perfectionists (i.e., MP and AP combined); the second con-

trast compared MP to AP while excluding NP. Neither
smoking nor alcohol use mediated any relationship. These
findings fail to support hypothesis 3.

Discussion
In line with previous research on perfectionism,31–37 we
found support for Parker’s31 tripartite model of perfection-
ism by classifying the students into three clusters (i.e., adap-
tive, maladaptive, and non-perfectionists), and attempting to
show how these clusters are related to substance use and
mental health over time.
The current work contributes to the perfectionism and
mental health literature in college students in several ways.
First, while previous literature has linked perfectionism to
mental health,17,41–45 no studies have examined this relation-
ship using a time-lagged design in college students. As such,
the design of this study allowed for the investigation of the
Figure 2. General health differences over time separated by cluster. effect that time may have on the relationship between per-
fectionism and college students’ mental health, and further
allowed analysis to compare perfectionistic clusters in terms
of substance use and whether their substance use changed
over time.
Consistent with the literature, our findings suggest that,
over time, adaptive perfectionists continue to report better
mental health than maladaptive perfectionists and non-per-
fectionists.38–42 However, as the semester progressed, a
slight, yet significant decrease in mental well-being was
found such that both adaptive perfectionists and non-perfec-
tionists have slight decreases in mental well-being while
maladaptive perfectionists are relatively stable. These find-
ings provide mixed support regarding perfectionism type
and mental health. In the context of COR Theory, it was
theorized that adaptive perfectionists may have more resour-
ces to cope with stress and may use more adaptive coping
strategies (e.g., healthier lifestyle habits) to effectively cope
with their stress, thereby decreasing the strain that may have
occurred from stressors.75 This was not supported in this
study, as adaptive perfectionists slightly decreased in their
mental well-being throughout the semester.
Figure 3. Alcohol differences over time separated by cluster. Previous research has suggested that maladaptive perfec-
tionists are more likely to use dysfunctional, emotion-based,
or avoidance-type coping habits to manage their stress,
alcohol use than NP (M ¼ 2.07, p < .05). No differences while adaptive perfectionists are more likely to use problem-
were found at Time 2. Repeated measures ANOVA (Figure 3) focused strategies.76,77 However, the results of this study do
revealed that there was a significant change in alcohol use over not fully support this notion and indicate that perfectionism
time (F(1, 310) ¼ 10.49, p ¼ .001) and a significant interaction type does not influence smoking behaviors but may play a
between time and cluster (F(2, 310) ¼ 3.61, p < .05). role in drinking behaviors. Literature suggests cognitive
Specifically, AP remained steady in alcohol use while NP and styles may influence engagement in smoking behaviors.78 As
MP slightly decreased. As such, hypothesis 2 was par- such, self-control, conscientiousness, and perfectionism are
tially supported. all different cognitive styles that may contribute to smoking
To test hypothesis 3, each type of substance use (i.e., behaviors. One possible explanation as to why the current
smoking and alcohol use) was tested as a mediator of the work did not find support for the relationship between per-
relationship between perfectionism cluster and mental health fectionism type and smoking can be described by Cronk
with model 4 of the PROCESS macro for SPSS.74 As such, and Piasecki79 who suggested a variety of antecedents to
orthogonal contrast codes were created for comparisons. smoking such as location, social connections, socializing,
Specifically, the first contrast was that of NP versus alcohol consumption, and stress. As we did not measure
