Mental health and health behaviors among college student mentors in a randomized controlled trial interrupted by COVID-19

Abstract

Objective:

The benefits of mindfulness-training and mentoring for college students have yet to be investigated. We aimed to provide an exploratory and descriptive account of their potential benefits during the COVID-19 pandemic.

Participants:

In February 2020, 49 undergraduates (M = 20.51 years-old; 94% female) participated in a randomized trial of 12-week mentoring + mindfulness or mentoring-as-usual. After five weekly mentoring-sessions, programs were interrupted by COVID-19; mentoring continued online.

Methods:

Undergraduates completed questionnaires about mental health, behaviors, and regulatory processes in February and July 2020, with additional COVID-19-related questions at follow-up.

Results:

Participants reported moderate COVID-19-related perceived stress, but mental health, health behaviors, and regulatory processes did not diminish over time, with no condition differences. Undergraduates described using contemplative practices and social support to cope with COVID-19-stress.

Conclusions:

Undergraduates showed stable mental health/health behaviors despite moderate COVID-19-related-stress. Future research on mentoring with a mindfulness component among a larger and more heterogeneous sample will be necessary.

Mindfulness-based interventions (MBIs) are consistently associated with improvements in mental health (eg, anxiety/depression) and health behaviors (eg, sleep, eating, physical activity).¹ This evidence mostly comes from clinically based randomized controlled trials (RCTs) and there is a need to understand whether MBIs’ benefits can be translated into other real-world settings.¹ Existing youth service-organizations (YSOs) provide a potential context for MBI delivery, but have yet to be investigated. YSOs like mentoring programs often serve low-income communities with elevated emotional/behavioral concerns.^2,3 Although mentoring programs have a wide variety of benefits for mentees and mentors,^3,4 they do not typically rely on evidence-based interventions and tend to have modest effects.⁵ Incorporation of MBIs in mentoring may offer a novel approach to improve the well-being of mentors and mentees.

College student mentors are a unique, understudied group who often experience high stress associated with university and mentoring experiences.^3,6 Mentoring youth with emotional/behavioral concerns can introduce unique stressors, as mentors are faced with a range of specials needs and trauma responses. College students have high rates of mental health difficulties, which can be exacerbated by stressors.^6,7 Although mentoring can improve interpersonal skills and problem-solving,^3,4 little work has been conducted on improving mentors’ mental health and health behaviors. Mentoring with a mindfulness-training component may be especially beneficial for mentor health, because fostering mindfulness, present-focused attention and non-judgmental awareness, can offer helpful regulatory support during times of stress.⁸ Systematic reviews and meta-analyses of MBIs among college students suggest that MBIs improve mental health, health behaviors, and regulatory outcomes like emotion regulation.^9,10 Incorporation of mindfulness-training into a mentoring program may, therefore, provide college student mentors with greater mindfulness and emotion regulation, and ultimately better mental health and health behaviors.

To test this notion, we implemented an RCT of mindfulness-training and mentoring (mindfulness + mentoring), compared to mentoring-as-usual, to investigate the distinct benefits of incorporating mindfulness in a mentoring program for undergraduate mentors’ well-being. Unfortunately, the original aim was interrupted by COVID-19. During COVID-19, anxiety/depression among college students was moderate-to-severe and increased in prevalence.^11,12 Perceived stress and family conflict also increased,^7,11 and health behaviors deteriorated, with increased food intake, emotional eating, sleep disturbances, and physical inactivity.^13–16 In a systemic review, common risk factors for distress during COVID-19 included being a student, identifying as female, and having frequent media exposure.¹⁷ Furthermore, greater COVID-19 perceived stress has been related to more media exposure, worse mental health, and more stress-related health behaviors.¹⁷ Emerging literature suggests that MBIs may be helpful to college students during COVID-19. For instance, a single-arm 16-day, virtual MBI for college students decreased stress/anxiety within-subjects during COVID-19, and a four-week (75 min/week) virtual MBI improved mindfulness and sleep, relative to a control, among predominately male college students.^18,19

