Background

Docosahexaenoic acid (DHA) is one of the most abundant polyunsaturated fatty acids in the human brain. Epidemiological research has linked high DHA consumption with a lower risk of Alzheimer's disease (e.g., Morris et al., 2003). People ingest DHA from foods such as fatty fish, walnuts, or flax seeds, or from dietary supplements. These are being marketed in different formulations of DHA, or DHA mixed with other omega-3 fatty acids such as eicosapentaenoic acid (EPA). Elevated blood levels of DHA correlate with reduced dementia risk in some but not all studies (Schaefer et al., 2006; Laurin et al., 2003). Animal studies have reported a reduction of amyloid, tau, and neuritic pathology with oral intake of DHA (Lim et al., 2005; Green et al., 2007; Calon et al., 2004).

Findings

DHA has been tested in clinical trials by itself and as part of other food-supplement formulations.

In 2000, Karolinska University Hospital began the OmegAD trial, which evaluated a six-month course of treatment with a DHA-containing fish oil formulation called EPAX 1050 TG in 204 people with mild to moderate AD, of whom 174 completed the study. The treatment and placebo groups did not differ on either of the main outcome measures, decline on the MMSE and ADAS-cog, or on neuropsychiatric symptoms overall. However, analysis of a small subgroup of the 32 mildest cases did suggest less decline on the MMSE, though not ADAS-cog, and a similar slowing of decline appeared to occur in the placebo group once switched to DHA after six months (Freund-Levi et al., 2006; see news and commentary; Freund-Levi et al., 2008). Substudies suggested that DHA treatment increased CSF levels of DHA and other fatty acids and decreased levels of tau, as well as changing expression of inflammation-related genes and release of certain cytokines in white blood cells (Freund-Levi et al., 2014; Vedin et al., 2012; Vedin et al., 2008).

From 2007 to 2009, the Alzheimer's Disease Cooperative Study conducted a study at 51 centers in North America to evaluate an 18-month course of 2 grams per day of DHA in 402 patients with mild to moderate AD, of whom 295 completed the trial. DHA had no effect, relative to placebo, on rate of decline on either the ADAS-cog or the CDR-SOB clinical/functional assessment. Analysis by participants' ApoE genotype indicated a slower cognitive decline in ApoE4 noncarriers, who may have been relatively less advanced in their disease (Quinn et al., 2010; Nov 2010 news).

This pharmacogenetic hint of a differential effect prompted several bioavailability studies at the Université de Sherbrooke, Quebec, Canada, between 2009 and 2011, which analyzed the percentage of DHA in lipoproteins, incorporation into plasma lipids, and pharmacokinetics by ApoE genotype in healthy young adults and people with MCI (e.g., Chouinard-Watkins et al., 2013; Plourde et al., 2014).

The largest clinical trial of DHA is MAPT, conducted in four cities in France. This three-year, secondary prevention study in 1,680 participants began in 2008 and enrolled people 70 and older who reported a subjective memory complaint and a mild functional loss, were frail and walked slowly, but did not meet an Alzheimer's diagnosis (Carrié et al., 2012). MAPT compared three interventions—800 mg DHA and 225 mg EPA daily alone, DHA/EPA plus a multidomain behavioral intervention, multidomain behavioral intervention alone—to placebo (Gillette-Guyonnet et al., 2009). Neither intervention, alone or in combination, significantly slowed cognitive decline as measured by a composite score of four tests of recall, orientation, processing and verbal fluency (Andrieu et al., 2017). An amyloid PET substudy compared the effects of the multidomain intervention and DHA/EPA according to brain amyloid status. The multidomain intervention, with or without DHA/EPA, improved composite scores after 36 months compared to placebo in the amyloid positive, but not negative, subset; DHA/EPA alone had no effect in either subset (Delrieu et al., 2019). In another PET analysis, cortical amyloid was lower after two years in people who got the multidomain intervention, with or without DHA/EPA, but not in those who received DPA/EPA alone (Hooper et al., 2020).

A subsequent analysis of MAPT data suggested higher odds of cognitive decline for people with low DHA/EPA levels (Bowman et al., 2019). Lower serum DHA was also correlated with more cerebral amyloidosis and atrophy (Yassine et al., 2016; Aug 2016 news). In response to those findings, MAPT investigators in April 2018 began LO-MAPT. This 18-month study is enrolling 400 older adults with low DHA/EPA status, who will be randomized to 1.53 g/day DHA/EPA or placebo. Participants must have subjective memory complaints or a family history of AD. The primary outcome is change in a cognitive composite of scores on the Free Cued and Selective Reminding Test, MMSE, and Category Naming Test. The trial offers an 18-month open-label extension, and will end in 2022.

From 2009 to 2011, the company NeuroBioPharm Inc. ran a six-month trial at 14 different sites in Canada to compare soft-capsule formulations of fish oil and krill oil, each containing 100 mg DHA, to placebo in 175 people with mild to moderate Alzheimer's disease. Krill oil contains DHA and EPA in a form claimed to be more bioavailable than fish oil (e.g. Schuchardt et al., 2011). This trial used the Neuropsychological Test Battery as primary outcome; data have not been published.

One DHA study, at Oregon Health and Science University, started in May 2014. According to published baseline data, the trial recruited 102 people age 75 and older who were cognitively impaired but did not have dementia, and suboptimal blood levels of DHA and EPA of less than 110 ug/ml. Participants were randomized to receive a three-year course of 1.65 grams/day of EPA plus DHA or soybean oil placebo (Bowman et al., 2019). Seeking to understand DHA's effect on vascular cognitive aging, this trial used white-matter hyperintensity as primary outcome; secondary and other outcome measures included other brain imaging modalities as well as blood-based indicators of endothelial health and neuropsychological tests. Top-line results were presented at the 2020 CTAD. Supplementation failed to slow accumulation of white-matter hyperintensities in the 45 placebo and 42 treated participants who had a least one follow-up MRI. At the end of the study, 24 of the treatment group had plasma DHA+EPA levels greater than 110 ug/ml. In a prespecified analysis, this group had fewer white-matter hyperintensities and maintained better white-matter integrity than the placebo group. There were no differences in total brain volume, ventricular volume, medial temporal lobe atrophy, excessive cognitive functions, or adverse events.

In June 2016, a pilot study at the University of Southern California began evaluating how much of a DHA supplement enters the central nervous system, and whether CSF levels are influenced by ApoE4 status (for review, see Yassine et al., 2017). The trial enrolled 31 adults older than 55 with a family history of dementia, who took 2g DHA per day or placebo for 26 weeks. The primary outcome was change in CSF DHA before and after supplementation. According to published results, this dose of DHA resulted in a 200 percent increase in plasma DHA compared to placebo, but only a 28 percent increase in CSF DHA (Arellanes et al., 2020). ApoE4 carriers had slightly lower CSF DHA. For EPA, CSF levels increased 43 percent, but were significantly lower in ApoE4 carriers than in noncarriers. Other endpoints including brain volume and cognitive scores did not differ between supplement and placebo groups.

In July 2018, the same group began a larger follow-up study in 320 healthy adults age 60-80, who have at least one dementia risk factor. Participants will take 2g DHA or placebo daily for two years. After six months of treatment, investigators will analyze CSF fatty acids; changes in MRI measures of structural and functional connectivity and cognition will be assessed at two years. The study will run through 2023.

For details of clinical trials of DHA in Alzheimer's, see clinicaltrials.gov.