6 S. K. NELSEN ET AL.

these variables, future research can investigate their impact The findings have important implications for understand-
on smoking behaviors. Additionally, the college student ing the role of perfectionism in the substance use and men-
sample may have under or over-reported their smoking hab- tal health of college students. First, college health
its, and may have had a misunderstanding of how to answer professionals would do well to view perfectionism as a
the Smoking History Questionnaire. It is recommended that multidimensional construct with both adaptive and mal-
future research use a different scale to investigate whether adaptive subtypes. Our findings suggest that, if the ways in
there may be a better way of measuring smoking history. which students cope with stress is determined by one’s per-
Since smoking and drinking behaviors commonly co- fectionism type, college health professionals can use this
occur among college students58 and using the rationale of information to provide students with health seminars
the COR theory, it was also predicted that adaptive perfec- regarding appropriate coping strategies and ways to engage
tionists would have lower levels of drinking use than mal- in healthier lifestyle habits. Seminars could include provid-
adaptive perfectionists and non-perfectionists. Results ing an overview of the different types of coping strategies,
indicated some important differences between clusters in the importance of using adaptive and problem-focused strat-
terms of alcohol use. Specifically, there was a significant dif- egies, and how the use of specific strategies impacts one’s
ference in alcohol use between clusters at Time 1, with health. Furthermore, in cases where college health professio-
adaptive perfectionists having lower levels of alcohol use nals are found dealing with students with perfectionism ten-
than non-perfectionists. This provides support for hypoth- dencies, it may be advantageous to gauge the level of
esis two and partially relates to Rice and Van Arsdale’s adaptiveness of the perfectionism. By doing so, they may
investigation59 of the association between perfectionism and provide better recommendations on how students should
drinking as a coping mechanism, suggesting that non-per- productively cope with feelings of stress.59 Additionally, col-
fectionists use a more negative form of coping (drinking) leges could help combat the increases of drug use during
than adaptive perfectionists. Contrary to hypothesis 2, adap- stressful school periods (i.e., starting college, midterms, and
tive perfectionists remained steady in alcohol use while mal- finals) by providing stress-relief activities or smoking and
adaptive perfectionists and non-perfectionists slightly drinking interventions for students. Implementing such
decreased in alcohol use. This provides some support for the activities and interventions may be particularly useful as
COR theory. Specifically, it may be that maladaptive perfec- increases of drinking and smoking typically occur during
tionists and non-perfectionists were drinking too much and social settings and at times of heighted stress among stu-
therefore had to reduce their alcohol consumption as the dents. Witkiewitz and colleagues suggest that schools could
semester progressed. In contrast, it is possible that adaptive provide real-time interventions via cell phones in response
perfectionists, who are using more adaptive forms of coping, to fluctuations in stress or change in environmental con-
did not have to alter their alcohol intake as it was not affect- texts.80 By doing so, they may encourage more adaptive cop-
ing their academic performance. This information may also ing strategies and dissuade the use of dysfunctional coping
be useful for colleges. It may be that college students had habits. Finally, as Newman and colleagues18 suggested,
higher levels of alcohol use during Time 1, because data col- rather than simply trying to curtail perfectionistic tendencies
lection occurred primarily before and during the college’s in students, faculty and college health professionals could
mid-term schedule. As such, many students may have been encourage healthier ways to pursue academic excellence.
misusing alcohol more to cope with the stress of school Our findings should be interpreted in light of the follow-
(e.g., midterm tests, projects, and evaluations). However, ing potential limitations. First, all measures within this study
because data collection for Time 2 occurred after midterms, were single-source and self-reported. This may have intro-
this may have resulted in reduced amounts of alcohol use, duced common method bias.81 Besides, this study contained
as students had fewer stressors occurring during their life. 16 different measures (as it was part of a larger study),
Lastly, when examining how substance use may mediate which may have caused fatigue during survey completion.
