A Dose Response Investigation of Docosahexaenoic Acid (DHA)

Docosahexaenoic acid (DHA) is vital for the structure of cell membranes, essential for brain and eye development, found at high levels throughout the brain and nervous system and is important for optimal cognitive development in children, the growth of the fetus during pregnancy and healthy breast milk during lactation. DHA has also been shown to aid in the protection against primary and secondary symptoms of cardiovascular disease as well as macular degeneration. However, whether DHA is deemed to be essential in the diet remains controversial. Meanwhile, the Institute of Medicine (now the National Academy of Medicine) set an adequate intake for DHA's precursor, alpha-linolenic acid (ALA), at 1.6 g/d and 1.1 g/d for healthy males and females, respectively. The investigators believe the confusion arises from the approach to studying omega-3 fatty acids, particularly DHA requirements. For these studies, DHA is supplemented, and often complex, multifactorial disease outcomes are used as endpoints. The doses of DHA used in these studies are often subjective and participants, including controls, have varying DHA intakes and levels. Supplementing a population, like vegetarians or vegans whose levels are similarly low would be useful. Classically, it was believed that that when DHA is consumed a portion is converted "backwards" to eicosapentaenoic acid (EPA) which then accumulates in a process called "retroconversion." However, using a novel approach called compound-specific isotope analysis (CSIA), the investigators observed that the rise in EPA is not primarily from retroconversion, but rather from a backlog of EPA primarily originating from ALA. Initially demonstrated in rodent models by CSIA, the team confirmed by similar modeling that DHA consumption in humans results in the same backlog of EPA. Critically, DHA induces a backlog in metabolism of EPA and dietary ALA appears to be essential for this backlog. Thus, the increase in EPA represents a biochemical feedback pathway that slows DHA synthesis in response to sufficient DHA.

Significance: By precisely determining the lowest amount of DHA required to activate this feedback mechanism it may be useful for future estimation of DHA requirements. Moreover, the problem with the amounts suggested for the unofficial recommendations is that if the global population were to consume these amounts it would be environmentally concerning as this nutrient is ecologically sensitive with its major source coming from already stressed ocean fish stocks, encouraging unsustainable fishing practices, something environmentalists and researchers have said are creating legitimate climate and health concerns.

Hypothesis: The dose at which EPA increases with increased DHA is the point at which elongation slows, indicating a significant feedback pathway is present.

Objectives: 1. To determine the dose-response for DHA to increase EPA. 2. Investigate the DHA dose and time that increases EPA using CSIA. 3. To measure EPA, DPAn-3, and DHA turnover and loss rates. 4. Provide data for exploratory analysis related to PUFA metabolism, ALA consumption and its effect on EPA and DHA levels, and DHA's effects on disease related biomarkers, C-Reactive Protein (CRP), and blood clotting.

Trial Design: Co-principle investigators Dr. Richard Bazinet and Dr. John Sievenpiper and co-investigators, Drs. David Jenkins and Adam Metherel along with PhD student Amy Symington, will conduct a double blinded, placebo-controlled, dose-response supplementation trial. Seventy-two healthy vegans or vegetarians who are 18-50 years old will be randomly assigned to 1 of 6 groups. Prior to randomization, participants will undergo a 2-week run-in phase to gather their questionnaire data. Once randomized, each group will take DHA supplements of either 0mg (placebo), 100mg, 200mg, 400mg, 800mg and 1000mg per day over the course of 8 weeks. These doses represent the range of DHA intake levels spanning average population intakes (<100 mg/d), frequently recommended intakes (250 - 500 mg/d), and DHA intake levels known to increase plasma EPA levels (1000 mg/d). A questionnaire will be administered to obtain information in relation to approximate ALA, EPA and DHA intake before and after the supplementation period. A 3-day weighed diet diary will be provided to obtain food intake over the course of the supplementation period to gauge ALA consumption. Participants will be asked to refrain from DHA rich foods during the supplementation period. This will not be an issue as this population generally do not consume DHA rich foods. Baseline intake of the essential fatty acid ALA will be assessed via questionnaire and verified by blood samples.

Participants: The study will recruit 72 healthy vegetarians and vegans to allow for the lowest possible baseline EPA and DHA levels. Exclusion criteria are as follows: consumption of EPA and/or DHA supplement within the past six months, have 3% or higher of DHA in their total plasma lipids, BMI <18 kg/m2 or >30 kg/m2, are menopausal or post-menopausal, pregnant or breastfeeding, chronic or communicable diseases (like multiple sclerosis, kidney and inflammatory bowel disease, Type 2 diabetes, cancer or heart disease), use of chronic anti-inflammatory medication, use of lipid-controlling medication, hypertriglyceridemia (>4 mmol/l) or hypercholesterolemia (LDL-C >5 mmol/l), anticipate major changes in lifestyle, smoker, heavy alcohol user (>3 drinks/day) and major surgery or has participated in an intervention trial in the last six months. This exclusion data is in line with previous meta-analyses and n-3 supplementation trials that have used the above exclusion criteria as they affect n-3 PUFA metabolism. Participants who sign the consent form who are later deemed to be ineligible due to BMI or % of DHA levels will be informed via email and/or phone. This is explained in the consent form.

