Omega-3 fatty acid, carotenoid and vitamin E supplementation improves working memory in older adults: A randomised clinical trial

Open AccessPublished:December 06, 2021DOI:https://doi.org/10.1016/j.clnu.2021.12.004

      Summary

      Background & aims

      Accumulating evidence suggests that omega-3 fatty acids (ω-3FAs), carotenoids and vitamin E can improve cognitive performance. However, their collective impact on cognition has not yet been investigated in healthy individuals. This study investigated the combined effect of ω-3FA, carotenoid and vitamin E supplementation on the cognitive performance of older adults.

      Methods

      Cognitively healthy individuals aged ≥65 years consumed daily 1 g fish oil (of which 430 mg docosahexaenoic acid, 90 mg eicosapentaenoic acid), 22 mg carotenoids (10 mg lutein, 10 mg meso-zeaxanthin, 2 mg zeaxanthin) and 15 mg vitamin E or placebo for 24 months in a double-blind, placebo-controlled, randomised clinical trial.

      Results

      Following 24-month supplementation, individuals in the active group (n = 30; aged 69.03 ± 4.41 years; 56.7% female) recorded significantly fewer errors in working memory tasks than individuals receiving placebo (n = 30; aged 69.77 ± 3.74 years; 70% female) (point estimate effect sizes ranged 0.090–0.105). Interestingly, as the cognitive load of the working memory tasks increased, the active group outperformed the placebo group. Statistically significant improvements in tissue carotenoid concentrations, serum xanthophyll carotenoid concentrations and plasma ω-3FA concentrations were also observed in the active group versus placebo (point estimate effect sizes ranged 0.078–0.589). Moreover, the magnitude of change of carotenoid concentrations in tissue, and ω-3FA and carotenoid concentrations in blood were related to the magnitude of change in working memory performance.

      Conclusion

      These results support a biologically plausible rationale whereby these nutrients work synergistically, and in a dose-dependent manner, to improve working memory in cognitively healthy older adults. Increasing nutritional intake of carotenoids and ω-3FAs may prove beneficial in reducing cognitive decline and dementia risk in later life.