Although we could not assess our original aim, we sought to add incrementally to the extant literature by exploring longitudinal changes from February to July (2020) in mental health, health behaviors, and regulatory processes (mindfulness/emotion regulation) in a sample of college student mentors as a whole and by study condition (mentoring + mindfulness versus mentoring-as-usual). We also explored COVID-19-related outcomes, including perceived stress, media exposure, and family conflict. A qualitative account of college students’ coping with COVID-19-related stress was also examined. Participation in a mentoring program was expected to protect students from population-wide declines in mental health, health behaviors, and regulatory processes.^7,11–16 Participation in mentoring + mindfulness was hypothesized to have more positive outcomes, compared to mentoring-as-usual. We also expected moderate-to-high COVID-19-related perceived stress and that stress would relate to greater COVID-19-related media exposure, family conflict, regulatory difficulties, and poorer mental health/health behaviors.

Method

Participants and procedures

Eighty college students involved in a service-learning mentoring program (a credit-bearing educational experience in which students served as mentors to at-risk community teens⁴) completed online surveys at the start of the mentoring program (~February 2020; baseline) and two-months after the program ended (~July 2020; follow-up). Forty-nine students (18–25 years, M = 20.51, SD = 1.34; 94% female) who were predominately middle-class and Caucasian (76%; 4% Asian/Pacific-Islander, 4% African-American, 4% American-Indian; 12% multiple/other races/ethnicities) participated in both surveys with no significant demographic differences between those who did and did not complete the follow-up survey (p-values > .11).

Across 12-weeks, college students engaged in activities with their adolescent mentee, including building positive relationships and supporting academic success.^3,4 Participants were randomized to mentoring-as-usual or mentoring + mindfulness: mindfulness-training was derived from Learning to BREATHE (L2B).²⁰ L2B was delivered for 30-min per night, embedded within mentoring sessions. College students received four L2B sessions (44% of intended dose) alongside their mentee before COVID-19 closures. The mentoring program, but not the mindfulness-training, continued online. The University Institutional Review Board (IRB) approved all procedures.

Measures

Participants reported internalizing/externalizing symptoms and attention problems on the Brief Problem Monitor.²¹ Post-traumatic stress disorder (PTSD) symptoms were measured on the Child PTSD Symptom Scale, which includes parallel items as the adult Post-traumatic Diagnostic Scale with simplified language.²² One item (‘feeling angry’) was omitted due to a programming error. Sleep quality and physical activity were self-reported on PROMIS Sleep-Related Disturbance and Pediatric Physical Activity short forms, respectively.^23,24 Participants completed the Reward-based Eating Drive Scale and Difficulties in Emotion Regulation Scale-Short Form.^25,26 The Five Facets of Mindfulness Questionnaire-Short Form was completed to measure non-reactivity, observing, acting with awareness, describing, and non-judging.²⁷ At follow-up, participants completed the Perceived Stress Scale adapted by the authors to measure appraisals of COVID-19-specific stressors (eg, “In the last month, how often have you felt nervous and stressed because of the novel coronavirus?”).²⁸ No psychometrically validated scales to assess COVID-19-impact were available at that time. To measure COVID-19-related media exposure, the number of hours per day that they spent listening to, watching, or reading about COVID-19 were reported; a parallel item measured media exposure of those around them. To measure COVID-19-related conflict with parents, participants reported on how often they fight with parents/caregivers about social distancing. Participants also responded to one open-ended question about coping with COVID-19-related stress (“What have you done to feel less stressed?”) and two Likert-scale questions about how well the strategies endorsed worked and difficulty of use.

Data analysis

No variables were non-normal. Baseline demographic characteristics were compared for mentoring + mindfulness vs. mentoring-as-usual, with no significant differences (p-values > .08). Generalized estimating equations (GEE), which are well-suited to handle repeated measurements,²⁹ were conducted in Stata with bias-corrected sandwich variance estimation to account for the small sample size.³⁰ GEE was used to analyze changes over time for the whole sample and by condition. Listwise deletion was used to handle missing data (baseline: <6% missing; follow-up: 12–20% missing). Descriptive statistics and bivariate correlations were examined to provide information about COVID-19-related impacts at follow-up. Basic content analysis was used to identify themes in the open-ended response about COVID-19-related coping.³¹ Descriptive information was generated for how well strategies worked and difficulty of use. Chi-square analyses, Fisher’s exact tests, and t-tests were conducted to analyze condition effects for all COVID-19-related measures.