the relationship between perfectionism type and mental Another limitation of this study includes the lack of diver-
health, results indicated that neither smoking nor drinking sity within the student sample, causing the sample to
behaviors explain the relationship. The current work used become more homogenous. In addition to the lack of diver-
COR theory’s conceptualization of loss and gain spirals in sity, the average age of the participants in the sample was
an attempt to explain how coping strategies may mediate roughly 19 years old. This may have skewed results as the
the relationship between perfectionism and mental health. legal limit of tobacco consumption is 18 and the legal limit
Particularly, adaptive perfectionists were expected to experi- of alcohol consumption is 21. As such, participants may
ence more gain spirals. Because adaptive perfectionists are have responded inaccurately to the measures of smoking
considered to have a greater ability to attain and maintain history and alcohol use, which would have an effect on anal-
resources, this would suggest that adaptive perfectionists yses. In addition, a one-item Smoking History Questionnaire
would have reduced levels of stress, thus, having greater was used and may have impacted participants’ under or
mental health in comparison to maladaptive perfectionists over-reporting of smoking. Although this work supports
and non-perfectionists.63 However, the results did not sup- previous literature that suggests perfectionism type influen-
port COR theory’s assumptions about resource gain spirals ces mental health, further research should continue to inves-
and loss spirals in relation to one’s perfectionism cluster and tigate perfectionism and health in a longitudinal manner to
mental health. better understand various mediators of the relationship.
 JOURNAL OF AMERICAN COLLEGE HEALTH 7

In conclusion, perfectionism type has been found to Rockville, MD: Center for Behavioral Health Statistics and Quality,
influence mental health and alcohol use. These findings pro- Substance Abuse and Mental Health Services Administration; 2019.
12. Lee CK, Corte C, Stein KF. Relationships between early alcohol
vide possible mechanisms in which colleges may better
experiences, drinker self-schema, drinking and smoking in col-
examine ways to increase students’ mental well-being among lege students. Subst Abus. 2018;39(4):426–433. doi:10.1080/
perfectionist clusters. The findings provide opportunities for 08897077.2018.144331426.
research to further investigate how other mediating variables 13. Hewitt PL, Mittelstaedt W, Wollert R. Validation of a measure
may play a role in the relationship between perfectionism of perfectionism. J Pers Assess. 1989;53(1):133–144. doi:10.1207/
s15327752jpa5301_14.
and mental health, which may help in creating, maintaining,
14. Curran T, Hill AP. Perfectionism is increasing over time: a
and retaining healthier students on college campuses. meta-analysis of birth cohort differences from 1989 to 2016.
Psychol Bull. 2019;145(4):410–429. doi:10.1037/bul0000138.
15. Curran T, Hill AP, Madigan DJ, Stornaes AV. A test of social
Conflict of interest disclosure learning and parent socialization perspectives on the develop-
ment of perfectionism. Pers Individ Differ. 2020;160:109925. doi:
The authors have no conflicts of interest to report.
10.1016/j.paid.2020.109925.
16. Kearney L, Baron A. Highlights of the Research Consortium 2002
Non-clinical Sample Study. Austin, TX: Counseling and Mental
Funding Health Center, University of Texas at Austin; 2003.
No funding was used to support this research and/or the preparation 17. Mehr KE, Adams AC. Self-compassion as a mediator of mal-
of the manuscript. adaptive perfectionism and depressive symptoms in college stu-
dents. J Coll Stud Psychother. 2016;30(2):132–145. doi:10.1080/
87568225.2016.1140991.
ORCID 18. Newman BN, Strickler JG, O’Brien C, Lui T, Lynch M.
Deconstructing perfectionism in college students: patterns of
Alper Kayaalp http://orcid.org/0000-0003-4762-2029 behavior, emotion, and cognition. Pers Individ Differ. 2019;145:
Kyle J. Page http://orcid.org/0000-0003-2458-1850 106–111. doi:10.1016/j.paid.2019.03.030.
19. Noble CM, Ashby JS, Gnilka PB. Multidimensional perfection-
ism, coping, and depression: differential prediction of depression
References symptoms by perfectionism type. J Coll Couns. 2014;17(1):80–94.
doi:10.1002/j.2161-1882.2014.00049.x.
1. Auerbach RP, Alonso J, Axinn WG, et al. Mental disorders 20. Suh H, Gnilka PB, Rice KG. Perfectionism and well-being: a
among college students in the World Health Organization World positive psychology framework. Pers Individ Differ. 2017;
Mental Health Surveys. Psychol Med. 2016;46(14):2955–2970. doi: 111(2017):25–30. doi:10.1016/j.paid.2017.01.041.
