Neurobehavioral Effects of Prenatal Mindfulness Training on Maternal Presence and Compassionate Love

This study was initiated at the PI's previous institution, the University of Illinois at Urbana-Champaign (UIUC), under the approval of the UIUC IRB (protocol 19461). Data collection for participants enrolled at UIUC was concluded in December 2021, and study recruitment and data collection for the remainder of the total sample is to commence at the PI's current institution, the Pennsylvania State University, in January 2022.

Study Procedures

At 20-28 weeks of pregnancy, participants recruited into the study provide informed consent and complete questionnaire measures either in person at the research lab or remotely via phone or Zoom (baseline assessment). Those randomized to MBCP then complete the 9-week class with their partner or other support person, and those assigned to treatment as usual complete an in-person or online birthing class of their choice from a list provided. After taking part in the birthing class, at approximately 37 weeks of pregnancy, women again complete questionnaires remotely (post-class assessment). At 3-4 months postpartum, women complete a final set of questionnaires and take part in a videorecording session with their infant, followed by a brain scanning session at the university-affiliated neuroimaging center.

Consent: Participants who are deemed eligible for the study are contacted to schedule a consent session (in-person or remote). A trained researcher discusses the study protocol and key details of informed consent. Following consent, the researcher discusses what the MRI session will be like and the participant fills out an online safety screener to identify any MRI contraindications. Participants are also asked about availability for birthing classes and to rank the available birthing classes according to their preference.

Birthing Class Assignment + Pre-Class Questionnaire (T1): Between the participant's 20th and 28th week of pregnancy, a study researcher randomly assigns the participant to the treatment (MBCP) or control (birthing class of the participant's choice) condition. The researcher contacts the participant to schedule a birthing class assignment session (T1; by phone or on Zoom). During the birthing class assignment, participants are told whether they have been assigned to the MBCP course or are able to choose a birthing class from the provided list. They are asked whether they have any questions about their assignment and provided with instructions on how to sign up for their class/seek reimbursement, as needed. Following the session, participants are sent a personalized link to the first questionnaire by email for them to complete.

Child-bearers randomized to MBCP complete the 9-week class with their partner or other support person. The class is taught by experienced midwives or social workers who have completed MBCP teacher training requirements, which include ongoing cultivation of a personal mindfulness practice as well as education and supervision specific to the MBCP program. Participants in this condition have the option of participating in one of several online MBCP classes available that accommodates their pregnancy timeline. Each three-hour MBCP class comprises a mix of instructor-guided mindful meditation practice (e.g., engaging in a body scan, breath meditation; prenatal yoga practice) and inquiry, in which instructors lead discussion of participants' experiences of the practice and how it applies to their lives. Classes also involve psychoeducation on the experience of and ways to engage with pregnancy, birth, and parenting from a mindfulness perspective. Participants are encouraged to engage in at-home practice: at least thirty minutes of formal and/or informal mindfulness practices modeled in class six days a week, and to read Mindful Birthing, a companion book written by MBCP developer Nancy Bardacke.

Child-bearers randomized to the treatment as usual group participate (also with a partner) in a childbirth class of their choice from a list provided. To maintain ecological validity of this treatment as usual condition, using an established comparative-effectiveness/pragmatic trial approach, no attempt is made to control the length or contact hours of the class, but a range of classes with varying time commitments are offered, and this information is collected at the post-class assessment and considered as a control variable in analyses. No data collection occurs during any of the classes.

Post-Class Assessment (T2): At approximately 37 weeks of pregnancy (following birthing class completion), participants are emailed a personalized link to a post-class questionnaire for them to complete at home.

Post-Birth Assessment (T3.1): Approximately three months after their infant is born, participants are emailed a personalized link to a post-birth questionnaire for them to complete at home.

Videorecording Visit (T3.2): If the participant indicates they are comfortable with in-person sessions, a researcher schedules a time to visit their home to videorecord the participant interacting with their infant. At the start, mothers are instructed to play with their infant as they normally would for ten minutes (freeplay). Mothers can feel free to play games, sing, use toys, or read books during freeplay. Following freeplay, mother are asked to engage in a structured peekaboo interaction with their infant, which is designed to elicit positive infant affect. During peekaboo, mothers are asked to hold their hands over their eyes and call their baby's name, then reveal their eyes and say "peekaboo." They repeat the above for two and a half minutes. This task may be repeated up to two more times, if needed.

Then mothers are asked to engage in an arm-restraint task (holding the infant's arms gently by their side to restrict movement while adopting a neutral, still expression), a task designed to elicit distress in infants at this age. This task lasts up to two and a half minutes, although the researcher may stop this task early if the baby is quite distressed. Two 12-second video segments each of infant positive and negative emotion expression (4 total) are selected for presentation in the scanner. Positive/negative video segments of an infant not involved in the study are also collected using similar procedures to create the "other infant" stimuli. Non-infant comparison stimuli are created by editing the unfamiliar infant videos using Matlab Image Scramble and time-domain audio scrambling with shufflewins to create scrambled videos with similar audiovisual characteristics but no recognizable infant features.

