We hypothesize that the antioxidant and cytoprotective functions of vitamin E combined with the cortisol-lowering effect of chocolate polyphenols and physical activity may help prevent the age-dependent decline of mitochondrial function and nutrient metabolism in skeletal muscle, key underpinning events in protein-energy malnutrition (PEM) and muscle wasting in the elderly. To test this hypothesis, a vitamin E functionalized dark chocolate rich in polyphenols will be developed with the collaboration of Nestlè Company, and its effects will be investigated combined with physical activity in a 6-month randomized case-control trial on pre-dementia elderly patients, a well-defined population of subjects at risk of undernutrition and frailty. Subjects stabilized on a protein-rich diet (0.9-1.0 g protein/Kg ideal body weight/day) and physical exercise program (High Intensity Interval Training specifically developed for these subjects), will be randomized in 3 groups (n = 34 each): controls (Group A) will maintain the baseline diet and cases will receive either 30 g/day of dark chocolate containing 500 mg total polyphenols (corresponding to 60 mg epicatechin) and 100 mg vitamin E (as RRR-alpha-tocopherol) (Group B) or the high polyphenol chocolate without additional vitamin E (Group C). Diet will be isocaloric and with the same intake of polyphenols and vitamin E in the 3 groups. Muscle mass will be the primary endpoint and other clinical endpoints will include neurocognitive status and previously identified biomolecular indices of frailty in pre-dementia patients. Muscle biopsies will be collected to assess myocyte contraction and mitochondrial metabolism. Laboratory endpoints will include the nutritional compliance to the proposed intervention (blood polyphenols and vitamin E status and metabolism), 24-h salivary cortisol, steroid hormones and IGF-1, and molecular indices of inflammation, oxidant stress, cell death and autophagy. These parameters will be investigated in muscle and blood cells by state-of-the-art omics techniques. Molecular and nutritional findings will also be confirmed in vitro using skeletal myotubes, blood leukocytes and neural cell lines. Clinical and laboratory results will be processed by a dedicated bioinformatics platform (developed with the external collaboration of the omics company Molecular Horizon Srl) to interpret the molecular response to the nutritional intervention and to personalize its application.

BACKGROUND Older adults are particularly vulnerable to undernutrition, a state resulting from defective food intake or uptake (nutrient deficiencies) leading to altered body composition and weight loss. Muscle wasting is a major drawback of this condition and a symptom of protein-energy malnutrition (PEM) and metabolic reprogramming of tissues that increase the ageing process. These changes sustain insulin resistance and impaired mitochondrial metabolism of critical organs and tissues, including skeletal muscle. Prevention of undernutrition is critical. Undernutrition correlates with an accelerated and general decline in health conditions (worsening both physical and cognitive/mental aspects), thus increasing the risk of frailty and finally accelerating physical and cognitive decline. Under these circumstances, the majority of people experience a significant loss of locomotor function, with a significant decline in quality of life, and a high risk of falls which often represents the terminal event in life. These factors can lower the frailty threshold for the oldest-old, with the consequent loss of adaptability, which is an essential feature of successful ageing. This process of deteriorating mobility is multifactorial and also includes decline in cognitive function, increased bone fragility, and reduced joint flexibility.

Strong evidence suggests that ageing and cognitive decline are associated with dysregulation of the hypothalamic-pituitary-adrenal axis (HPA axis), with a clear increase in cortisol levels. The effects of hypercortisolism are far-reaching, affecting the skeletal muscle widely, thus leading to significant sarcopenia and fragility. It is well known that HPA axis activity is impaired in Alzheimer's disease (AD) patients. This dysregulation induces an increase in cortisol levels. High levels of cortisol, one of the most catabolic hormones, also lead to noticeable sarcopenia. Mechanistic aspects include antagonistic effects on the insulin axis (secondary insulin resistance) and consequent metabolic reprogramming of tissues to gluconeogenesis sustained by non-carbohydrate precursors that include amino acids derived from the proteolytic degradation of muscle proteins. Comorbidity in frail people can further sustain the secretion and metabolic effects of cortisol, especially undernutrition and PEM. Some of us previously showed that cortisol levels are significantly higher in patients with AD and severity of the behavioural symptoms, and more importantly, changes in body mass, significantly correlated with cortisol levels. Therefore, cortisol levels, which are not regularly evaluated in AD patients, would help predict patients at risk of weight loss.

Moreover, recent findings revealed a significant decrease in cortisol levels in response to chronic physical activity in healthy individuals and patients with dementia. Physical activity treatment (PT) is a non-pharmacological treatment with great potential to attenuate the cognitive decline in healthy elderly. In patients with Mild Cognitive Impairment (MCI) it was observed that 6 months of PT significantly ameliorates BMI, 6' Walking Test (6MWT), systolic and diastolic blood pressure, glucose, cholesterol, and triglycerides. Importantly, PT may preferably be undertaken as high aerobic intensity (85-95% of maximal heart rate) intervals (HIT), as this yields superior effects on the cardiovascular system compared with PT of moderate or low intensity. HIT has successfully been applied in older individuals (8,9), and in frail populations such as patients with heart failure.