10.1017/S0033291716001665. 21. Adkins KK, Parker W. Perfectionism and suicidal preoccupation. J
2. Castillo LG, Schwartz SJ. Introduction to the special issue on col- Pers. 1996;64(2):529–543. doi:10.1111/j.1467-6494.1996.tb00520.x.
lege student mental health. J Clin Psychol. 2013;69(4):291–297. 22. Egan SJ, Wade TC, Shafran R. The transdiagnostic process of
doi:10.1002/jclp.21972. perfectionism. RPPC. 2012;17(3):279–294. doi:10.5944/rppc.vol.
3. Gallagher RP. National Survey of College Counseling Centers 17.num.3.2012.11844.
2014. Project Report. The International Association of 23. Eum KUn, Rice KG. Test anxiety, perfectionism, goal orienta-
Counseling Services (IACS), University of Pittsburgh, PA. 2015. tion, and academic performance. Anxiety Stress Coping. 2011;
4. Oswalt SB, Lederer AM, Chestnut-Steich K, Day C, Halbritter A, 24(2):167–178. doi:10.1080/10615806.2010.488723.
Ortiz D. Trends in college students’ mental health diagnoses and 24. Hamachek DE. Psychodynamics of normal and neurotic perfec-
utilization of services (2009–2015). J Am Coll Heal. 2020;68(1): tionism. Psychology. 1978;15:27–33.
41–51. doi:10.1080/07448481.2018.1515748. 25. Hewitt PL, Flett GL. Perfectionism in the self and social context:
5. Rafal G, Gatto A, DeBate R. Mental health literacy, stigma, and conceptualization, assessment, and association with psychother-
help-seeking behaviors among male college students. J Am Coll apy. J Pers Soc Psych. 1991;60(3):456–470. doi:10.1037/0022-3514.
Health. 2018;66(4):284–291. doi:10.1080/07448481.2018.1434780. 60.3.456.
6. American College Health Association. American College Health 26. Burns DD. The perfectionist’s script for self-defeat. Psych Today.
Association-National College Health Assessment III: Undergraduate 1980;14:34–52.
Student Reference Group Data Report Spring 2020. Silver Spring, 27. Hollender MH. Perfectionism. Comp Psychiatrry. 1965;6(2):
MD: American College Health Association; 2020. 94–103. doi:10.1016/S0010-440X(65)80016-5.
7. Healthy Minds Network. The impact of Covid-19 on College 28. Hollender MH. Perfectionism: a neglected personality trait. J
Student Well-being. 2020. https://healthymindsnetwork.org/wp- Clin Psychiatry. 1978;39(5):384.
content/uploads/2020/07/Healthy_Minds_NCHA_COVID_Survey_ 29. Frost RO, Marten P, Lahart C, Rosenblate R. The dimensions of
Report_FINAL.pdf. perfectionism. Cogn Ther Res. 1990;14(5):449–468. doi:10.1007/
8. Zhai Y, Du X. Mental health care for international Chinese stu- BF01172967.
dents affected by the COVID-19 outbreak. Lancet Psychi. 2020; 30. Slade PD, Owens RG. A dual process model of perfectionism
7(4):e22. 30089-4. doi:10.1016/s2215-0366(20). based on reinforcement theory. Behav Modif. 1998;22(3):
9. Mochrie KD, Whited MC, Cellucci T, Freeman T, Corson AT. 372–390. doi:10.1177/01454455980223010.
ADHD, depression, and substance abuse risk among beginning 31. Parker WD. An empirical typology of perfectionism in academ-
college students. J Am Coll Health. 2020;68(1):6–10. doi:10.1080/ ically talented children. Am Educ Res J. 1997;34(3):545–562. doi:
07448481.2018.1515754. 10.3102/00028312034003545.
