Abstract
Introduction The main objective of the study will be to evaluate the effects of two widely used standardized mindfulness-based programs [Mindfulness-Based Stress Reduction (MBSR) and Compassion Cultivation Training (CCT)], on epigenetic, neurobiological, psychological, and physiological variables.
Methods The programs will be offered in an intensive retreat format in a general population sample of healthy volunteer adults. During a 7-day retreat, participants will receive MBSR and CCT in a crossover design where participants complete both programs in random order. After finishing their first 3-day training with one of the two programs, participants will be assigned to the second 3-day training with the second program.
The effects of the MBSR and CCT programs, and their combination, will be measured by epigenetic changes (i.e., DNA methylation biomarkers), neurobiological and psychophysiological measures (i.e., EEG resting state, EKG, respiration patterns, and diurnal cortisol slopes), self-report questionnaires belonging to different psychological domains (i.e., mindfulness, compassion, well-being, distress, and general functioning), and stress tasks (i.e., an Arithmetic Stress Test and the retrieval of negative autobiographical memories). These measures will be collected from both groups on the mornings of day 1 (pre-program), day 4 (after finishing the first program and before beginning the second program), and day 7 (post-second program). We will conduct a 3-month follow-up using only the set of self-report measures.
Discussion This study aims to shed light on the neurobiological and psychological mechanisms linked to meditation and compassion in the general population. The protocol was registered at clinicaltrials.gov (Identifier: NCT05516355; August 23, 2022).
1. Introduction
In recent years there has been an outburst of scientific studies on the positive effects of mindfulness-based interventions (Mindfulness-Based Interventions, MBIs) on physical and mental health [1]. Two of the most researched programs related to the development of mindfulness and compassion are, respectively, Mindfulness-Based Stress Reduction (MBSR) [2], focused on attentional training and a non-judgmental attitude, and Compassion Cultivation Training (CTT) [3,4], more focused on the structured training of a compassionate and self-compassionate attitude. These are highly structured programs, of short duration (i.e., typically delivered in 8 weeks/16 hours format), that have shown positive effects, both for general and clinical populations, as reflected in many meta-analyses and systematic reviews [5–9], including programs using briefer formats (e.g., single-session to 2-week multi-session formats) [10].
The fundamental objective of this study is to explore the potential changes in epigenetic variables, psychological distress, well-being, brain activity, cortisol levels, and reactivity to stress, that can be observed following MBIs. The way to address this question is through an intensive intervention in a retreat format, a procedure that allows control of important variables (e.g., food, environmental conditions, etc.), and administering all the components of the standard MBSR and CCT programs in their entirety. There is already meta-analytic evidence on the positive effects of these intensive multi-day programs in healthy individuals [11]. Although there is some evidence that attentional changes might be involved in the psychological changes observed after a meditation retreat [12], there is a need to examine the mechanisms of action through which meditation practice produces its effects [13]. Most theoretical models published to date emphasize the central role of attention regulation, which is thought to underpin emotional and cognitive flexibility [14], which in turn enhances emotion regulation processes [15], and the ability to maintain non-judgmental awareness of thoughts, feelings, and experiences. Theoretical models also emphasize the role of compassion and self-compassion [16] in promoting change due to MBI practices [17–21]. However, the extant evidence comparing the self-reported psychological effects of MBSR and CCT programs indicates that both programs have similar effects [22,23], although reached through different mechanisms [24].
Regarding the biological effects of MBIs, there is consistent meta-analytic evidence showing that they produce significant physiological changes in parameters like a reduction in oxygen consumption, blood pressure, and heart rate [25], which are contrary to those that occur during the stress response [26]. These changes have also been observed in three-day intensive meditation retreats [27]. Similarly, although the evidence for brain volumetric change is uncertain [28], there is data on changes in frontoparietal and default network connectivity during resting state tests [29,30]. There is also promising evidence that intensive intervention practices, in a retreat format, are associated with significant changes in gene expression that might be potentially regulated by epigenetic mechanisms. For example, an 8-day silent guided meditation retreat can modify gene expression, in particular genes related to the immune response [31]. However, the brain mechanisms and molecular activity are not yet well understood. Along the lines of some previous research on detectable epigenetic changes in brief intensive meditation interventions aimed at reducing stress [32–36], we intend to test whether epigenetic changes are one of the mediating mechanisms in the observable psychological changes following an intervention. We will also analyze whether changes in brain connectivity (measured with EEG in resting state) and in the diurnal cortisol rhythms occur throughout the programs.
