The goal of this clinical trial is to better understand the relationship between blood-brain barrier (BBB) function and stress resilience and to investigate a potential causal role for BBB function in stress resilience in humans. The researchers will look for particpants that are young adults to answer the following questions to reach our goal:

• Is better BBB integrity, measured with neuroimaging, associated with better stress resilience in the short term and in the long term?
• Does the administration of Metformin improve BBB integrity?
• Is improved BBB integrity a possible link between Metformin and better short- and long-term stress resilience?

The participants will do the following during our study:

• Participants will fill in online surveys on stressor exposure and mental health once every four weeks over 36 weeks
• Participants will visit the study site 4 times:
• The first time for a screening process which includes questionnaires, a medical exam and a blood sample.
• The next 2 visits will include a blood sample, a medical exam, an MRI scan and a list of questionnaires
• The last visit will include another blood sample, a medical exam and a list of questionnaires.
• Between visit 2 and 3, the participant will take either Metformin or Placebo

STUDY DESIGN

Prospective, randomized, parallel-group, placebo-controlled, double-blind, multi-center experimental study to investigate the relationship between individual variation in BBB integrity and individual variation in stress resilience in young healthy adult participants at risk for stress-related mental health problems, making use of a 12-week oral administration of Metformin (850 mg twice daily at full dose) as an experimental tool to enhance individual variation in BBB integrity (experimental manipulation). The duration of the study for a participant will be 36 weeks in total. Participants will be assessed at three different study sites (i.e. Mainz, Germany, Zürich, Switzerland, Warsaw, Poland).

The study is divided into 10 phases of 4 weeks each, and has the following study flow and intervention scheme:

• Screening phase: Potential participants will be screened and assessed for eligibility. If eligible, informed consent will be obtained.
• Time point T0, week 1 - Baseline: Medical exam, blood sampling, randomization, neuroimaging, questionnaires, distribution of study drug. Start of the online monitoring of stressor exposure and mental health, which participants will perform every four weeks (time points T0 to T9), throughout the study.
• Time points T0 to T3, week 1 - 12 - Experimental Manipulation phase: Four weeks of dose increase (T0 to T1), i.e., first and second week 500 mg once daily, third and fourth week 500mg twice daily. Full-dose administration (850 mg twice daily) for eight weeks (T1 to T3).
• Time point T3, week 12 - End of Experimental Manipulation: Medical exam, blood sampling, neuroimaging and questionnaires. Adverse Event (AE) monitoring.
• Time points T4 to T9, week 12 to 36 - Follow-up: separated into early (T4 to T6) and late (T7 to T9) follow-up period.
• Time point T9, week 36 - End of follow-up: Medical exam, blood sampling, questionnaires and AE monitoring.

OUTCOME MEASURES

Stress resilience is quantified as inverse stressor reactivity (SR). Short-term stress resilience is the average inverse SR score in the 12 weeks after the Experimental Manipulation phase (online monitoring time points T4 to T6). Long-term stress resilience is the average inverse SR score in the 24 weeks after the Experimental Manipulation phase (time points T4 to T9). SR is calculated as an individual's deviation (residual) from a regression of mental health problems [z-standardized sum score of anxiety and depression symptoms, as assessed with the Patient Health Questionnaire - Anxiety and Depression Scale (PHQ-ADS) during online monitoring] on stressor exposure [z-scored sum of occurrences of all daily hassles (DHs), as assessed with the Mainz Inventory of Microstressors (MIMIS) during online monitoring] in the study sample.

Whole-brain BBB integrity is quantified as inverse whole-brain water exchange rate (kw) from blood to brain tissue (BBB permeability), as assessed with neuroimaging before (time point T0) and after the Experimental Manipulation phase (time point T3).

CALCULATING THE SR SCORE

To build the SR score, first, the stressor exposure score E is calculated for every four-weekly monitoring time point after T0 (T1 to T9) as the z-scored sum of occurrences of all daily hassles (DHs), assessed with the MIMIS. Life event (LE) occurrence, assessed with a life-events list, is expected to be less frequent than DH occurrence but, if reported, expected to correlate with DH occurrence. This will be used to corroborate the validity of the DH-based E score. For each monitoring time point, a mental health problem score P is also calculated as the z-standardized sum score of anxiety and depression symptoms, as assessed with the PHQ-ADS. To quantify how strongly participants' mental health problems relate to stressor exposure, the researchers then calculate the relationship between E and P in the analysis sample for a given study phase by fitting a linear mixed model to predict P by E over all included participants and monitoring time points Tx, with random slopes and intercepts for participants. The E-P regression line is then determined by the fixed effect estimates for the EP slope and intercept and serves as the normative E-P relationship for the analysis sample over the study phase of interest. The form of regression that explains most variance in P (linear or quadratic) will be used. At each Tx, the participants' individual E scores will be entered into the normative E-P line equation, giving us their expected P score when assuming normal stressor reactivity. The SR score is the individual's average residual onto the regression line for the study phase of interest.

MEASURING BBB INTEGRITY

A pseudo-continuous ASL (pCASL) MRI sequence will be used to estimate BBB permeability. By reading out pCASL signals at different post-labeling delays (PLD) under different diffusion gradients, perfusion signals in capillary and brain tissues can be reliably estimated and subsequently fitted into a model to estimate the water exchange rate (kw) from blood to brain tissue, an index quantifying BBB permeability. Analysis will use a toolbox provided by the sequence developer (http://loft-lab.org/index-5.html) and includes image reconstruction, preprocessing (motion-correction, coregistration, skull-stripping), and kw modeling. Whole-brain averaged kw values quantifying overall BBB permeability will be calculated. BBB integrity is quantified as inverse permeability.