10. National Institute on Alcohol Abuse and Alcoholism. SBIRT Has 32. Gilman R, Ashby JS. A first study of perfectionism and multidi-
Broad Impact on Adolescent Health. Published February, 2020. mensional life satisfaction among adolescents. J Early Adoles.
https://niaaa.scienceblog.com/275/sbirt-has-broad-impact-on-ado- 2003;23(2):218–235. doi:10.1177/0272431603023002005.
lescent-health/. 33. Gilman R, Ashby JS, Sverko D, Florell D, Varjas K. The relation-
11. National Survey on Drug Use and Health (NSDUH). Key ship between perfectionism and multidimensional life satisfaction
Substance Use and Mental Health Indicators in the United States: among Croatian and American youth. Pers Individ Differ. 2005;
Results from the 2018 National Survey on Drug Use and Health. 39(1):155–166. doi:10.1016/j.paid.2004.12.014.
8 S. K. NELSEN ET AL.

34. Moate RM, Gnilka PB, West EM, Rice KG. Doctoral student per- 55. Wills TA, Shiffman S. Coping and substance use: a conceptual
fectionism and emotional well-being. Meas Eval Couns Dev. framework. In: Shiffman S, Wills TA, eds. Coping and Substance
2019;52(3):145–155. doi:10.1080/07481756.2018.1547619. Use. Orlando: Academic Press; 1985:3–24.
35. Rice KG, Ashby JS, Gilman R. Classifying adolescent perfection- 56. Flett GL, Hewitt PL. Perfectionism and Maladjustment: An Overview
ists. Psychol Assess. 2011;23(3):563–577. doi:10.1037/a0022482. of Theoretical, Definitional, and Treatment Issues. Perfectionism:
36. Rice KG, Lopez FG, Richardson CME. Perfectionism and per- Theory, Research, and Treatment. Washington, DC: APA; 2002:5–31.
formance among STEM students. J Voc Behavior. 2013;82(2): 57. Flett GL, Goldstein A, Wall A-M, Hewitt PL, Wekerle C, Azzi
124–134. doi:10.1016/j.jvb.2012.12.002. N. Perfectionism and binge drinking in Canadian students mak-
37. Stoeber J, Otto K. Positive conceptions of perfectionism: ing the transition to university. J Am Coll Health. 2008;57(2):
approaches, evidence, challenges. Pers Soc Psychol Rev. 2006; 249–253. doi:10.3200/JACH.57.2.249-256.
10(4):295–319. doi:10.1207/s15327957pspr1004_2. 58. Pritchard ME, Wilson GS, Yamnitz B. What predicts adjustment
38. Aldea MA, Rice KG. The role of emotional dysregulation in per- among college students? A longitudinal panel study. J Am Coll
fectionism and psychological distress. J Couns Psychol. 2006; Health. 2007;56(1):15–21. doi:10.3200/JACH.56.1.15-22.
53(4):498–510. doi:10.1037/0022-0167.53.4.498. 59. Rice KG, Van Arsdale AC. Perfectionism, perceived stress, drink-
39. Rice KG, Ashby JS. An efficient method for classifying perfec- ing to cope and alcohol-related problems among college stu-
tionists. J Couns Psychol. 2007;54(1):72–85. 7285. doi:10.1037/ dents. J Coun Psyc. 2010;57(4):439–450. doi:10.1037/a0020221.
60. Dunkley DM, Blankstein KR, Halsall J, Williams M, Winkworth
0022-0167.54.1.72.
G. The relation between perfectionism and distress: hassles, cop-
40. Wu T, Wei M. Perfectionism and negative mood: the mediating
ing and perceived social support as mediators and moderators. J
roles of validating from others versus self. J Couns Psychol. 2008;
Couns Psyc. 2000;47(4):437–453. doi:10.1037/0022-0167.47.4.437.
55(2):276–288. doi:10.1037/0022-0167.55.2.276.