Although research on the comparative effects of MBSR and CCT has shown the existence of some particular mediators [24], the effects of a sequential intervention of both types of training on the same individuals are not known. This sequential approach is important to evaluate whether there are additive effects of the programs and whether the order of the intervention (one more focused on attentional aspects and the other more on socio-emotional factors) can facilitate psychological and neurobiological changes. Thus, we aim to analyze the existence of differential psychobiological changes in two types of intensive psychological programs (MBSR and CCT), and whether the sequence of the programs (using a cross-over design), does significantly affect the results.
As, to the best of our knowledge, this is the first study using a sequential design with two intensive meditation modalities, it is difficult to formulate specific hypotheses regarding the separate and combined effects of the MBSR and CCT programs on the selected outcomes. Thus, considering the two MBIs jointly, we expect that both programs will significantly: a) modify the DNA methylation and expression levels of specific genes and a reduction of the DNA methylation-based biological clock; b) reduce participants’ levels of psychological distress while increasing their levels of psychological functioning (i.e., mindfulness, compassion, emotion regulation, and well-being); c) change the EEG spectral profile of the resting state brain activity (i.e., changes in the balance between Alpha (8-14 Hz) and Theta (4-8 Hz) frequency bands and, regarding the organization of the brain functional networks, a modulation of the difference between the activation of the default mode network and the one corresponding to the salience network); d) alter awakening and bedtime cortisol levels leading to a steeper Diurnal Cortisol Slope (DCS); e) improve the efficacy of mood regulation following a negative mood induction (using an autobiographical memory task); f) reduce participants’ reactivity in an objective experimental stress test (i.e., Arithmetic Stress Test). In addition, given the multidisciplinary nature of the present project, we intend to provide a comprehensive view of potential changes in various areas of biological and psychological functioning and their interrelationships.
In general, we expect an overall pattern of results in which changes will be significantly higher for the combined effects of both MBIs than for each program separately. Moreover, based on previous literature comparing standard MBSR and CCT programs [24], we expect no major differences in primary outcomes between both programs, although the mechanisms of change may be different; in general, changes in MBSR will be mediated by cognitive variables (e.g., mindfulness) whereas changes in CCT will be mediated by socio-emotional variables (e.g., compassion). We also expect that the self-reported psychological changes associated with the MBIs will remain 3 months after the programs.
2. Methods
2.1 Study design overview
Participants will be randomly assigned to two groups starting with MBSR or CCT. A stratified randomization procedure, using the Research Randomizer Program (https://www.randomizer.org), performed by the principal investigator (CV), will be used to maintain the same percentage of male and female participants in each group. Participants from both groups will take part in a 7-day retreat, including training in both meditation-based practices. Using a cross-over design, the participants will undergo their second training during the last three days of the retreat. Biological and psychological measures will be collected from both groups in the mornings of day 1 (before beginning the programs), day 4 (after the first program and before the beginning of the second program), and day 7 (at the end of the program). Also, before going to bed, participants will be given a short questionnaire with eight questions about their general functioning during the day and their satisfaction with the progress of the program. Furthermore, a 3-month follow-up assessment will be conducted in both groups only for psychological questionnaires. (See figure 1).
2.2 Participants: Recruitment and selection
Individuals from the general population will be recruited through the Nirakara-Lab website, a university-associated research center specializing in Mindfulness and Compassion-Based programs. Participants will be invited to join the study on the university’s official website offering the MBSR and CCT programs (www.nirakara.org). To control for potential self-selection biases, which is a common shortcoming of MBIs research [37–39], there will be statistical controls on key issues such as years and type of previous meditation practice, if any, sociodemographic information, and mental health status. Individuals interested in participating in the retreat will be interviewed by members of the research team (RZ and CA) to screen their suitability for the study. Inclusion criteria will be being a healthy adult between 25 and 65 years old who is interested in meditation and compassion techniques to regulate their mood and stress. Exclusion criteria include having a current or past self-reported diagnosable serious or disabling mental disorder (in particular, PTSD, major depression, psychotic disorders, and/ or use of alcohol or drugs disorders) or having a current or past (i.e., less than 5 years ago) medical or physical condition that may affect the immune system (i.e., autoimmune disease, chronic severe infections, HIV, cancer), pregnancy, serious chronic illnesses, as well as the consumption of psychotropic drugs. Individuals with current habits (i.e., smoking, alcoholism, substance abuse) or taking specific medications (e.g., corticoids) that might affect the immune system are also excluded. Also, individuals planning to travel from a different time zone or long-travel times will be excluded as that might affect the immune system.