Brain Scanning Session (T3.3): After the home visit, participants are scheduled for an MRI session at the university-affiliated neuroimaging center. During this session, participants complete five study scans: two resting state scans (6 minutes, 12 seconds each), a structural scan (5 minutes, 21 seconds), and two functional scans where they view the video clips of their own and another unknown infant, as well as non-infant stimulus blocks (8 minutes, 12 seconds each). The total duration of the scan is 45 minutes.

Functional Scanning - Regional brain activity is assessed with blood oxygen level dependent echoplanar images (BOLD-EPI), T2*-weighted gradient echo sequence, TR = 2 s, TE = 25 ms, flip angle = 90 deg, 38 slices of thickness 3.0 mm, with a slice gap of 0.3 mm, 92 x 92 voxel matrix, FOV = 230 mm.

Structural Scanning - Functional data are mapped onto high-resolution T1-weighted structural images using 3-DMagnetization-Prepared RApid Gradient Echo (MPRAGE) pulse sequence, TR = 23000 ms, TE = 2.32 ms, TI = 900 ms, flip angle = 8 deg, 192 sagittal slices of thickness 0.9 mm, 256 x 256 matrix, FOV = 240 mm.

Mothers view the videos via a mirror positioned over their eyes and hear accompanying sounds via headphones (sound check prior to scanning to ensure audibility). Functional runs present each stimulus block 6 times (3 presentations each of unique positive and negative own infant/other infant/scrambled video segments), in counterbalanced order. Video stimuli accompanied by sound were chosen based on previous research showing stronger neural responses to dynamic (as opposed to still) emotion faces, and the investigators' own work showing mindfulness-related effects on mothers' neural response to similar infant videos. Stimulus presentation protocols were informed by previous maternal neuroimaging research. Following scanning, mothers are asked through both open-ended and forced-choice questions what they were doing during the task (distraction or suppression of emotion, mindfulness practice, etc.). They also rate their own and their infant's emotional valence and intensity during each video segment. Reported activities and emotion ratings will be considered as possible covariates in fMRI analysis.

Study Data Analysis

Baseline characteristics of the sample, stratified by experimental group, will be examined to ensure even distribution of key variables. Completers and drop-outs will be compared on baseline variables, and those showing group differences will be considered as covariates in tests of hypotheses. Primary analyses will be conducted with the full sample according to intent-to-treat principles under CONSORT guidelines.

Aim 1: MBCP > self-reported maternal presence and compassionate love. To test Hypothesis 1 that MBCP mothers will show higher levels of these qualities, both following the course and later with their infants, hierarchical linear modeling (HLM) will be used to examine the main effect of experimental group (expressed as a dummy-coded variable indicating MBCP participation) on slopes of self-report measures of presence and compassionate love across all three assessments, as well as post-intervention levels. These models will demonstrate whether MBCP results in increased levels of these qualities, both more generally and specifically with the mother's own infant.

Aim 2: MBCP > neural response to one's infant. To test Hypothesis 2 that MBCP mothers will show increased neural activation to their infant's emotion cues in brain regions supporting the above qualities-i.e., ACC, dlPFC, insula, vPFC, and striatum-group comparisons will be conducted using the FSL program.

At the intrasubject level, multiple regression with OLS and adjustment for autocorrelation will test contrasts of stimulus-specific regressors. Boxcar models reflecting onset-offset of each infant stimulus will be convolved with an optimal basis set for HRF, generated using FLOBS. Results will be averaged across the 2 runs using fixed-effects analysis. Based on both theorized MBCP mechanisms of action and previous research on mindfulness-related maternal brain effects, two aspects of maternal neural response will be examined as outcomes: (1) more general responsiveness to their infant, indexed by contrasts of activation to own > other infant and own-infant > non-infant videos, and (2) valenced responses to their infant, indexed by contrasts of positive > negative own-infant videos.

At the group level, MBCP-related differences in each of these responses will be tested via MBCP > Control and Control > MBCP contrasts with mixed-effects analysis (FLAME) in the whole brain. Cluster threshold correction with Z > 2.6 and FDR < .05 will be used to define regions of significantly different activation, and correspondence with brain atlas-defined anatomical regions will determine whether hypothesized sites of elevated responsiveness are supported.

Aim 3: Identify MBCP moderators. To test Hypothesis 3 that mothers who begin the class with more risk characteristics and/or who receive a higher dosage of mindfulness practice will show greater benefits of MBCP, interaction terms involving moderator variables will be added to the above models. Based on the hypotheses outlined above, investigators expect significant positive interactions between MBCP participation and single relationship status, distress, and practice dosage.