The effects of mindfulness meditation on burnout in clinical genetic counselors: A three‐arm randomized controlled trial

Abstract

Burnout is common among genetic counselors (GCs). Clinician burnout has been found to adversely affect individual well‐being, patient care, and the likelihood of staying in a role. Both individual and system‐level solutions are needed to address burnout. Mindfulness meditation (MM) is one individual‐level solution that has shown promise for reducing burnout in clinicians, but has not been studied in GCs. We conducted a decentralized, parallel, three‐arm randomized controlled trial comparing MM to a novel active control meditation (ACM) and a no‐meditation control (NMC), with 1:1:1 randomization. Participants were clinical GCs in the United States. MM and ACM participants were asked to do 10 min of daily app‐based meditation for 8 weeks. ACM was designed to control for nonspecific aspects of MM by mimicking MM length and structure without including mindfulness techniques. The primary outcome, burnout, was assessed using the Professional Fulfillment Index. Secondary outcomes included other indicators of professional well‐being, such as stress and professional fulfillment. Outcomes were assessed via an intention‐to‐treat approach, with multiple imputation for missing outcome data. Outcome analyses controlled for baseline trait mindfulness. Three hundred and ninety‐seven participants (median age 31 years; 97.7% female, 94.2% White) were randomized, and 76% completed post‐intervention outcome measures. At baseline 57% of GCs met criteria for burnout. There was no difference in burnout reduction between MM and ACM groups (p = 0.53). However, multiple measures suggest that ACM did not perform well as an active meditation control, thus the primary hypothesis could not be effectively tested. In prespecified secondary analyses, MM reduced burnout (Cohen's d = −0.84, p < 0.001) compared to NMC, a passive control. Similar results were seen for stress. These findings suggest MM may be beneficial for GC professional well‐being; however, further research on MM for GCs is needed with more diverse study samples and better active controls.

What is known about this topic

Burnout is common among genetic counselors. Mindfulness meditation has shown promise for reducing clinician burnout but has not been studied in genetic counselors.

What this paper adds to the topic

Mindfulness meditation reduced burnout and stress among clinical genetic counselors when compared to a passive control. Comparison to an active control was not successful due to methodological issues.

1. INTRODUCTION

Burnout is common among genetic counselors (GCs), with a point prevalence over 50% (Allsbrook et al., 2016; Bernhardt et al., 2009; Caleshu et al., 2022; Johnstone et al., 2016; Lee et al., 2015; Udipi et al., 2008). Burnout is characterized by emotional exhaustion, depersonalization, and reduced sense of personal accomplishment (Maslach, 2001). Beyond its impact on individual well‐being, burnout is a significant factor in clinicians leaving direct patient care (Hayes et al., 2006; National Academies of Sciences, Engineering, and Medicine, 2019; Williams et al., 2001). GCs consistently report burnout as the primary reason they consider leaving the field (National Society of Genetic Counselors, 2023). Furthermore, clinician burnout has been associated with decreased quality of patient care and increased medical errors across healthcare settings (National Academies of Sciences, Engineering, and Medicine, 2019; Niconchuk & Hyman, 2020; Shanafelt et al., 2010). Clinician burnout imposes significant costs on healthcare institutions through increased turnover, reduced productivity, and loss of expertise (Office of the Surgeon General (OSG), 2022). The annual attributable cost of physician burnout in the United States has been estimated at approximately $4.6 billion (Han et al., 2019). The National Academies released a report in 2019, which synthesized the large body of literature on clinician burnout and proposed a multilayered system‐based model of the origins of burnout, noting that interventions are needed at all levels of the healthcare system (National Academies of Sciences, Engineering, and Medicine, 2019). Furthermore, they emphasize that professional well‐being goes beyond just the absence of burnout, to include the presence of positive states like professional fulfillment, engagement, and a sense that one's work is meaningful.

Mindfulness meditation has emerged as a promising individual‐level intervention that, in combination with system‐level interventions, can improve professional well‐being in clinicians (Shoker et al., 2024). Mindfulness involves paying purposeful attention to present‐moment experiences with an attitude of acceptance and non‐judgment (Kabat‐Zinn, 2003). This capacity is most often cultivated through mindfulness meditation (MM), which involves techniques such as focused attention on the breath or other sensations, noting when the mind has wandered, and gently redirecting attention to present‐moment experience, without judgment (Bishop, 2002). Studies examining MM interventions among clinicians have demonstrated reductions in stress, anxiety, burnout, and emotional exhaustion (Goldberg et al., 2022; Irving et al., 2009; Lamothe et al., 2016; Shoker et al., 2024). Our previous research found that self‐reported mindfulness among GCs was negatively associated with burnout and compassion fatigue, and positively associated with empathy and work engagement (Silver et al., 2018). Thus, MM in clinicians at risk for burnout may be an important strategy for retention and sustainable quality care in GCs.

Despite these promising findings, a significant limitation of MM research is the reliance on passive control groups, which fail to control for nonspecific factors such as expectancy effects or structured self‐care time (Bishop, 2002; Davidson & Kaszniak, 2015; Goldberg et al., 2017, 2019). Meditation trials with passive controls can overestimate effect sizes by 30%–40% compared to those with active controls (Goldberg et al., 2022; Goyal et al., 2014). Inert active controls that mimic nonspecific aspects of MM without incorporating mindfulness techniques are essential for accurately evaluating mindfulness's unique contribution to outcomes. However, these have been challenging to design and no gold standard active control for MM exists (Davidson & Kaszniak, 2015; Kinser & Robins, 2013).