Fatty acid analyses: Blood samples for baseline and follow-up meetings will be obtained following an overnight 12-hour fast. Blood samples will be collected at days 0, 3, 7, 14, 28 and 56. The samples will be collected by a registered nurse. Whole blood samples will be saved and stored from each participant. Then, the remaining whole blood samples from each participant will be centrifuged and the plasma and red blood cells (RBCs) isolated from each sample. They will be stored at -80°C in an air tight container until required for analysis. Plasma and RBCs samples will be analyzed for fatty acid concentrations using gas chromatography (GC)-flame ionization detection and delta carbon-13 (δ13C) will be assessed using GC-isotope ratio mass spectrometry. Plasma and RBCs will be analyzed with changes in plasma EPA levels as the primary outcome and plasma δ13C EPA signature, measured by compound specific isotope analysis (CSIA), as the secondary outcome. The δ13C signature of EPA will also be measured to see that it aligns with the δ13C signature of plasma ALA and not the DHA supplement.

Sample Size: Power was calculated based on results from the literature which has consistently reported changes in plasma EPA following DHA supplementation for 6 weeks with 12 subjects, and is consistent with our findings. Using means, standards deviations and sample sizes from the Metherel et al study in 2019 and these two previous studies, an alpha = 0.05 and desired power of 0.8 we determined sample sizes of just over 9 as sufficient to yield a statistically significant effect. As such, to account for a projected 20-25% dropout rate we will recruit 12 participants to ensure we are appropriately powered for our primary outcome.

Secondary analysis: Depending upon recruitment the investigators will examine if males and females respond differently in an exploratory analysis, recognizing they may not be powered to detect potentially real, but small sex-related differences. The investigators will use whole blood samples collected to determine FADS1, FADS2, ELOVL2 and ELOVL5 isoforms which have been shown to affect blood levels of EPA and DHA as a sub analysis.

The investigators will also provide data to do additional analysis related to n-3 PUFA turnover and loss rates, participants' ALA consumption and its effect on n-3 PUFA metabolism. In addition, some of the beneficial outcomes that may occur with DHA supplementation according to current research will be determined, including blood pressure, heart rate, triglycerides, and LDL, HDL and total cholesterol.

Recruitment: Participants will be recruited by way of email and social media to various vegetarian/vegan focused groups and organizations.

Anthropometric analyses: Height will be measured with a wall-mounted stadiometer.

Body weight will be assessed by beam scale. Waist circumference will be assessed using the Heart and Stroke Foundation methodology. Blood pressure (BP) and resting heart rate will be measured. To collect this measure, participants will remain seated in a quiet, temperature-controlled room for at least 5 minutes to achieve resting heart rate and BP. Subsequently, BP will be measured oscillometrically using the OMRON Intellisense HEM-907 according to JNC VII criteria. BP will be measured in triplicate, with each measurement separated by one minute, and the average of the three measures will be taken.

Biochemical analyses: Plasma and RBC samples for EPA and DHA will be separated by centrifuge and immediately frozen at -80 C at the University of Toronto for later analysis. Although plasma is the primary focus for the outcome, both plasma and RBC samples will be analyzed for comparison using GC-FID to determine EPA and DHA levels and δ13C will be analyzed using GC-IRMS.

Baseline and follow-up questionnaire: To obtain approximate ALA, EPA and DHA intake both before and after the supplementation period, a questionnaire will be provided to participants via email to fill out following the first phone/zoom meeting and at the final visit on day 56.

Statistical Analysis

Primary outcomes. Plasma and red blood cell (RBC) samples will be analyzed using GC-FID to determine n-3 PUFA concentrations within each group. Statistical analysis will be performed by a two-way ANOVA (dose x time) with repeated measures on time to determine interaction and/or main effects. In addition, the increase in EPA levels will be analyzed by segmented regression (or breakpoint analysis), as is done with the indicator amino acid oxidation technique between doses and within each time point to identify the dose of DHA supplementation that results in a sudden increase in EPA levels, the DHA dose that initiates feedback inhibition. Repeated measures post hoc Tukey test will be used to determine significant differences in DHA% and EPA% between visits as well as between groups. A Games-Howell test would be administered during the Tukey test if a group happens to have uneven participant numbers due to dropouts.

Secondary outcomes. δ13C will be analyzed using GC-IRMS. The δ13C signature of EPA will be measured to see that it aligns with the δ13C signature of plasma ALA and not the DHA supplement. Turnover rates and half-lives of EPA, DPAn-3 and DHA will be calculated by GraphPad Prism version 10.0 and the rate of loss of EPA, DPAn-3 and DHA in nmol/ml/day from the plasma will be calculated using the following formula: Jout = 0.693CFA/t1/2.

Exploratory and adherence outcomes. Repeated measures mixed effect models will be used to assess changes in all exploratory outcomes without controlling for false discovery rate. Pairwise comparisons between groups will be performed using Tukey-Kramer adjustment or other appropriate statistics. Effect modification by ALA consumption, sex, and genetics will be explored and blood samples will be analyzed for compliance.

Subgroup Analysis. A priori analysis will be conducted by age, sex, ethnicity, baseline BMI, baseline waist circumference, genetic variants and ALA consumption.