The screening procedure will be conducted as follows: 1) First contact will be made through an online form expressing interest and commitment to participating in the study, where participants will be provided with some information on the retreat (date, location, and program characteristics) and will complete information on demographics, physical and psychological health history, and meditation experience. A waiting list will be formed for participants who did not sign up in time and wished to be considered in case a place became available. An expected final sample of 52 people will take part in the retreat being randomised to the MBSR training or CCT training for the first part of the retreat. All participants will complete the first arm of the study and then will be allocated to the opposite condition for the second study arm (following a cross-over design).
Participants will pay the cost of the retreat (i.e., meals and accommodation only) and will not receive any economic compensation for their contribution to the study. Participation requires informed consent, provided by the principal investigator (PI) or authorized full members of the research team, in which all the conditions of the study are reported, following all the stipulations of the WMA Declaration of Helsinki for human studies. The models used in the study will be available upon request. In case of any adverse effect, which is very unlikely in this type of interventions, participants may leave the study at any time without having to offer any explanation and without any kind of penalty. Furthermore, the team will facilitate any kind of help requested by participants during the retreat. The collection of biological information is innocuous, and blood and saliva samples will be collected by nursing professionals.
The study has been approved by the University Ethics Committee (Ref. 22/449-E, July 2022) and has been pre-registered (clinicaltrials.gov, Identifier: #NCT (hidden for review). Any modification to the protocol will be reported both to the Ethics Committee and the trial registration site. Participants will be informed in advance of the results of the study to be published in scientific journals.
2.3 Sample size
According to G*Power software, we need 21 participants per group to obtain a power of .95 to detect a mean effect size of .37 (based on a meta-analysis on mindfulness retreats, [11]), with an alpha standard error probability of .05. In anticipation that there may be last-minute dropouts in pre-registrants or incomplete data in the sample collection or the analysis of the planned psychological or biological data, we have increased the sample by an additional 20% to 26 participants for each condition (52 in total). This sample size further corresponds to the average size of standard meditation retreats [11] and in studies of brain activity in meditators [29,30].
2.4 Components and structure of the programs
Each program will comprise standard MBSR and CCT protocols in an intensive format to fit a retreat mode. Throughout the retreat, participants will be in silence, not talking to each other or using electronic devices. Each program will last 3 days in total, and all participants will go through both programs. On each retreat day, participants will spend a total of 4.25 hours of sitting meditation, 2 hours of walking meditation, 2 hours of explanations and discussing with the instructors, doubts or problems related to the practices and 1 hour of mindful body movements (yoga/qigong). In total, participants will have 9 hours of practice each day. (A description with the contents of the programs is shown in Table 1).
2.4.1 Mindfulness-Based Stress Reduction (MBSR)
The MBSR is an 8-week standardized program [40] that aims to cultivate awareness by paying attention to the present moment with acceptance and a non-judgmental disposition. Training is delivered by two certified instructors at the University of Massachusetts Centre for Mindfulness (https://www.umassmed.edu/cfm/) and Mindfulness Center at Brown University (https://www.brown.edu/public-health/mindfulness/home).
2.4.2 Compassion Cultivation Training (CCT)
The CCT is an 8-week standardized program [3,41] aimed at cultivating compassion and empathy toward oneself and others, consisting of daily formal and informal practices. Training is delivered by two certified instructors at the Compassion Institute and Nirakara-Lab.
2.5 Study outcomes
(See Table 2 for all measures and time points).