We sought to test the hypothesis that MM could improve professional well‐being among GCs providing direct patient care in the United States. We designed a study comparing MM to both an active control meditation (ACM) and a no‐meditation control (NMC) group. The ACM meditation was novel, informed by prior studies (Lindsay et al., 2019; Manocha et al., 2011; Ruscio et al., 2016) and designed to match the MM intervention in structure, duration, and frequency, while excluding explicit mindfulness techniques. Our primary aim was to test whether MM would be more effective than a well‐matched ACM at reducing burnout among clinical GCs. The NMC group was included to allow comparison to a no‐meditation passive control if the novel ACM arm did not perform well as an active control.

2. METHODS

2.1. Study design

We conducted a double‐blinded, parallel three‐arm decentralized randomized controlled trial (RCT) with randomization 1:1:1 to three study arms: MM, ACM, and NMC. Participants were blinded to the study hypotheses. Individuals performing data analysis were blinded as to which arm participants were randomized. No protocol deviations or changes occurred. The primary outcome was burnout. Secondary outcomes included stress, professional fulfillment, professional self‐care, resilience, reactive distress, and a desire to spend less time caring directly for patients.

2.2. Study arms

2.2.1. Mindfulness meditation (MM) arm

The intervention was 8 weeks of 10 min a day of MM via Headspace, a web and smartphone app that provides MM exercises, guided by audio instructions from a teacher with extensive training in meditation practice and meditation instruction. Participants were asked to do specific guided meditations in Headspace that included MM techniques like awareness of breath and sensations, noting of thinking or distraction, and gently returning attention to the breath. They were instructed to first do the Headspace Basics courses (30 guided meditation sessions that include the fundamentals of MM) then, once they had finished with Basics courses, to select from guided meditations that apply these fundamentals to managing anxiety, patience, restlessness, and navigating change (Radin et al., 2025). The intervention duration was chosen because most of the evidence for MM is from trials of 8‐week interventions and sufficient evidence is not yet available for shorter intervention periods. Prior studies have shown that MM guided by Headspace can improve psychological symptoms in general (Bostock et al., 2019; Flett et al., 2020; Zawadzki et al., 2025), and professional well‐being in healthcare workers specifically (Radin et al., 2025; Taylor et al., 2022).

2.2.2. Active control meditation (ACM) arm

Given that contemplative science currently lacks a gold standard control for MM, we developed a novel active control that was untested prior to this study. It was intended to be inert, meaning it did not include the active ingredient of mindfulness training. We created meditation exercises designed to mirror the structure (web‐based, audio‐guided), duration, and dose (10 min per day over 8 weeks) of MM, without mindfulness techniques. We drew from active control designs in prior studies on MM, where the control included elements of mind‐wandering, introspection, and/or intentional thinking (Lindsay et al., 2019; Manocha et al., 2011; Ruscio et al., 2016). Audio instructions guided participants to close their eyes and get lost in their thoughts or reflect on their day, with no instruction regarding mindfulness (i.e., attention to breath and body sensations, acceptance of thoughts and feelings that arise, returning attention to the breath after noticing mind wandering to distraction). Mind wandering and getting lost in thoughts was chosen since it is distinct from mindfulness techniques which often encourage awareness of thought and nonjudgmental observation of them, without getting lost in them or engaged with them, as well as returning attention to body sensations or other present‐moment anchors once aware that attention has drifted to thoughts. Since this active control was novel, we included manipulation checks (change in mindfulness) and assessments of its credibility and expectancy (described below) to help assess its performance as an active control.

2.2.3. No‐meditation control (NMC) arm

Participants were asked not to engage in any meditation exercises. NMC was included for secondary comparisons to MM in case ACM did not perform well as an active control.

2.3. Sample size determination

The target sample size was determined based on a power analysis conducted for the primary outcome, burnout, which was measured via the Professional Fulfillment Inventory (PFI), comparing MM to ACM. To detect a 0.5 standard deviation difference in burnout scores, at a significance level of p < 0.05 and with 80% power, 213 participants are needed. We increased this to 387 to account for an expected lost‐to‐follow‐up rate of 25%, lack of published data on PFI scores in this population, and an estimate that ~15% of our sample would have a regular meditation practice and thus may not respond to the intervention (Silver et al., 2018).

2.4. Participants

Recruitment occurred from September 2019 to July 2020 via NSGC e‐blasts, NSGC forum posts, Twitter, and emails to participants in a prior study on this topic by our group (Silver et al., 2018). Recruitment materials noted that the study was on poor professional well‐being, including burnout, but did not mention mindfulness as the intervention. Eligibility criteria included self‐described English proficiency and being a GC in the United States who counsels patients. The study was limited to GCs who counsel patients because we also sought to test the hypothesis that MM can improve key counseling variables (e.g., working alliance, empathy, unconditional positive regard; see NCT03723018). These secondary outcomes related to direct patient care will be reported in a subsequent manuscript.

2.5. Study procedures

Study procedures are outlined in Figure 1 and the CONSORT flow diagram is in Figure 2. This was a decentralized trial with participants completing all study procedures online.

FIGURE 1.

Study procedures. Timeline and flow of study procedures for the three‐arm randomized controlled trial. Participants completed the T0 (baseline) survey then were randomized 1:1:1 (MM, active control, no‐meditation control). After the 8‐week intervention period, participants were sent the T1 (post‐intervention) survey. Participants in the MM arm completed an additional survey (T2) 26 weeks after randomization to assess maintenance of intervention effects.