ADDITIONAL OUTCOME MEASURES

For additional exploratory analyses, blood samples will be taken and used to determine blood-based molecular and cellular markers of immune system, metabolic, and BBB function (including proteins, peptides, RNAs, metabolites). In these additional analyses, BBB integrity will be assessed using both blood-based biomarkers and regional neuroimaging measures (esp. of hippocampus, ventral striatum, and prefrontal cortex), and the association of immunological covariates with BBB integrity and stress resilience will also be considered. Further, socio-demographic and psycho-social variables will be assessed via questionnaire and tested for their association with BBB integrity and stress resilience.

SAMPLE SIZE

Because the association of BBB function in stressor-exposed individuals at risk for stress-related mental health problems with stress resilience (inverse SR score) has never been investigated, the researchers base their power calculation for the primary endpoint on the sensitivity to detect a medium effect size of Cohen's d = 0.3. With a power of 0.8, a two-sided alpha-level of 0.05, and a linear mixed model with eight predictor variables (3 variables of interest and 5 covariates), n=109 participants are needed. Taking into account 10% attrition, our aim is to recruit a total number of 122 participants (61 in each group).

STRATEGIES FOR DATA HANDLING AND THE DISSEMINATION OF RESULTS

Data collection and data management will be conducted in compliance with the principles of the declaration of Helsinki (1996) and relevant national and regional regulations and will only follow study protocols approved by the regional ethics committee at each partner site/country. The investigator at each partner site will be responsible for data management in accordance with these principles. The investigators will collect and process MRI imaging data (digital: DICOM), biosamples (blood), biological data from the analysis of biosamples (digital: concentrations of plasma proteins/peptides, RNAs, metabolites, and other molecules), and computerized (digital: offline and online collection) and paper-and-pencil (analogue) questionnaire and interview data. Data will be made findable with a study description and metadata following fairsharing.org conventions. Open accessibility and data re-use are complicated by the sensitive personal nature of most data and the principled non-anonymizability esp. of MRI data. Participant consent forms will permit sharing of pseudonymized data and biosamples among the sites and with collaborators within the European Union and countries with an adequacy decision under Article 45 GDPR, under the supervision of a Data Use and Access Committee (DUAC). All anonymized data will be made openly accessible through public depositories (in particular, OSF) upon publication or at latest two years after the termination of the project and will remain accessible for at least ten years upon completion of the study.

Inclusion criteria: (participants meeting all of the following criteria will be considered for enrollment in the study)

1. Absence of mental disorder diagnosis.
2. University students.
3. GHQ-28 > 20
4. Three or more adverse life events acc. to LE list in the past
5. Beck Depression Inventory (BDI) ≤ 14 & Columbia-Suicide Severity Rating Scale (C-SSRS) ≤ 1. Thereby concurrent depression and suicidality are excluded.
6. Age 18 to 25 years
7. Ability of participant to understand character and individual consequences of the study (MMSE Folstein > 28)
8. Signed and dated informed consent of participant

Exclusion criteria: (participants presenting 1 of the following criteria will not be enrolled in the study)

1. Life-time and current diagnosis of any severe mental disorder determined by M.I.N.I. diagnostic interview.
2. Known history of brain injuries or neurodevelopmental disorder.
3. Evidence of neurodegenerative disorder (e.g., Parkinson).
4. Multimorbidity or significant organ (esp. liver or renal) dysfunction or manifest diabetes or substance abuse (esp. alcohol).
5. Contraindication to metformin such as renal insufficiency (Creatinin-Clearance< 60ml/min), recent (<3 month) ischemic events (e.g. myocardial infarction or stroke).
6. Women of childbearing age, who do not practice a medically accepted contraception (i.e., systematic contraceptives, diaphragm, condoms with spermicide, sexual abstinence) during the study and during a 2 years post-study period and who do not present a negative pregnancy test (serum or urine).
7. History of hypersensitivity to the study drug, to any drug with similar chemical structure, or to any excipient present in the pharmaceutical form of the study drug.
8. Diabetes type 2 (would result in interference with the experimental manipulation)
9. Participation in other studies employing a drug during the present study or within the last three months.
10. Current use of antidiabetic, weight-loss, or psychoactive medication or substances.
11. Pacemaker, implanted medical pumps, implanted cardiac catheters or acute or unstable heart disease (angina pectoris).
12. Intracranial implant (aneurysm clips, shunts, stimulators, cochlear implants or electrodes) or other metallic objects inside or near the head (mouth excluded) that cannot be removed.
13. Claustrophobia or another contraindication to MRI.
14. Insufficient German language skills.

Data will be made findable with a study description and metadata following fairsharing.org conventions. Data will be made interoperable by using common standardized formats (in particular, BIDS for MRI) and a detailed codebook. Open accessibility and data re-use are complicated by the sensitive personal nature of most data and the principled non-anonymizability esp. of MRI data. Participant consent forms will permit sharing of pseudonymized data and biosamples among the sites and with collaborators under the supervision of a Data Use and Access Committee (DUAC). All anonymized data will be made openly accessible through public depositories (in particular, OSF).

All anonymized data will be made openly accessible through public depositories (in particular, OSF) upon publication or at latest two years after the termination of the project and will remain accessible for at least ten years upon completion of the study.

Collaborators can get access to the pseudonymized data under the supervision of a Data Use and Access Committee (DUAC). All anonymized data will be made openly accessible through public depositories (in particular, OSF).