2.5.1 Epigenetic changes
Samples of peripheral blood will be obtained and processed in situ, during the retreat. Participants’ blood samples will be collected in two standard tubes (10ml each) that contain EDTA as an anticoagulant. Blood samples will be processed immediately using the standardized Ficoll gradient centrifugation protocol to isolate mononuclear cells. These cells are cryopreserved directly (with 10% DMSO culture medium) and stored appropriately (in a -80 C freezer for short periods and liquid nitrogen for longer storage). Nucleic acids will be extracted in the laboratory of the Biomedical Epigenetics group, IDIBAPS (Barcelona), using DNA and RNA extraction kits. Once the nucleic acids are obtained, they will be stored in suitable freezers (−20 or -80°C) until their use for molecular profiling techniques, such as DNA methylation arrays (Infinium Methylation EPIC Bead Chip from Illumina, herein called “EPIC methylation arrays”), and gene expression by massive RNA sequencing (RNA-seq).
2.5.2 Psychological measures
a) Self-report measures
These measures, commonly used in previous MBI studies [22], cover selected psychological distress and functioning constructs (i.e., stress, anxiety, depression, mindfulness, compassion, well-being, and emotion regulation) and all will be administered via Qualtrics with individualized links:
Psychological distress. Depression Anxiety Stress Scales (DASS-21) [42].
Psychological well-being. Satisfaction with Life Scale (SWLS) [43].
Mindfulness. State Mindfulness Scale (SMS) [44].
Anxiety. State-Trait Anxiety Inventory (STAI-20) (Spielberger et al., 1983).
Somatic Symptoms. Patient Health Questionnaire (PHQ-15) [45].
Sleep difficulties. DSM-5 sleep difficulties [46].
Compassion. State self-compassion State Self-compassion Scale (SSCS) [47].
Fear of Compassion. The Yourself subscale from the Fear of Compassion Scale [48].
Emotion Regulation. State Difficulties in Emotion Regulation (S-DERS-21) [49].
Current mood. Positive and Negative Affect Schedules (PANAS) [50].
Daily assessments of psychological functioning. A 7-item scale to assess participants’ average state of mindfulness, mind wandering, compassion to others, self-compassion, well-being, the richness of life, and utility of the contents of the program every day.
Affect in the last 3 days. Hedonic and Arousal Affect Scale (HAAS) [51].
Meditation Adverse Effects. Meditation-Related Adverse Effects Scale – Mindfulness-Based Program (MRAESMBP) [52].
Program satisfaction. Client Satisfaction Questionnaire (CSQ-8) [53].
b) Psychological stress tasks
b.1) Stress Reactivity
Arithmetic Stress Test: a standardized laboratory stress induction procedure, that is part of the Trier Social Stress Test [54,55], will be used to assess participants’ stress reactivity. Individuals are asked to repeatedly subtract a given 2-digit number (i.e., 13) from a 4-digit number (e.g., 1022). These numbers were changed for each of the three assessment days. Participants are asked to write their responses on a personal computer as quickly as possible. Time is limited to 3 minutes to increase the stressful nature of the task. Pre-post stress level changes in mood will be assessed with the PANAS and STAI.
b.2) Emotion regulation
A 10-min retrieval of autobiographical negative memories. Participants will receive a link through Qualtrics for a baseline assessment (three days before the retreat) where they are asked to think about three negative autobiographical memories of events for which they felt responsible and made them feel negative emotions such as shame, guilt, and sadness. Participants will be asked to assign personal keywords (e.g., ‘traffic accident’), for each of the three memories, synthesizing the event. Each of the three negative events will be randomly assigned to the three respective assessment days so that participants were given the corresponding “keywords” when they begin the retrieval task.
Participants are given 10 minutes to write down the memory as specifically as possible. Pre-post changes in their sadness, happiness, anxiety, serenity, shame, and guilt affect states will be assessed using a 0 (nothing) to 100 (very much) visual analogue scale. These two laboratory psychological tasks will be completed by each individual while alone in a quiet room, using a portable computer.
2.5.3 Neurobiological laboratory measures
a) Brain electrical activity. Recording of EEG activity in resting state (alternating open and closed eyes, 6 min each condition) with four EE-225 64chEEG+24chBIP 16kHz devices with a laptop and eego EEG recording software. The sampling rate will be set at 1000 Hz and the reference is 7Z of the equidistant layout.
b) Respiratory activity. Respiration rate, which is related to attention training and psychological well-being [56], will be monitored by a respiration chest-band, during the EEG assessment session and will be registered using a piezoelectric sensor.
c) Cardiac activity EKG activity will be monitored with dedicated electrodes during the EEG assessment session.
These recordings will be completed by everyone in laboratory spaces dedicated to the EEG assessments.