FIGURE 2.

CONSORT diagram depicting participants' flow through the trial. Flow of participants through screening, randomization, intervention, and follow‐up phases of the three‐arm randomized controlled trial. MM = mindfulness meditation, ACM = active control meditation, NMC = no‐meditation control. ^aNumber of online eligibility surveys completed. The number of unique individuals who completed the eligibility survey may be less than 773 due to some individuals filling it out more than once. ^bIn order to be randomized, eligible participants had to complete the T0 survey. ^cReasons participants did not meditate were not collected. Eight sessions was used as a cutoff because in the most studied mindfulness intervention, Mindfulness‐Based Stress Reduction, 8 meditation sessions would be equivalent to only doing meditation in the weekly group sessions. ^dParticipants were considered lost to follow‐up if they did not complete the T1 measure of burnout. Reasons for not completing that measure were not collected.

2.5.1. Randomization

Participants were randomized after completing the baseline (T0) survey so that randomization did not bias their T0 responses. Blocked randomization was used in order to increase the likelihood of balanced study group sizes. The randomization sequence was generated by the study statistician and randomization was assigned using the randomization module in REDCap (Harris et al., 2009, 2019).

2.5.2. Data collection

Participants completed a survey at baseline (T0) and again after the 8‐week intervention period (T1). Multiple reminders were sent when surveys were not completed. MM participants additionally completed a follow‐up survey (T2) 26 weeks after randomization (Figure 1). Surveys were administered via REDCap.

2.6. Performance of ACM

Given the novel nature of the ACM arm and the known challenges in designing active controls for meditation, we collected data to help assess how well the ACM meditation performed as an active control. This included credibility and expectancy, change in mindfulness from T0 to T1, and adherence to meditation instructions.

2.7. Instrumentation

Table S1 lays out the scales used, their properties (length, reliability, internal consistency), and when each was administered. Brief descriptions of each are provided below. Internal consistency of all scales in the present sample was assessed using Cronbach's α (Table S2).

2.7.1. Mindfulness

The Five‐Factor Mindfulness Questionnaire (FFMQ) is a trait measure of dispositional mindfulness, developed using input from other existing mindfulness measures (Baer et al., 2006). It is sensitive to change with interventions and has been used in many studies of MM.

2.7.2. Credibility and expectancy

The Credibility and Expectancy Questionnaire (CEQ) evaluates how believable participants find the intervention and their expectation of benefit. It is widely used in behavioral intervention trials to assess nonspecific effects such as placebo or expectancy (Devilly & Borkovec, 2000).

2.7.3. Adherence to meditation

For both ACM and MM, data on initiation of meditation sessions was collected via the platform that delivered the meditation. At T1, MM and ACM participants' adherence to meditation instructions was assessed. We used investigator‐created Likert‐based questions to assess how often participants practiced techniques specific to each arm (e.g., focus on breath for MM and mind‐wandering for ACM) then calculated a score for adherence to each type of meditation (MM, ACM). An open‐ended question asked participants to describe what they did during meditation.

2.7.4. Perceived harms and benefits

At T1, MM and ACM participants were asked if they experienced any harms or negative effects, or any benefits and, in an open ended question, what those were.

2.7.5. Prior experience with meditation and mindfulness

At T0, participants were asked if they'd undergone prior meditation training and how often they meditate. At T1, participants were asked if they had prior mindfulness training.

Investigator‐created questions were used to measure demographics (T1).

2.7.6. Primary outcome

2.7.6.1. Burnout

We used the burnout subscale of the PFI, which consists of the work exhaustion and interpersonal disengagement subscales (Trockel et al., 2018). The PFI uses a 5‐point Likert scale (0–4, where 0 = not at all true and 4 = completely true). A score ≥14 indicates likely burnout, as defined by the scale developers. It was selected because it is briefer than other measures of professional well‐being (yet maintains convergent validity with them), is designed to detect change in intervention studies, and allows for measurement of professional fulfillment in addition to burnout (Miller, 2024; Trockel et al., 2018). It also provides a cut‐off for burnout that assists in assessing point prevalence of burnout while the current version of the Maslach Burnout Inventory (MBI) does not.

2.7.7. Secondary outcomes

2.7.7.1. Professional fulfillment

The professional fulfillment subscale of the PFI “measures the intrinsic positive reward a clinician experiences from their work, encompassing dimensions such as happiness, meaningfulness, satisfaction, self‐worth, and sense of control” (Trockel et al., 2018).

2.7.7.2. Stress

The Perceived Stress Scale (PSS) is a measure of stress that has been extensively used in intervention studies, especially those targeting stress reduction (Cohen et al., 1983).

2.7.7.3. Professional self‐care

The Professional Self‐Care Scale (PSCS) measures the extent to which professionals engage in activities that may support their well‐being. Subscales include professional support, professional development, life balance, cognitive awareness, and daily balance (Dorociak et al., 2017).

2.7.7.4. Resilience

The Connor‐Davidson Resilience Scale (CD‐RISC) evaluates personal qualities and resources that help individuals adapt to adversity. It is widely used to measure resilience in clinicians and other populations (Connor & Davidson, 2003).

2.7.7.5. Reactive distress

Reactive distress refers to discomfort and distress experienced in response to another person's negative experience. It was measured using the personal distress subscale of the Interpersonal Reactivity Index (IRI), which assesses an individual's tendency to experience distress in response to observing others' difficult experiences (Davis, 1983).