      Study id number

      Keywords

      1. Introduction

      Due to the increasing prevalence of Alzheimer's disease (AD) and its associated economic, societal and caring burden, emphasis is now being placed on preventative strategies to delay its onset and reduce the risk of developing the disease. Accumulating evidence suggests that good nutrition (e.g. fruits, vegetables, fish) and healthy dietary patterns are important for improving cognitive performance [
      • Zwilling C.E.
      • Talukdar T.
      • Zamroziewicz M.K.
      • Barbey A.K.
      Nutrient biomarker patterns, cognitive function, and fMRI measures of network efficiency in the aging brain.
      ,
      • Loughrey D.G.
      • Lavecchia S.
      • Brennan S.
      • Lawlor B.A.
      • Kelly M.E.
      The impact of the mediterranean diet on the cognitive functioning of healthy older adults: a systematic review and meta-analysis.
      ], and are associated with a reduced risk of AD [
      • Jennings A.
      • Cunnane S.C.
      • Minihane A.M.
      Can nutrition support healthy cognitive ageing and reduce dementia risk?.
      ,
      • Solfrizzi V.
      • Custodero C.
      • Lozupone M.
      • Imbimbo B.P.
      • Valiani V.
      • Agosti P.
      • et al.
      Relationships of dietary patterns, foods, and micro- and macronutrients with Alzheimer's disease and late-life cognitive disorders: a systematic review.
      ,
      • Yuan C.
      • Chen H.
      • Wang Y.
      • Schneider J.A.
      • Willett W.C.
      • Morris M.C.
      Dietary carotenoids related to risk of incident Alzheimer dementia (AD) and brain AD neuropathology: a community-based cohort of older adults.
      ]. Importantly, advances in science and technology have increased our capacity to fully understand the unique neuroprotective mechanisms of specific nutrients that are likely driving these positive results. Some dietary components selectively accumulate in the brain where they play important physiological functions. These include omega-3 fatty acids (ω-3FAs) [
      • Singh M.
      Essential fatty acids, DHA and human brain.
      ,
      • Weiser M.J.
      • Butt C.M.
      • Mohajeri M.H.
      Docosahexaenoic acid and cognition throughout the lifespan.
      ], xanthophyll carotenoids (oxygen-containing, plant-based pigments) and vitamin E [
      • Craft N.E.
      • Haitema T.B.
      • Garnett K.M.
      • Fitch K.A.
      • Dorey C.K.
      Carotenoid, tocopherol, and retinol concentrations in elderly human brain.
      ,
      • Johnson E.J.
      • Vishwanathan R.
      • Johnson M.A.
      • Hausman D.B.
      • Davey A.
      • Scott T.M.
      • et al.
      Relationship between serum and brain carotenoids,-tocopherol, and retinol concentrations and cognitive performance in the oldest old from the Georgia Centenarian Study.
      ,
      • Lieblein-Boff J.C.
      • Johnson E.J.
      • Kennedy A.D.
      • Lai C.S.
      • Kuchan M.J.
      Exploratory metabolomic analyses reveal compounds correlated with lutein concentration in frontal cortex, Hippocampus, and occipital cortex of human infant brain.
      ]. Previous observational and interventional work that has separately examined the effects of these nutrients on cognitive function has yielded promising, yet mixed, results [
      • Johnson E.J.
      • Vishwanathan R.
      • Johnson M.A.
      • Hausman D.B.
      • Davey A.
      • Scott T.M.
      • et al.
      Relationship between serum and brain carotenoids,-tocopherol, and retinol concentrations and cognitive performance in the oldest old from the Georgia Centenarian Study.
      ,
      • Bowman G.L.
      • Silbert L.C.
      • Howieson D.
      • Dodge H.H.
      • Traber M.G.
      • Frei B.
      • et al.
      Nutrient biomarker patterns, cognitive function, and MRI measures of brain aging.
      ,
      • Kryscio R.J.
      • Abner E.L.
      • Caban-Holt A.
      • Lovell M.
      • Goodman P.
      • Darke A.K.
      • et al.
      Association of antioxidant supplement use and dementia in the prevention of Alzheimer's disease by vitamin E and selenium trial (PREADViSE).
      ]. Overall, the evidence to date suggests that these nutrients can work independently to improve cognitive performance, primarily due to their antioxidant and anti-inflammatory properties. Interestingly, previous exploratory work has shown that a combination of the ω-3FA docosahexaenoic acid (DHA) and the xanthophyll carotenoid lutein can work synergistically to improve cognition in older women [
      • Johnson E.J.
      • McDonald K.
      • Caldarella S.M.
      • Chung H.Y.
      • Troen A.M.
      • Snodderly D.M.
      Cognitive findings of an exploratory trial of docosahexaenoic acid and lutein supplementation in older women.
      ]. The present study, the Cognitive impAiRmEnt Study (CARES), was designed to examine the potential synergistic effects of a combination of ω-3FAs (namely DHA and eicosapentaenoic acid [EPA]), xanthophyll carotenoids (specifically lutein, zeaxanthin and meso-zeaxanthin) and vitamin E (d-α-tocopherol) on the cognitive performance of cognitively healthy older adults.

      2. Materials & methods

      2.1 Classification of evidence

      This study provides Class II evidence that 24-month supplementation with 430 mg DHA, 90 mg EPA, 10 mg lutein, 2 mg zeaxanthin, 10 mg meso-zeaxanthin and 15 mg vitamin E (d-α-tocopherol) is effective in improving cognitive performance, namely working memory, in cognitively healthy older adults.

      2.2 Study design and procedures

      CARES Trial 2 (Trial 1 published previously [
      • Power R.
      • Nolan J.M.
      • Prado-Cabrero A.
      • Coen R.
      • Roche W.
      • Power T.
      • et al.
      Targeted nutritional intervention for patients with mild cognitive impairment: the cognitive impAiRmEnt study (CARES) trial 1.
      ]) was a parallel group, double-blind, placebo-controlled, block-randomised clinical trial. Volunteers, primarily from the South-East catchment area of Ireland, were recruited through regional and national advertisement campaigns. Eligibility criteria included: age ≥65 years; no self or family collateral report of memory loss; no rapidly progressive or fluctuating symptoms of memory loss; no established diagnosis of early dementia; no consumption of cognitive enhancement therapies (e.g. cholinesterase inhibitors); no history of stroke disease; no depression (under active review); no psychiatric illness (under active review of psychotropic medications); no glaucoma (acute angle); not consuming carotenoid or fish/cod liver oil supplements; and no fish allergy.
      Prior to enrolment, all individuals that expressed an interest in participating in the trial completed a screening assessment (performed by RP) to confirm eligibility. This included assessing cognition using the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) and the Montreal Cognitive Assessment (MoCA). Individuals that fulfilled the criteria for each assessment were invited to participate in the trial. Individuals with borderline scores were referred to a consensus panel consisting of a Consultant Geriatrician, Psychiatrist of Old Age and Clinical Neuropsychologist for assessment of eligibility [
      • Albert M.S.
      • DeKosky S.T.
      • Dickson D.
      • Dubois B.
      • Feldman H.H.
      • Fox N.C.
      • et al.
      The diagnosis of mild cognitive impairment due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease.
      ,
      • Dubois B.
      • Albert M.L.
      Amnestic MCI or prodromal Alzheimer's disease?.
      ]. Eligible individuals were invited to enrol into the study (Fig. 1).
      Fig. 1
      Fig. 1Consolidated Standards of Reporting Trials flow diagram for CARES Trial 2.
      Of the 60 participants enrolled at baseline, 9 were lost at follow-up and 1 was excluded. Among participants in the active group, 2 were no longer interested in participating. Among participants in the placebo group, 2 were no longer interested in participating, 2 developed early-stage age-related macular degeneration (AMD) and 2 developed other health issues. One adverse event was recorded during the trial. One participant (female, aged 77 years at baseline) reported severe diarrhoea 4 weeks after commencing the trial. Of note, this participant was a survivor of cancer of the rectum. Upon trial completion, details of the intervention code revealed that this participant was enrolled into the placebo group. Thus, an attrition rate of 15% was recorded. Final visit (i.e. 24-month) data of 1 participant (male, aged 77 years at baseline) in the placebo group were removed prior to rANOVA analysis as meso-zeaxanthin was detected in 24-month (but not 12-month) serum (0.186 μmol/L). The presence of meso-zeaxanthin suggested carotenoid supplementation and was retrospectively confirmed via telephone with the participant.