61. Zhang B, Cai T. Coping styles and self-esteem as mediators of
41. Ashby JS, Rice KG, Martin JL. Perfectionism, shame, and depres-
the perfectionism - depression relationship among Chinese
sive symptoms. J. Couns. Dev. 2006;84(2):148–156. doi:10.1002/j. undergraduates. Soc Behav Pers. 2012;40(1):157–168. http://doi.
1556-6678.2006.tb00390.x. org/mm7. doi:10.2224/sbp.2012.40.1.157.
42. Elion AA, Wang KT, Slaney RB, French BH. Perfectionism in 62. Bush K, Kivlahan DR, McDonell MB. The AUDIT alcohol con-
African American students: relationship to racial identity, GPA, sumption questions (AUDIT-C): an effective brief screening test
self-esteem, and depression. Cultur Divers Ethnic Minor Psychol. for problem drinking. Arch Intern Med. 1998;158(16):1789–1795.
2012;18(2):118–127. doi:10.1037/a0026491. doi:10.1001/archinte.158.16.1789.
43. Sherry SB, Hewitt PL, Flett GL, Harvey M. Perfectionism dimen- 63. Hobfoll SE, Halbesleben J, Neveu JP, et al. Conservation of
sions, perfectionistic attitudes, dependent attitudes, and depres- resources in the organizational context: the reality of resources
sion in psychiatric patients and university students. J Couns and their consequences. Annu Rev Organ Psychol Organ Behav.
Psychol. 2003;50(3):373–386. doi:10.1037/0022-0167.50.3.373. 2018;5(1):103–128. doi:10.1146/annurev-orgpsych-032117-104640.
44. Wei M, Heppner PP, Russell DW, Young SK. Maladaptive per- 64. Rice KG, Lapsley DK. Perfectionism, coping, and emotional
fectionism and ineffective coping as mediators between attach- adjustment. J Coll Stu Dev. 2001;42:157–168.
ment and subsequent depression: a prospective analysis. J Couns 65. Dunkley DM, Sanislow CA, Grilo CM, et al. Perfectionism and
Psychol. 2006;53(1):67–79. doi:10.1037/0022-0167.53.1.67. depressive symptoms 3 years later: negative social interactions, avoi-
45. Wei M, Mallinckrodt B, Russell DW, Abraham TW. Maladaptive dant coping, and perceived social support as mediators. Compr
perfectionism as a mediator and moderator between attachment Psych. 2006;47(2):106–115. doi:10.1016/j.comppsych.2005.06.003.
and negative mood. J Couns Psychol. 2004;51(2):201–212. doi:10. 66. Acharya L, Jin L, Collins W. College life is stressful today -
1037/0022-0167.51.2.201. Emerging stressors and depressive symptoms in college students.
46. Deuling JK, Burns L. Perfectionism and work-family conflict: J Am Coll Health. 2018;66(7):655–664. doi:10.1080/07448481.
self-esteem and self-efficacy as mediator. Pers Individ Diff. 2017; 2018.1451869.
116:326–330. doi:10.1016/j.paid.2017.05.013. 67. Abbey JD, Meloy MG. Attention by design: using attention
47. Flaxman PE, Menard J, Bond FW, Kinman G. Academics’ expe- checks to detect inattentive respondents and improve data qual-
riences of a respite from work: effects of self-critical perfection- ity. J Oper Manage. 2017;53:63–70.
ism and perseverative cognition on postrespite well-being. J Appl 68. Slaney RB, Rice KG, Mobley M, Trippi J, Ashby JS. The revised
Psychol. 2012;97(4):854–865. doi:10.1037/a0028055. almost perfect scale. Meas Eval Couns Dev. 2001;34(3):130–145.
48. Gnilka PB, McLaulin SE, Ashby J. Coping resources as mediators doi:10.1080/07481756.2002.12069030.
of multidimensional perfectionism and burnout. Cons Psychol J. 69. Rice KG, Ashby JS, Slaney RB. Perfectionism and the five-factor
model of personality. Assessment. 2007;14(4):385–398. doi:10.