2.7.7.6. Desire to reduce clinical time

This was assessed using an investigator‐created question with a Likert scale that asked participants if they would like to reduce the amount of time they spend providing clinical care, including all aspects of clinical care.

2.8. Data analysis

We used a prespecified intention‐to‐treat analysis for the primary outcome and all secondary outcomes. Outcomes were analyzed using linear regression with the T1 score for each outcome as the dependent variable, study arm as a predictor, and the T0 score for the outcome as a covariate. T0 mindfulness (FFMQ) was also included in the model to control for dispositional or previously developed mindfulness, which has been found to moderate the effects of mindfulness interventions (Sieder et al., 2024). For each outcome, three pairwise comparisons were performed: MM vs. ACM (primary comparison), MM vs. NMC, and ACM vs. NMC. While intervention effects were assessed via linear regression, we also report the following unadjusted analyses for descriptive purposes and in accordance with CONSORT: Differences between groups in T1 means of outcome variables, means of each outcome for each group at each timepoint, and Cohen's d. The protocol predefined use of multiple imputation to handle the missing T1 outcome data if >5% of values were missing. Multiple imputation was performed using the mice package in R using predictive mean matching with imputation of five datasets. The imputation model included randomization arm, baseline burnout, baseline mindfulness, and relevant demographic variables to address observed associations with being lost to follow‐up. Convergence was verified by examining trace plots (which showed stable chain mixing and no trends across iterations), and the distributions of imputed values were similar to those of observed data.

We performed inductive coding and quantitative content analysis on open‐ended responses. We used an alpha level of 0.05 for all statistical tests. Secondary outcomes were not included in sample size calculations and were analyzed without adjustment for multiple testing; results should be interpreted as exploratory and hypothesis‐generating.

Reporting of this trial follows CONSORT 2025 guidelines (Table S12) (Hopewell et al., 2025). No patients or members of the general public were involved in the trial design, conduct, or reporting. However, GCs led and contributed to the design, conduct, analysis, and reporting of this trial (CC, MAC, JA, JS). Study materials were piloted with multiple GCs prior to implementation. The study protocol is available upon request. The statistical analysis plan is described in the Data Analysis section above.

3. RESULTS

3.1. Participants

A total of 397 participants were randomized (132 MM, 133 ACM, 132 NMC). The median age was 31 (IQR = 10), and nearly all participants were female, White, and non‐Hispanic/Latine (Table 1). Most participants (64.5%) had some prior exposure to meditation; however, only a minority of participants (8.3%) meditated on a weekly basis or had some prior mindfulness training (11.7%; Table S3). At T0, 57% of participants had PFI scores above the threshold for burnout.

TABLE 1.

Participant characteristics.

	Total	MM	ACM	NMC
Age a	31.0 (10.0)	31.0 (8.5)	31.0 (10.0)	31.0 (10.0)
Gender b
Female	388/396 (98.0%)	127/131 (96.9%)	131/133 (98.5%)	130/132 (98.5%)
Male	8/396 (2.0%)	4/131 (3.1%)	2/133 (1.5%)	2/132 (1.5%)
Race c
White	374/397 (94.2%)	125/132 (94.7%)	131/133 (98.5%)	118/132 (89.4%)
Asian	24/397 (6.0%)	7/132 (5.3%)	2/133 (1.5%)	15/132 (11.4%)
Black or African American	2/397 (0.5%)	1/132 (0.8%)	1/133 (0.8%)	0/132 (0%)
Other	4/397 (1.0%)	4/132 (3.0%)	0/133 (0%)	0/132 (0%)
Hispanic or Latine	7/396 (1.8%)	2/132 (1.5%)	2/132 (1.5%)	3/132 (2.3%)
Year of Graduation a	2015 (8)	2014 (9)	2015 (8.25)	2014 (8)

Abbreviations: ACM, active control meditation; MM, mindfulness meditation; NMC, no‐meditation control.

^a Median (interquartile range).

^b Transgender and nonbinary options were provided but not selected; Male and Female were the options used in the surveys when they were administered. We have retained them here to ensure we are reporting the terms participants selected; however, they are not the current standard of the journal for gender, which is Man and Woman.

^c Choose all that apply.

Figure 2 presents the CONSORT flow diagram of participant progression through the trial. A quarter of the participants (24.2%) did not complete T1 outcome measures. More participants were lost to follow‐up in the ACM (32%) and MM (25%) groups than the NMC group (15%) (p = 0.04). Participants lost to follow‐up had higher baseline burnout scores (p = 0.02) and lower baseline mindfulness (subscales: acting with awareness, p = 0.005, non‐judging, p = 0.002; non‐reactivity, p = 0.03). More participants were lost to follow‐up among those who did not engage in any meditation sessions in both the MM (69.2% vs. 13.2%) and ACM (77.8% vs. 9.1%) arms compared to those who meditated at least once (p < 0.001).

Of note, the study started in 2019 and continued through 2020. For 39% of participants, the T1 survey and at least a portion of the intervention period occurred during the early months of the coronavirus pandemic. Neither T0 burnout nor T1 burnout was associated with completing the relevant survey during the pandemic.

3.2. Adherence to and experience with meditation

3.2.1. Credibility and expectancy

The Credibility Expectancy Questionnaire (CEQ) was completed by 78.9% (104/132) of MM participants and 63.9% (85/133) of ACM participants. ACM participants rated their form of meditation as less credible and held lower expectations for benefit than MM participants did for their form of meditation (credibility: p < 0.0001; expectancy: p < 0.0001).

