Background

Flavanols are naturally occurring bioactive plant products in tea, wine, and chocolate. The predominant flavanols in cocoa include catechin, epicatechin, and procyanidins, which are thought to give these foods heart and other health benefits. These compounds and their metabolites are credited with improving vascular function, insulin resistance, blood pressure, and with having antioxidant, anti-inflammatory, neurogenic, and neuroprotective effects. Several epidemiological studies and small trials of short duration suggest that cocoa flavanols may preserve cognition in older adults (e.g. Brickman et al., 2014, review by Socci et al., 2017). The CoCoA study, a randomized trial, reported a boost in cognition after eight weeks of 520-993 mg daily cocoflavanol supplement in 90 cognitively healthy adults (Mastroiacovo et al., 2015), or 90 with MCI (Desideri et al., 2012). More recently, in a study of 211 healthy adults, treatment with 12 weeks of cocoa flavanols did not improve performance on the primary outcome of an object-recognition task, but did improve a secondary outcome of hippocampal-dependent list learning (Sloan et al., 2021).

In preclinical work, cocoa extract was reported to reduce Aβ oligomerization in vitro (Wang et al., 2014). Anti-aggregation activity was previously reported for the catechin EGCG extracted from green tea leaves. No results are published on treatment of Alzheimer’s mouse models with cocoa extracts, but epicatechin-containing grape seed extracts were shown to improve plaque and tangle pathology, neuroinflammation, oxidative stress, and cognitive behavior in mouse models of amyloidosis (Wang et al., 2008; Wang et al., 2009; Wang et al., 2010).

Findings

In September 2016, the Cocoa Supplement and Multivitamin Outcomes Study of Cognition (COSMOS-Mind) began enrolling 2,262 people 65 or older without dementia to evaluate the effect of three years of daily supplementation on global cognition. The four-arm study compared a patented cocoa extract plus a multivitamin (Centrum Silver, Pfizer), to the extract plus placebo, multivitamin plus placebo, or placebo only. The cocoa extract contains 500 mg of flavanols with 80 mg of epicatechin. Participants were recruited by mail across the U.S. They took cognitive assessments via telephone at baseline and annually. The primary outcome was a cognitive composite comprising the Telephone Interview for Cognitive Status-modified (TICSm), and telephone versions of delayed recall, oral trail-making test, category and verbal fluency, digit span, and digit ordering. Other outcomes were scores for executive function and memory, and incident MCI and AD-related dementia. The study design is published (Baker et al., 2019). This is an ancillary study of the ongoing COSMOS trial, which examined the effect of cocoa extract and a multivitamin on cardiovascular and cancer endpoints in more than 21,000 older adults (Rautiainen et al., 2018; Sesso et al., 2022).

COSMOS-Mind finished in February 2021; results were shown in November at CTAD. Cocoa extract had no effect on global cognitive function over three years. In contrast, the multivitamin improved cognitive measures, compared to placebo. Seventy-seven percent of participants completed the trial. Most were white, and had at least some college or higher education. Results are published (Baker et al., 2023). On secondary outcomes, the study found no difference in the incidence of MCI or dementia with cocoa supplement compared to placebo (Sachs et al., 2023). A subgroup analysis suggested possible cognitive benefits of cocoa flavanols in people with poorer diet quality (Vyas et al., 2023).

In August 2016, another substudy, COSMOS Web, began to measure changes in cognition via an online test battery in 3,947 COSMOS participants randomized to the cocoa extract, a multivitamin, both, or placebo for three years. The primary outcome was change in episodic memory after one year, measured by the ModRey Assessment of Immediate Recall and Learning and Episodic Memory (Hale et al., 2019). The study included additional cognitive tests and MRI measures of brain structure and blood flow as secondary endpoints. It finished in December 2020. According to published results, cocoa flavanols did not change the primary outcome. However, the supplement did improve ModRey performance in the subgroups of people with poorer-quality diets and with low flavanol intake at baseline (Brickman et al., 2023). Increases in a urine biomarker of flavanol intake correlated with cognitive improvement after one year of supplementation.

From 2015-2019, a placebo-controlled study tested one year of a combination of flavanol-rich dark chocolate and omega-3 fatty acid-rich fish oil in older adults with subjective cognitive impairment or mild cognitive impairment (Vauzour et al., 2023). The trial found no cognitive benefit or change in brain structure from the supplementation, which provided 500 mg flavan-3-ols, 1.1 g DHA, and 0.4 g EPA daily. In a separate study, 200 mg of flavanols plus a berry extract, taken daily for 12 weeks, was reported to improve scores on tests of executive function in healthy adults (García-Cordero et al., 2021).

In January 2019, a study at Columbia University began testing the effects of a cocoa extract on neuroinflammation in normally aging adults. The 146 participants, age 50-69, will take 650 mg Cocoapro flavanols or placebo for three months. The primary outcome is levels of the inflammatory marker HMGB1 in blood, with secondary outcomes of other markers of inflammation. Cognitive endpoints include the ModRey and others. This trial will run until mid-2024.

For details on these trials, see clinicaltrials.gov.