2.5.4 Salivary sampling and cortisol measurement
Participants will be given Salivette® cortisol saliva sample collection tubes (Sarstedt, Germany) that contain a sterile cotton swab for sample absorption along with detailed verbal and written instructions concerning sample collection. Diurnal Cortisol Slope and cortisol levels at bedtime and awakening will be measured at three different time points of the retreat (see Figure 1). Then, the saliva samples will be stored in the refrigerator until they are delivered to the laboratory after a few hours. Once in the laboratory, the samples will be centrifuged at 3000 rpm for 5 min, resulting in around 1.5 – 2.0 ml clear supernatant of low viscosity that will be stored at − 80°C until the analyses of the salivary cortisol levels. Cortisol, the principal stress glucocorticoid produced by the hypothalamic-pituitary-adrenal axis (HPA) [57], will be assessed using a commercially available enzyme-linked immunosorbent assay (Salimetrics®) having an approximate sensitivity of <0.007 μg/dL.
2.6 Procedure
In arriving at the retreat centre, everyone will be informed about the retreatment rules (e.g., accommodation, silence rules, use of electronic devices, etc.) as well as his/her specific time schedules for each of the laboratory sessions. All laboratory sessions will be done on the morning of each assessment day. We will set up different laboratories for the tasks (one for blood extractions, one with nine computers to complete the self-report questionnaires, four EEG separate labs with portable equipment, and another four separate labs for the psychological tasks with one computer per room). The schedule of the assessments will be kept the same for each participant on the three assessment days. The order of the tasks will be the same for all participants (i.e., blood extraction, self-report questionnaires, EEG recording, autobiographical memory task and arithmetic stress task). The expected total duration of each assessment session is approximately 90 minutes.
2.7 Statistical analysis plan
Repeated-measures ANOVAs will be conducted to analyse changes associated with the programs. For the autobiographical memory task and all the biological measures (i.e., genetic measures, EEG, EKG, respiratory rate, and cortisol in saliva), a 2 (Order of program) x 3 (Time) repeated measures ANOVA will be conducted. For the self-report psychological measures, a 2 (Order of program) x 4 (Time) ANOVA will be conducted. We will use age, sex, and previous meditation experience as covariates in the main analyses. Assumptions of normality, homogeneity of variances, sphericity, and homogeneity of covariances will be checked. Further analyses will include post-hoc, p-corrected comparisons (e.g., Bonferroni). All analyses will be performed with SPSS or R Studio statistical software. All tests will be conducted using two-tailed tests using a standard significance value of p<0.05.
The EPIC methylation arrays’ raw data will be normalised and filtered by a robust bioinformatics pipeline procedure [58]. We will then perform an unsupervised analysis of the data via hierarchical clustering, i.e., PCA and t-SNE (by R Studio). In parallel, taking advantage of the known DNA methylation signatures related to distinct cell types or molecular processes, we will: 1) characterize the proportion of different immune cells; and 2) determine the epigenetic age with Horvath’s model or other epigenetic clocks to capture the influence of lifestyle on the DNA methylome [59,60]. This may reveal a link between the MBI programs and the immune response, inflammation, and cells’ physiological condition. Finally, we will focus on supervised analyses to identify differential DNA methylation biomarkers in the two groups (MBRS first vs CCT first) and time points: day 1(baseline), day 4 (program switch), and day 7 (post retreat).
The EEG signals will be assessed both at the sensor level and the source’s space. Source space reconstruction would be performed using the eLoretta procedure over the MNI template. The analysis will be conducted two-fold: 1) the power spectral profiles of each condition (eyes closed and eyes open) will be computed using the fast Fourier transform algorithm; 2) the functional networks will be calculated using a phase synchronization algorithm (i.e., phase locking value). The functional networks will be created using standard regions of interest defined over the AAL atlas.
Diurnal Cortisol Slope (DCS) will be calculated individually for each participant as a function of their respective bedtime and awakening times. DCS will be calculated as awakening cortisol subtracted from bedtime cortisol, divided by each subject’s total waking hours.
Statistical analyses will be conducted only for data from participants who have completed pre-assessment and attended the full retreat program. Intention-to-treat (ITT) analyses will be conducted with all participants, regardless of session or outcome measure completion. ITT mixed models (restricted maximum likelihood (REML) estimation) will be used to account for missing data. Binary logistic regression will be used to evaluate the assumption that data is Missing at Random (MAR). To safeguard the confidentiality of participants’ data, all samples will be anonymized with alpha-numeric codes. These codes will not carry any personal information.