3.2.2. Adherence

The median number of meditation sessions was 13 (IQR 30.8) for MM and 5 (IQR 16.0) for ACM (p = 0.0001). More MM participants initiated at least one meditation session compared to ACM participants (Table S4). Participants in both arms reported a high level of adherence to their assigned meditation instructions (Table S5, Figure S1). The number of meditation sessions initiated was not correlated with either T1 burnout (MM: p = 0.25; ACM: p = 0.45) or T1 mindfulness (MM: p = 0.06; ACM: p = 0.19). At T1, 29.0% (29/100) of MM participants and 22.5% (20/89) of ACM participants reported they had done meditation during the intervention period other than what was prescribed to them in the trial. However, they did very few non‐study meditation sessions (MM: median 1.0 (IQR 1.0); ACM: median 1.0 (IQR 0.25)).

3.2.3. Perceived harms and benefits

Harms or negative effects associated with meditation were similarly reported across groups (MM 12.0% vs. ACM 15.7%; p = 0.6). Among self‐reported harms and negative effects, more participants reported difficulty with the assigned meditation type in the ACM group (20.0%) than in the MM group (1.0%; p = 0.001; Table S6). This included anxiety, agitation, and the focus on thoughts being unhelpful. More MM participants (91.0%) perceived benefit from meditation during the study compared to ACM participants (69.3%; p < 0.001). More MM participants intended to continue to meditate after the trial (MM 88.0%, ACM 75.3%; p = 0.04).

3.2.4. Change in trait mindfulness

Mindfulness increased from T0 to T1 by 10.7% in MM participants and 12.0% in ACM participants (Table S7). Of note, change in mindfulness in ACM participants was not associated with self‐reported use of mindfulness techniques during meditation (p = 0.5).

3.3. Primary outcome ‐ burnout

In the prespecified primary analysis using linear regression to evaluate the intervention effect, controlling for baseline burnout and mindfulness, no intervention effect was observed for MM, compared to ACM. However, in prespecified secondary analyses, compared to NMC, both MM and ACM resulted in decreases in burnout (Table 2, Figure 3). Unadjusted mean differences between groups at T1 were 0.28 (95% CI −1.5 to 2.1) for MM vs. ACM, −4.0 (95% CI −5.9 to −2.2) for MM vs. NMC, and −3.8 (95% CI −5.7 to −1.8) for ACM vs. NMC. Effect sizes for change in burnout from T0 to T1 were large for MM (Cohen's d = −0.84, 95% CI −1.1 to −0.6) and medium for ACM (d = −0.70, 95% CI −1.0 to −0.4). Similar findings were seen for both components of burnout (Figure 3, Table S10). Mean T2 burnout scores for MM were lower than T0 (p < 0.001) but higher than T1 (p = 0.03), suggesting partial maintenance of MM benefit (Table 3). Scores for all PFI subscales at all time points, by arm, are in Table S8.

TABLE 2.

Linear regression results for intention‐to‐treat outcome analysis for burnout.

Comparison	Term	Coefficient (β)	Standard error	p‐Value
MM vs. ACM	Study arm	−0.53	0.83	0.525
	T0 burnout	0.28	0.06	<0.0001
	T0 mindfulness	0.02	0.08	0.812
ACM vs. NMC	Study arm	−3.36	0.78	<0.0001
	T0 burnout	0.49	0.06	<0.0001
	T0 mindfulness	−0.05	0.09	0.548
MM vs. NMC	Study arm	−4.05	0.77	<0.0001
	T0 burnout	0.44	0.06	<0.0001
	T0 mindfulness	−0.05	0.09	0.593

Note: Sample size: N = 397.

Abbreviations: ACM, active control meditation; MM, mindfulness meditation; NMC, no‐meditation control 2.

FIGURE 3.

Burnout and its components by study arm and timepoint. Mean burnout scores and component subscales for each study arm at T0 (baseline) and T1 (after intervention). Error bars represent ± standard deviation from the mean. Panel (a) shows burnout subscale scores from the Professional Fulfillment Index. Panel (b) shows the interpersonal disengagement component of burnout. Panel (c) shows the work exhaustion component of burnout. ACM, active control meditation; MM, mindfulness meditation; NMC, no‐meditation control. The mindfulness meditation(MM) and active control meditation (ACM) groups both showed significant reductions in burnout compared to the no‐meditation control (NMC) group, with no significant difference between MM and ACM groups. Sample sizes: MM n = 100, ACM n = 90, NMC n = 111 at T1.

TABLE 3.

Burnout at each timepoint, by study arm.

Timepoint	Arm	Mean (95% CI)	Burned out
T0	MM	16.5 (15.3–17.7)	75/131 (57.3%)
	ACM	15.2 (14.2–16.3)	60/133 (45.5%)
	NMC	14.5 (13.3–15.7)	80/132 (60.2%)
T1	MM	11.4 (10.1–12.6)	35/100 (35.0%)
	ACM	11.6 (10.3–13.0)	35/90 (38.8%)
	NMC	15.4 (14.0–16.8)	66/111 (59.4%)
T2	MM	13.0 (11.4–14.5)	37/87 (38.1%)

Note: Burnout Cutoff: A score ≥14 indicates the individual is likely experiencing burnout, as defined by scale authors.

Abbreviations: ACM, active control meditation; MM, mindfulness meditation; NMC, no‐meditation control.