2017;69:209–222.
1177/1073191107303217.
49. Ocampo ACG, Wang L, Kiazad K, Restubog SLD, Ashkanasy
70. Brown RA, Lejuez CW, Kahler CW, et al. Distress tolerance and
NM. The relentness pursuit of perfectionism: a review of perfec-
duration of past smoking cessation attempts. J Ab Psych. 2002;
tionism in the workplace and an agenda for future research. J
111(1):180–185. doi:10.1037/0021-843X.111.1.180.
Org Behav. 2020;41:144–168. doi:10.1002/job.2400. 71. Grzegorek JL, Slaney RB, Franze S, et al. Self-criticism, depend-
50. Hobfoll SE. Conservation of resources. A new attempt at concep-
ency, self-esteem, and grade point average satisfaction among
tualizing stress. Am Psychol. 1989;44(3):513–524. doi:10.1037// clusters of perfectionists and nonperfectionists. J Coun Psych.
0003-066x.44.3.513. 2004;51(2):192–200. doi:10.1037/0022-0167.51.2.192.
51. Hobfoll SE, Shirom A. Conservation of resources theory: applica- 72. Rice KG, Mirzadeh SA. Perfectionism attachment and adjust-
tions to stress and management in the workplace. In: ment. J Couns Psychol. 2000;47(2):238–250. doi:10.1037/0022-
Golembiewski RT, ed. Handbook of Organization Behavior. 2nd 0167.47.2.238.
ed. New York, NY: Dekker; 2001:57–81. 73. Martinent GC, Ferrand C. Cluster analysis of perfectionism
52. Molnar DS, Reker DL, Culp NA, Sadava SW, DeCourville NH. among competitive athletes. Psychol Rep. 2006;99(3):723–738.
A mediated model of perfectionism, affect, and physical health. J doi:10.2466/PR0.99.7.723-738.
Res in Pers. 2006;40(5):482–500. doi:10.1016/j.jrp.2005.04.002. 74. Hayes AF, Preacher KJ. Statistical mediation analysis with a mul-
53. Goldberg DP, Hillier VF. A scaled version of the General Health ticategorical independent variable. Br J Math Stat Psychol. 2014;
Questionnaire. Psychol Med. 1979;9(1):139–145. doi:10.1017/ 67(3):451–470. doi:10.1111/bmsp.12028.
S0033291700021644. 75. Folkman S, Moskowitz JT. Coping: pitfalls and promise. Annu
54. Lazarus RS, Folkman S. Stress, Appraisal, and Coping. New York: Rev Psychol. 2004;55(1):745–774. [Database] doi:10.1146/annurev.
Springer; 1984. psych.55.090902.141456.
 JOURNAL OF AMERICAN COLLEGE HEALTH 9

76. O’Connor R, O’Connor D. Predicting hopelessness and psycho- 79. Cronk NJ, Piasecki TM. Contextual and subjective antecedents
logical distress: the role of perfectionism and coping. J Couns of smoking in a college student sample. Nicotine & Tobacco Res.
Psychol. 2003;50:362–372. 2010;12(10):997–1004. doi:10.1093/ntr/ntq136.
77. Wills TA. Stress, coping, and tobacco use in early adolescence: 80. Witkiewitz K, Desai SA, Steckler G, et al. Concurrent drinking
relationships to substance use in urban school samples. Health and smoking among college students: an event-level analysis.
Psych. 1986;5(6):503–529. doi:10.1037/0278-6133.5.6.503. Psychol Addict Behav. 2012;26(3):649–654. doi:10.1037/a0025363.
78. Casper RC, Hedeker D, McClough, et al. Personality dimensions 81. Shadish WR, Cook TD, Campbell DT. Experimental and Quasi-
in eating disorders and their relevance for subtyping. J Am Acad Experimental Designs for Generalized Causal Inference. Boston:
Child Adolesc Psychiatry. 1992;31:830–840. Houghton-Mifflin; 2002.