Personal data linked to the alpha-numeric code will not be included in any files used for data analyses and will remain separate from the main database. Data displayed in public repositories (e.g., GitHub or datadryad.org), linked to the resulting publications, will always be anonymized. Data supporting each published study will be included within the article and/or supporting materials. To improve blinding integrity [61], personnel who will analyse the data collected from the study are not aware of the treatment applied to any given group. The authors will be fully responsible of handling and analyzing data independently from any sponsor.
3. Discussion
With this study, we aim to shed light on two main aspects. Firstly, we will analyse the precise interaction between psychological and psychobiological changes associated with MBSR and CCT programs. Secondly, following the rationale clinical trials that have compared the effect of the order of administration of two different modalities of interventions [62,63], we will conduct a cross-over design to explore whether it is important the order of intervention of mindfulness and compassion practices.
This is an ambitious study that attempts to analyze the effects of two meditation programs (MBSR and CCT) by integrating multidisciplinary perspectives. The study will offer a unique insight into the effects on epigenetic variables, using the most advanced analysis methods in the field [58]. Also, adding to current knowledge in the field [64], we aim to correlate epigenetic changes with changes in electrical brain activity and psychological variables, comparing the results before and after the interventions. Thus, we believe that this study offers a great possibility of data integration at various levels that is not common in this type of research. Furthermore, the study has been carefully designed to try to minimize potential biases that are common in meditation research [38,65]. In this regard, the study employs methods of randomization, selection of well-validated measures, and blinding of statistical analyses.
However, we recognize that the protocol has some limitations. Participants in mindfulness studies are typically self-selected, which is a difficult problem to control in research on meditation procedures. Although the cross-over design will allow us to answer important research questions (e.g., the existence of additive effects of the two modalities), it is not certain that the results can be easily transferable to other settings. Also, the retreats include compact and intensive training that is offered in a short period of time, lacking homework exercises. However, although more time may be required to produce lasting psychobiological changes, previous studies support the idea that brief retreat interventions can detect change [35,66]. Finally, the influence of the exposure to other contingencies associated to the retreat (e.g., silence, rest days, etc.) may contribute to the changes that may be found [67–70]. We also recognize the limitation that the follow up is restricted to psychological measures only. Apart from the enormous economic cost of repeating the epigenetic measures, which we cannot afford with our limited funds, it would be impossible to achieve as the participants live in different parts of the country and were recruited only for a one-week retreat.
We hope that, despite these potential limitations, the results may shed light on the neurobiological and psychological mechanisms linked to meditation and compassion in the general population and may contribute to improving the scientific basis of these practices.
Data Availability
All relevant data from this study will be made available upon study completion.
Acknowledgements
We thank Nirakara Lab’s staff, Ignacio Jiménez and Claudia Caldera for their help throughout the project and all the personnel who will help us to administer the tasks and tests in our labs (Leyre Castillejo, Brenda Nadia Chino, Claudia Cogollos, Alberto del Cerro, Elena Diaz, Lucía Hernández, Marina Izquierdo, Ana Obeso, Ana Mar Pacheco, Ester Santamaria, Lucia Torres), as well as Joaquin Martinez, Head of the Hematology Department of the Hospital Universitario 12 de Octubre (Madrid, Spain) for letting us use his laboratory space, and his technicians Alicia Giménez, Laura Carneros Blanco and Lucía Sastre for their agreement to process biological samples. The support of Miguel Angel Casermeiro (Fundación General UCM) was also very valuable. Special thanks to Yago Piedra, director of the Real Colegio Maria Cristina (El Escorial, Madrid), and all his staff, for their support in preparing the retreat. We also thank Frank Zanow (ANT Neuro, Berlin, Germany) who kindly offered his help to provide technical support for the EEG recordings.
This research has been partially funded by a grant from Ministry of Science (PID2019-108711GB-I00) to Carmelo Vazquez, and a Spanish Ministry of Science FPI predoctoral fellowship (PRE2020-092011) to Rosaria M. Zangri.The funders did not and will not have a role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Footnotes
Declaration of interest: The authors declare no competing interest and are fully responsible of all data analyses and subsequent scientific reports and communications of any type.
References