3.4. Secondary outcomes

For perceived stress, compared to NMC, both MM and ACM resulted in greater reductions, with no difference between MM and ACM (Table S11, Figure S2). Unadjusted mean differences in stress scores between groups at T1 were 1.2 (95% CI ‐0.5 to 3.0, p = 0.16) for MM vs. ACM, −3.7 (95% CI −5.3 to −2.0, p < 0.001) for MM vs. NMC, and −4.9 (95% CI −6.6 to −3.2, p < 0.01) for ACM vs. NMC. Effect sizes for change in stress from T0 to T1 were large for both MM (Cohen's d = −0.88, 95% CI −1.16 to −0.60) and ACM (d = −0.90, 95% CI −1.19 to −0.60). Mean T2 stress scores for MM were comparable to T0 (p = 0.4) and higher than T1 (p < 0.001), suggesting a lack of sustained benefit. Of note, all participants completed the T2 survey during the first year of the coronavirus pandemic.

For the professional self‐care scale, compared to NMC, MM resulted in significantly greater self‐care. However, there was no difference for MM vs. ACM and for ACM vs. NMC (Table S11). The MM intervention effect was driven by increases in the cognitive awareness subscale. No other subscales showed intervention effects. T2 cognitive awareness subscale scores for MM participants were higher than T0 (p = 0.0001) and comparable to T1 (p = 0.1), suggesting that benefit was sustained. The effect size for change in cognitive awareness from T0 to T1 for MM was medium (Cohen's d = 0.70, 95% CI 0.43 to 0.98).

No differences were observed between study arms for professional fulfillment, resilience, desire to reduce time spent seeing patients, or reactive distress (Table S11). Scores for all secondary outcomes at all time points, by arm, are in Tables S8 and S9.

4. DISCUSSION

The primary aim of this study was to examine the effects of 8 weeks of MM training on self‐reported burnout, as compared to an active control that did not include mindfulness training. Both MM and ACM yielded similar reductions in burnout, which was not what we expected. The non‐superiority of MM vs. ACM may be explained by factors related to the performance of the active control condition. ACM was designed to control for nonspecific aspects of meditation—such as dedicated self‐care time, structured practice, and expectation of benefit—while excluding explicit mindfulness training (Davidson & Kaszniak, 2015; Kinser & Robins, 2013). However, mindfulness increased similarly in both MM and ACM groups. Furthermore, lower credibility and expectancy scores suggest that ACM failed to control for the expectation of benefit, a critical component of a well‐matched active control (Bishop, 2002). Given these findings, the ACM failed to serve as an appropriate active control training, preventing a valid test of our primary hypothesis regarding the specific effects of MM versus nonspecific meditation effects. In prespecified secondary analyses, compared to a passive control (NMC), MM yielded significantly increased mindfulness, decreased burnout, and decreased perceived stress.

The poor performance of ACM as an active control is consistent with broader challenges in meditation research regarding the design of appropriate control conditions and the lack of a gold standard active control for MM (Davidson & Kaszniak, 2015; Goldberg et al., 2019; Van Dam et al., 2018). The unexpected increase in mindfulness among ACM participants reveals ACM was not in fact inert. However, Figure S1 shows that despite both arms showing increased mindfulness scores, participants were indeed engaging in fundamentally different practices. It may be that ACM inadvertently included mindfulness components (Goldberg et al., 2016; Tran et al., 2022). A 2022 meta‐analysis of 146 RCTs found that self‐reported mindfulness increased in (non‐mindfulness) active controls and that change in mindfulness accounted for change in mental health outcomes (Tran et al., 2022). One possibility is that ACM instructions inadvertently promoted mindfulness through the cultivation of meta‐awareness—the awareness of one's current mental state, including where attention is focused, a key component of mindfulness (Dunne et al., 2019; Jankowski & Holas, 2014). Although designed to encourage mind‐wandering and reflection without mindfulness techniques, the ACM's periodic audio prompts to “get lost in your thoughts” may have paradoxically increased participants' awareness of their mental states. Additionally, the intentional nature of mind‐wandering in ACM may differ from spontaneous mind‐wandering. Research by Seli and colleagues found that intentional mind‐wandering is positively associated with mindfulness as measured by the FFMQ, while unintentional mind‐wandering shows negative associations (Seli et al., 2015).

Given the increase in mindfulness and decrease in burnout observed in the ACM arm, it is worth considering whether ACM may be an effective intervention for addressing GC burnout. Notably, multiple data points suggest that ACM is less acceptable than MM for at least a subset of GCs. In addition to having lower credibility and expectancy than MM, there were more participants that did not meditate during the trial, fewer meditation sessions per participant, fewer participants who perceived meditation benefited them, and fewer participants who intended to continue to meditate after the trial. Only 1.0% of MM participants reported difficulty with their assigned meditation type, compared to 20.0% of ACM participants (p = 0.001). ACM participants described experiencing anxiety and agitation, and finding the focus on thoughts unhelpful—experiences that are both harmful and that could limit long‐term adherence and effectiveness. These data suggest that the form of meditation used in the ACM may not be a viable option for many GCs. Our overall attrition rate of 24% aligns with the pooled mean of 24.7% (95% CI: 20.6%–29.3%) from a meta‐analysis of 70 mindfulness app RCTs (Linardon, 2023).