Along with physical activity, macro and micronutrients, are reported to interact with the activity of HPA-axis and to help reducing cortisol levels. Chocolate polyphenols appear to have significant effects with impact on both mental well-being and metabo-inflammatory symptoms of chronic exposure to such stress hormone (3,11-13). Cocoa-derived flavonoids can lower the levels of the active hormone cortisol. Mechanistically, these natural molecules inhibit 11β-hydroxysteroid dehydrogenase (11β-HSD) type 1, an enzyme involved in reducing cortisone to the active form cortisol. The intake of these and many other micronutrients and homeostatic factors decrease with aging due to general worsening of quantity and quality of food intake. Together with micronutrients, protein intake is a critical aspect and a major risk factor for PEM and frailty. Nutritional supplementation for frail people has been shown to slow their functional decline, improving both muscle mass and strength, particularly if this is combined with physical activity. It is now established that nutritional recommendations, including adequate protein and micronutrient intake, are important for a better quality of life in the elderly, which is common management approach of older people who are frail or at risk for developing frailty.

Vitamin E is a fat-soluble essential micronutrient with unique properties as antioxidant and cell protection factor. It is present in cellular membranes of all tissues to scavenge peroxyl radicals formed by free radical attack on polyunsaturated fatty acids. This function is particularly important to prevent mitochondrial damage and the uncontrolled release of free radicals from these organelles in the muscle. Its intake and function as cell protection factor and immune system modulator can be compromised in the elderly (20); moreover, preclinical and human experimental studies show that vitamin E positively influences myoblast proliferation, differentiation, survival, membrane repair, mitochondrial efficiency, muscle mass, muscle contractile properties, and exercise capacity. Furthermore, recent studies on the human metabolism of vitamin E demonstrated that the biotransformation of this vitamin in human tissues forms bioavailable long-chain metabolites with a role as tissue detoxification (PXR and PPAR-gamma agonist activity) and anti-inflammatory (LOX-5 inhibition) mediators. Therefore, for multiple reasons, vitamin E supplementation in the diet as a measure to support physical training in preventing age-associated PEM is worth investigating.

AIMS The study aims to investigate if regular consumption of vitamin E-functionalized and polyphenol-rich chocolate can support physical exercise high-protein diet to slow down the progression of protein-energy undernutrition in pre-dementia elderly people. Specifically, the primary aim is to investigate whether regular consumption of vitamin E-functionalized and polyphenol-rich chocolate and regular exercise practice boost lower limb muscle mass in pre-dementia elderly people. The secondary aims are to investigate the effect of regular consumption of vitamin E-functionalized and polyphenol-rich chocolate and regular exercise practice on muscle strength, cognitive function, vascular function, metabolic and physical functions, as well as mitochondrial respiration, circadian cortisol curve, blood hormones, and inflammatory status in blood and mRNA in pre-dementia elderly people.

PROCEDURE The study will be a randomized, double blinded, controlled trial with parallel groups including active control and shame groups. One hundred and fifty individuals with MCI and subjective cognitive decline without functional deficits will be screened for eligibility and those that comply with inclusion and exclusion criteria will be confirmed and the informed consent will be allocated for testing and undergo preliminary evaluations (T00). After preliminary evaluation, all the individuals included in the study will undergo a 4 to 6-week "Run-in" phase during which the high protein diet (HPro) will be introduced and all subjects will be trained to implement the High-Intensity Training physical exercise (HIT) program that will be developed during the study. Immediately after the "Run-in", a pre-intervention (T0) evaluation will be undertaken. Consequently, participants stabilized on the HPro Diet + HIT which will be the common treatment for all participants, will be randomly assigned (utilizing an online statistical computing web program) to one of the three arms of the nutritional intervention in which the effect of vitamin E (VE) will be investigated separate or combined with the effect of chocolate polyphenols (HPP) compared to control treatment. The intervention will last six months; assessments will be performed after three months (halfway through the intervention) and at the end of the intervention (T1 and T2, respectively). A follow-up assessment will be performed three months after the end of the intervention after the restoration of baseline diet and physical activity conditions (T3). Each group will include 34 participants; a 20% dropout has been estimated based on previous studies.