Results

There were no significant changes in mental health, health behaviors, nor regulatory factors over time (p-values > .10) or by condition (p-values > .15; Table 1). Supplemental Figure 1 displays mean change scores from baseline to follow-up. The average college student perceived feeling moderately stressed about COVID-19 (M = 18.35, SD = 6.89). Seventy-four percent reported ‘never’ or ‘rarely’ fighting with parents about COVID-19-related social distancing. Forty-six percent reported <1 h of daily COVID-19 media exposure; 46% reported 1–3 h; and 8% reported >3 h. Another 44% reported being around others who were listening to or reading about COVID-19 for 1–3 h per day. There were no differences in stress, conflict, or media exposure by condition (p-values > .88).

Table 1.

Summary of generalized estimating equations predicting change over time by condition.

	Time			condition			Time*condition			constant			Racea
	b	SE	p	b	SE	p	b	SE	p	b	SE	p	b	SE	P
Internalizing	−.21	.38	.58	−.98	.93	.29	.29	.62	.65	4.26	.61	<.001	-	-	-
Externalizing	.12	.25	.63	.47	.60	.43	.03	.44	.94	1.32	.34	<.001	-	-	-
Attention Problems	−.54	.34	.12	−.65	.75	.39	−.10	.44	.83	3.42	.52	<.001	-	-	-
PTSDb Symptoms	−1.08	1.99	.59	−2.65	4.12	.52	.79	2.79	.78	14.78	2.74	<.001	-	-	-
Sleep	.32	.68	.67	−.57	1.65	.73	−.05	.93	.96	11.55	1.12	<.001	.53	1.04	.62
Physical Activity	1.16	.85	.18	2.56	1.87	.18	−1.62	1.25	.20	10.44	1.18	<.001	-	-	-
Reward-based Eating	−2.23	1.42	.12	−3.30	2.98	.27	.25	1.94	.90	21.89	2.07	<.001	-	-	-
Emotion Reg. Difficultiesc	−.35	.38	.36	−.22	1.34	.87	−.82	.57	.15	12.63	.88	<.001	1.65	1.22	.18
Non-Reactivity	.10	.46	.84	1.82	1.42	.20	−.60	.89	.50	15.58	.78	<.001	-	-	-
Non-Judgment	.44	.72	.54	.59	1.45	.69	.55	.96	.57	14.39	1.04	<.001	-	-	-
Describing	.05	.46	.92	−.87	1.21	.47	.55	.70	.43	18.38	.87	<.001	-	-	-
Observing	.70	.42	.10	−.43	1.36	.76	−.76	.75	.32	13.57	.82	<.001	.92	.82	.27
Act Aware	.28	.52	.60	.95	1.26	.45	−.85	.84	.32	17.00	.81	<.001	-	-	-

Note. Time represents pre- and post-assessment change. Condition represents mentoring + mindfulness vs. mentoring-as-usual. b = Unstandardized estimate; SE = Standard error.

^aRace was dichotomized to run analyses (0:Caucasian; 1:Other) and only included in analyses where it predicted significant change;

^bPost-traumatic stress disorder;

^cEmotion regulation difficulties..

At follow-up, COVID-19-related stress was negatively correlated with mindful non-reactivity (r = −.44, p = .005) and positively correlated with PTSD symptoms (r = .39, p = .02; but with no other variables; Supplemental Table 1).

Participants provided 106 total open-ended responses about coping. Eighty percent reported <1 strategy. From reported strategies, eight coping themes emerged: contemplative practice (n = 37, 35%), social support (n = 18, 17%), health behaviors (n = 13, 12.3%; eg, sleeping/exercise), nature/walking (n = 10, 9%), multi-media consumption (n = 9, 8.5%), self-expression (n = 9, 8.5%; eg, journaling), distraction (n = 6, 5.7%), and solitude (n = 4, 3.8%). Of the most frequently reported theme, contemplative practice, there were five sub-categories: breathing (n = 16, 43%), mindfulness (n = 9, 24%), meditation (n = 4, 11%), yoga (n = 3, 8%), prayer (n = 3, 8%), and purposeful attention to emotion (n = 2, 5%). Endorsement of contemplative practice was higher in mentoring + mindfulness (80%) than mentoring-as-usual (45%; p = .08). Most college students reported that coping skills worked ‘frequently’ or ‘extremely well’ (n = 27; 69.3%) and were ‘not at all’ difficult to use (n = 18, 46%).