Although our primary hypothesis could not be effectively tested, prespecified secondary analyses provide evidence for MM's positive impact on professional well‐being in clinical GCs. Given the known challenges in designing active controls for MM, we included a passive control arm (NMC) to serve as an alternate control in secondary analyses in case ACM did not perform effectively as an inert active control. Compared to NMC, MM reduced both burnout and stress with large effect sizes. At baseline, 57.3% of MM participants were burned out, dropping to 35.0% at T1. In contrast, the proportion of GCs experiencing burnout remained stable in the NMC arm. These findings need to be interpreted with caution, given that they arise from secondary analyses, and because it has been well demonstrated that comparisons to a passive control can both overestimate effect sizes and identify effectiveness erroneously (Davidson & Kaszniak, 2015; Goldberg et al., 2022; Kinser & Robins, 2013). However, recent active‐controlled trials support the validity of the benefits of MM that we observed. Several randomized studies have demonstrated that mindfulness interventions outperform a vareity of active comparators for clinician burnout and stress, including dedicated break time (Ireland et al., 2017), stress education (Cascales‐Pérez et al., 2021), and protected time for self‐directed activities (West et al., 2014). A 2022 systematic review found mindfulness superior to active controls for stress in healthcare providers (Cascales‐Pérez et al., 2021; Goldberg et al., 2022). This convergent evidence from active‐controlled studies suggests our observed intervention effect for MM versus NMC is less likely to be spurious and more likely to represent true benefit for clinical GCs.

We did not see a dose‐dependent relationship between the number of meditation sessions and T1 burnout or mindfulness, which is consistent with prior studies (Strohmaier, 2021). However, recent studies suggest a longer time frame may be needed to detect a relationship between meditation dose and benefit (Bowles et al., 2022; Bowles & Van Dam, 2025). The fact that participants who did not complete the study had higher baseline burnout suggests that self‐initiated app‐based MM may not be accessible to GCs who are more severely burned out. They may instead need more structured support such as MM facilitated in the work setting, peer support, or systems‐level interventions, which are discussed further below (Klatt et al., 2015; Panagioti et al., 2017; West et al., 2014).

Among secondary outcomes, MM reduced stress and increased scores on the cognitive awareness subscale of the Professional Self Care Scale, but it did not impact professional fulfillment, resilience, reactive distress, or desire to reduce clinical time. MM may not be a good fit for these outcomes or the trial may have been underpowered to detect an intervention effect for them.

The 57% baseline burnout rate, measured during the COVID‐19 pandemic for many participants, is consistent with prior studies on GC burnout, which estimate point prevalence at 35%–87%, depending on instrument and sample (Gostic et al., 2024; Injeyan et al., 2011; Johnstone et al., 2016; Lyon et al., 2025; National Society of Genetic Counselors, 2024). The only other study using the PFI in GCs reported 17% burnout, though in a small subspecialty sample in Australia (Yeates et al., 2025). Of note, most studies on GC burnout do not report the point prevalence of burnout in their sample (Allsbrook et al., 2016; Bernhardt et al., 2009; Lee et al., 2015; Silver et al., 2018; Stanley et al., 2026; Udipi et al., 2008; Wadman et al., 2022).

GCs often have access to MM resources provided by their employers. This can include classes such as Mindfulness‐Based Stress Reduction, apps such as Headspace and Calm, or brief recurring guided MM sessions in the workplace. Studies across a variety of work settings have found investment in mindfulness programs leads to increased productivity and reduced healthcare usage by employees, which can help make the business case for employers to provide such programs (Huberty et al., 2022; Klatt et al., 2016). A recent qualitative meta‐synthesis examined factors that affect successful implementation of MM for healthcare providers (Knudsen et al., 2024). The authors provide recommendations including, but not limited to, leadership support, providing protected time during work and dedicated space, making participation optional, offering group‐based trainings, providing training in brief exercises that can be done during clinical work, and providing in‐person and online options. In terms of feasibility, it is notable that our results and those from other studies suggest that the benefits of MM can be realized with short (~10 min) meditation sessions done a few times a week (Bostock et al., 2019; Champion et al., 2018; Fincham et al., 2023). In addition, self‐directed app‐based meditation, which can be more accessible to busy clinicians than live classes, has been shown to be effective for a variety of outcomes, including stress, burnout, and other aspects of professional well‐being (Bostock et al., 2019; Champion et al., 2018; Gál et al., 2021; Flett et al., 2020; Radin et al., 2025; Taylor et al., 2022; Zawadzki et al., 2025).

Even when time and resources for MM are provided by an employer, MM is still ultimately an individual‐level intervention. Focusing solely on individual‐level interventions is insufficient to effectively address clinician burnout and is also unethical (De Simone et al., 2021; National Academies of Sciences, Engineering, and Medicine, 2019; Pijpker et al., 2019; Spilg, 2024). This is evident in our own data, with a third of MM participants still experiencing burnout at T1. As articulated in the National Academy of Medicine's report on clinician burnout (National Academies of Sciences, Engineering, and Medicine, 2019), burnout is fundamentally a systems‐level problem requiring systems‐level solutions. Meta‐analyses and systematic reviews have found that work‐system interventions have larger effects on clinician burnout than individual‐level interventions (De Simone et al., 2021; Panagioti et al., 2017) and that combinations of the two are most effective (Pijpker et al., 2019). Our previous work identified multiple work‐system factors contributing to GC burnout, including insufficient administrative support, lack of autonomy, and not feeling valued by non‐GC colleagues (Caleshu et al., 2022). The broader literature on clinician burnout suggests additional work‐system interventions, including a culture where clinicians can speak up and have meaningful involvement in organizational decisions that affect them, improvements in the usability of technology, ensuring workload aligns with resources, peer‐support programs that enhance sense of community and meaning at work, reduction in documentation burden, and increased automation of low‐in‐scope tasks (De Simone et al., 2021; National Academies of Sciences, Engineering, and Medicine, 2019; Pijpker et al., 2019; Thomas Craig et al., 2021).