SAMPLE SIZE CALCULATION Considering an alpha = 0.05, a power = 0.8 and the 20% of estimated drop out, we aim to recruit 102 subjects (34 in each group). Main outcome is "muscle mass", and for all the groups treatment duration will be 6 months. In 6 months in the target population the loss of muscle mass is assumed to be 1.0-1.5% (+/- 0.5%) [5] [48]. In Control group (Group A), which includes people undergoing targeted exercise, the expected increase is 2% 1.5% (+-0.5%) [5] [48]. In treatment groups (Groups B and C) the median average expected increase at second follow-up is 4% 2% (+-0.5%), and 1.5% 4% (+-0.5%) at 6 months [6] [49]. The rate of lost at follow-up, derived from previous studies, is 20% (+-2%) [7] [50]. Correlation between repeated measures is assumed to be 0.5, variance explained by the between-subjects effect 6.25 and error variance 65.

All estimates were performed using Stata v.16.1 (StataCorp LP, College Station, TX, USA) by "power repeated" command.

STATISTICAL ANALYSIS Statistical analysis will be conducted under the supervision of an expert in biostatistics (dr Gili, at Coordinator Unit) and with the support of LIPOSTAR software provided by external collaborator C2. A two-way repeated measures ANOVA, including age and gender as covariates, with "time" as within-group factor and "treatment" as between-groups factor will be utilized to calculate difference between groups. In the presence of significant effects, a multiple comparisons tests with Bonferroni's correction will be performed. The familywise alpha level for significance will be set at 0.05 (two-tails), with Bonferroni's correction when needed, for all the analyses.

SIDE EFFECTS Sides effects may be related to the assessment procedures: strength, voluntary activation, and electrically evoked potential tests may cause muscle soreness and discomfort during the procedures. In case of persistent discomfort the procedure will be immediately stopped. Also, side effect might be caused by blood draw and muscle biopsy: subjects may experience some side effects related to the blood draw in the draw site, which normally gets between the following days. Also, subjects may experience some soreness in the site of the biopsy, muscle tightness and fatigue in the few days after biopsy was taken. In the case of these events, subject will be monitored and the family doctor will be informed.

DATA AND SAFETY MONITORING COMMITTEE A log-diary will be kept by each participant and will be checked weekly by the investigators and collaborators. In the diary participants will include information about possible adverse events caused by assessment procedures or related to the diet and training, any important points about the response to the interventions, any possible discomfort experienced during or after the training, or notes regarding diet and supplementation. Prof. Gianluca-Svegliati Baroni of the Gastroenterology Division of the University Hospital of Ancona, Italy will serve as external scientific supervisor of the clinical trial. He is an expert in clinical and preclinical studies of human nutrition and metabolism. He will advise on specific Code: CHOKO-AGE Data: 10/06/2021 Version:1 30 tasks and monitor the different phases of clinical trial from organization to implementation of activities, data gathering and evaluation/interpretation. The quality assurance standards of University of Verona will be adopted to monitor the clinical trial. A delegate of this University will be nominated to perform the monitoring of the different phases of the trial utilizing internal SOPs. The entire set of clinical procedures, operator's activity and collection of experimental data will be verified during a series of visits by the monitor that will occur at the beginning and the end of each time point in the study (Time T00 to T3).

Inclusion Criteria:

• Presence of Mild Cognitive Impairment or Mild Dementia. Recruited individuals will be assessed by means of Neuropsychological tests (Mini Mental State Examination, evaluations criteria from Diagnostic and Statistical Manual for Mental Disorder-5) which will be performed by an expert Neuropsychologist

Exclusion Criteria:

• Presence of kidney or liver failure, or any other liver or kidney disease;
• Presence of gastro-intestinal disorders (i.e. irritable bowel syndrome);
• Presence of food intolerance;
• Presence of heart failure, angina, pulmonary disease, cancer and cancer-related cachexia;
• Presence of coagulation disorders;
• Addictive or previous addictive behaviour, defined as the abuse of cannabis, opioids or other drugs, carrier of infectious diseases;
• Presence of musculoskeletal diseases;
• Suffering from mental illness, inability to cooperate;
• Suffering from known cardiac conditions (e.g. pacemakers, arrhythmias, and cardiac conduction disturbances) or peripheral neuropathy;
• Regular users of any proton pump inhibitors (e.g., omeprazole, lansoprazole, pantoprazole), antibiotics, anticoagulant medication or antiplatelet medications in high dose (es: acetylsalicylic acid >200mg x day);
• Mini Mental State (MMSE): results >= 10 points

All Individual participant data (IPD) collected during the trial will be available after identification. Also, Study Protocol, Statistical Analysis Plan and Informed Consent Form will be shared. The availability period will start immediately following the publication and will not end. The declared data will be shared with Researchers who provide a methodologically sound proposal to achieve this proposal's aims and for IPD meta-analysis. Request and proposal have to be sent to francesco.galli@unipg.it and massimo.venturelli@univr.it.

The availability period will start immediately following the publication and will not end.

The IPD will be shared with Researchers who provide a methodologically sound proposal to achieve this proposal's aims and for IPD meta-analysis. Request and proposal have to be sent to francesco.galli@unipg.it and massimo.venturelli@univr.it.