Discussion

The goal of this brief report was to provide a descriptive, exploratory account of longitudinal changes in mental health, health behaviors, and regulatory process among college students involved in a service-learning mentoring program, with and without a brief mindfulness-training component, during COVID-19. We also explored COVID-19-related outcomes and qualitative accounts of coping with COVID-19-related stress. Although there is evidence that COVID-19 has exacerbated mental health concerns globally,¹⁷ college student mentors showed no significant change in mental health, health behaviors, or regulatory processes. In line with the existing literature,¹¹ contemplative practices, as well as social support, were among the most frequently reported coping skills utilized by students. It is possible that participation in a mentoring program, with or without mindfulness-training, provided students with adequate social support and coping to promote resilience during COVID-19. Indeed, participation in mentoring has demonstrated similar benefits (eg, improved interpersonal skills/self-esteem),⁴ but the benefits of a service-learning mentoring program have yet to be investigated within the context of a global pandemic. Considering current literature on the benefits of mindfulness for college students during COVID-19,^18,19 the current findings also highlight how four 30-min sessions of an MBI (prior to COVID-19) may be insufficient to produce differential findings five-months into the pandemic. Tentatively, these results generate the hypothesis that it could be helpful for those struggling with COVID-19-related stress to become involved in mentoring activities. However, a longer, more rigorous investigation of the incorporation of MBIs within mentoring programs may be warranted to parse out the benefits of mindfulness for undergraduate mentors.

Findings from qualitative content analysis that contemplative practices and social support were among the most frequently reported coping skills yield information about common types of contemplative practices utilized (eg, breathing/mindfulness, yoga) and information about the relative ease and usefulness of these coping skills. These strategies may be effective and easy to incorporate into daily life,¹¹ suggesting it may be valuable to encourage increased practices like mindfulness and yoga, as well as reliance on social support. Despite the abbreviated dosage, undergraduates in mentoring + mindfulness showed a non-significant trend toward endorsing more contemplative coping skills, with double the frequency (80% vs. 45%) than those in mentoring-as-usual. Going forward, it will be important to test the potential, distinctive benefits of mentoring + mindfulness compared to mentoring-as-usual with a more complete administration of the mentoring + mindfulness program.

As anticipated, greater PTSD symptoms and lower mindful non-reactivity related to greater COVID-19 perceived stress. PTSD symptomatology can be high during epidemics,¹⁴ and dispositional mindfulness appears to assist with regulating distress during periods of high stress, like COVID-19.³² Alternatively, we expected that COVID-19 perceived stress would relate to media exposure, conflict with parents, and other stress-related behaviors/regulatory factors,^11,17 but none of these variables were significantly correlated. Thus, PTSD symptoms and the facet of mindful non-reactivity may be important to investigate further.

Several limitations may impact the generalizability of this study. The sample size was small, and there was no non-mentoring control group. Mindfulness-training was also truncated due to COVID-19, which limited the intended differentiations between the treatment and control group. Furthermore, data relied on self-report only and the sample was mostly female, predominately middle-class, and Caucasian, limiting generalizability.

In sum, college student mentors did not experience significant changes in mental health, health behaviors, or regulatory processes. Although the average student reported feeling moderately stressed about COVID-19, college students also reported utilizing coping skills such as contemplative practices and social support to lessen stress. These results tentatively suggest that it may be important to encourage college students struggling with COVID-19 to utilize mindfulness and social engagement to cope with uncontrollable stressors. More research is needed to test the benefits of a mentoring program with a mindfulness component among a larger and more heterogeneous sample.

Supplementary Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/07448481.2022.2086007.