4.1. Limitations

Several limitations warrant consideration when interpreting our findings. We did not collect reasons participants did not complete the study or did not progress from eligibility screen to completing the T0 survey. It is reasonable to suspect the results are impacted by self‐selection bias. Since participants who were lost to follow‐up had higher burnout and lower mindfulness at baseline, outcome data was unlikely to be missing completely at random, which is a key assumption of multiple imputation. However, we included these baseline variables in the imputation model, which can mitigate bias. Convergence was verified by examining trace plots (which showed stable chain mixing and no trends across iterations), and the distributions of imputed values were similar to those of observed data, suggesting multiple imputation was appropriate in this context. Consistent with the ITT approach, no participants were excluded based on adherence, and imperfect adherence likely attenuates the observed treatment effects. Our measurement of meditation adherence may overestimate actual practice, as the modalities available for collecting this data could only measure session initiation, not completion. Additionally, 39% of participants completed their post‐intervention assessment during the early months of the COVID‐19 pandemic. While we found no association between pandemic timing and burnout, the unprecedented stressors of this period may have influenced results in unmeasured and/or unmeasurable ways. Since secondary outcomes were not included in sample size calculations and were analyzed without adjustment for multiple testing, results should be interpreted as exploratory and hypothesis‐generating. While we aimed to blind both participants and analysts, this was likely not fully effective. The daily balance subscale of the Professional Self Care Scale had low internal consistency in our sample. Our participants were nearly all White, cis‐gender, and female. This homogeneity limits generalizability to GCs of other identities, who may have different experiences of both burnout and meditation. Future research must prioritize recruiting diverse samples to understand whether meditation's benefits extend equally across all GCs.

4.2. Future directions

In addition to the need for studies on GCs of varying identities, research on MM for GC students and GCs not involved in direct patient care is also needed. Further work is needed to develop and validate active controls that successfully control for nonspecific aspects of meditation without inducing mindfulness. This methodological work is essential for determining the specific versus nonspecific benefits of MM (Goldberg et al., 2022). Studies on clinician MM have found improvements in care quality and patient outcomes; investigating this within the genetic counseling field would be valuable (Braun et al., 2019). Lastly, studies are needed on a range of work system‐level interventions to improve GCs' burnout and other aspects of professional well‐being (National Academies of Sciences, Engineering, and Medicine, 2019).

5. CONCLUSION

Our primary hypothesis could not be effectively evaluated due to methodological challenges with the active control. Nonetheless, this randomized controlled trial provides evidence that meditation can significantly reduce burnout and stress among GCs. Prespecified secondary analyses showed that MM improved burnout and stress among clinical GCs when compared to the passive control. These findings, when considered within the context of the mounting evidence for the positive impact of clinician meditation on burnout, point to the need to consider MM as an individual‐level intervention to mitigate burnout in the genetic counseling field. However, meditation must be positioned as one component of a comprehensive approach to GC professional well‐being, complementing rather than replacing necessary systems‐level reforms.

AUTHOR CONTRIBUTIONS

Colleen Caleshu: Conceptualization; data curation; formal analysis; funding acquisition; investigation; methodology; project administration; visualization; writing – original draft; writing – review and editing. MaryAnn Campion: Conceptualization; methodology; supervision. Jehannine (J9) Austin: Conceptualization; methodology; supervision. Philippe Goldin: Conceptualization; methodology. Julia Silver: Conceptualization. Aad Tibben: Conceptualization; methodology; supervision.

CONFLICT OF INTEREST STATEMENT

Headspace provided in‐kind six‐month app subscriptions to all study participants. Headspace had no role in the study design, data collection, analysis, interpretation, or writing of the manuscript. Authors disclose the following conflicts of interest: employment and ownership interest in Genome Medical (CC); none to declare (MAC); none to declare (JS); none to declare (AT); none to declare (PG). J9 Austin is editor in chief of the Journal of Genetic Counseling and co‐author of this article. They were excluded from editorial decision‐making related to the acceptance of this article for publication in the journal.

ETHICS STATEMENT

This study was approved by the Stanford University Institutional Review Board. The study conformed to the principles of the Declaration of Helsinki and was registered on ClinicalTrials.gov on 2018‐10‐25 (NCT03723018).

Human studies and informed consent: All participants provided informed consent prior to their inclusion in the study. Participant anonymity has been preserved and all identifying information has been excluded from the manuscript.

Animal studies: No animal studies were done.

POSITIONALITY STATEMENT

Our research team consists of GCs in research, clinical, leadership, and educational roles, as well as research and clinical psychologists. The team has expertise in genetic counseling, mindfulness meditation, clinical and research psychology, and randomized controlled trials. Some team members have a mindfulness meditation practice while others do not. As GCs conducting research within our own profession, we bring insider perspectives on the stresses and professional well‐being challenges facing GCs, while also acknowledging that our shared professional identity and varied personal experiences with burnout may influence our research approach and interpretation of findings. Across the team, we hold a range of social identities, some of which are majority identities in genetic counseling (e.g., White, cis‐gender female, heterosexual, nondisabled) and some of which are marginalized in the profession (e.g., queer, agender, disabled). We are geographically located in the United States, the Netherlands, and Canada. We share a methodological commitment to rigorous research.

Supporting information

Figures S1–S2

Tables S1–S12

DATA AVAILABILITY STATEMENT

Data are available from the authors upon reasonable request.