      2.3 Randomisation and intervention

      Eligible individuals were assigned to the active or placebo group using block randomisation with no stratification. Random allocation sequencing in block sizes of 10 and in a 1:1 randomisation ratio was performed using a trial management system (Trial Controller) designed by our research centre and overseen by a Statistician (JS). In addition to completing the random allocation sequencing for the trial, the Trial Controller was used to document patient information (name, study code and contact details), assist with the scheduling of study visits and support the organisation and management of active and placebo capsules used in the trial. Comprehensive security and access controls in relation to the storage of the electronic data and the prevention of unauthorised access were implemented for this software.
      Capsule dispensing was performed by members (CK and LOB) of UPMC Whitfield Pharmacy, Waterford, Ireland. Using the Trial Controller system these individuals had access to patient study codes, assigned intervention group and capsule batch numbers. Importantly, pharmacy members had no contact with participants and no access to participant names or contact details. By comparison, researchers directly involved in CARES had access to participant details and study codes, but no access to information regarding intervention allocation or capsule batch numbers. The researcher (RP) received a box of tablets from the pharmacy members (CK or LOB) with a subject identification label (i.e. both the researcher and study participant were blinded to the intervention). The intervention code was only revealed at study completion.
      Participants were randomised to either the active intervention (n = 30) containing 1 g fish oil (of which 430 mg DHA and 90 mg EPA), 22 mg xanthophyll carotenoids (of which 10 mg lutein, 10 mg meso-zeaxanthin and 2 mg zeaxanthin) and 15 mg vitamin E (D-⍺-tocopherol) (now commercially known as Memory Health in the USA and reMIND in the UK and Europe) or placebo (sunflower oil) (n = 30) group for 24 months. Previous research has shown that the carotenoid formulation used in the present study is the most efficacious in terms of achieving a response in retinal tissue concentrations (i.e. in the macula lutea) [
      • Sabour-Pickett S.
      • Beatty S.
      • Connolly E.
      • Loughman J.
      • Stack J.
      • Howard A.
      • et al.
      Supplementation with three different macular carotenoid formulations in patients with early age-related macular degeneration.
      ,
      • Meagher K.A.
      • Thurnham D.I.
      • Beatty S.
      • Howard A.N.
      • Connolly E.
      • Cummins W.
      • et al.
      Serum response to supplemental macular carotenoids in subjects with and without age-related macular degeneration.
      ]. Both the discs of retinal photoreceptors [
      • Shindou H.
      • Koso H.
      • Sasaki J.
      • Nakanishi H.
      • Sagara H.
      • Nakagawa K.M.
      • et al.
      Docosahexaenoic acid preserves visual function by maintaining correct disc morphology in retinal photoreceptor cells.
      ] and the grey matter of the brain [
      • Skinner E.R.
      • Watt C.
      • Besson J.A.
      • Best P.V.
      Differences in the fatty acid composition of the grey and white matter of different regions of the brains of patients with Alzheimer's disease and control subjects.
      ] are enriched in phospholipids with DHA. In contrast, the presence of EPA in both visual and cognitive tissues is residual. Therefore, a fish oil formulation with the highest DHA content achievable to improve the DHA composition of these tissues was chosen. Fifteen milligrams of vitamin E was chosen as it is the maximum amount allowed by the European Food Safety Authority. Moreover, previous research has demonstrated a greater carotenoid response in blood when combined with fish oil [
      • Johnson E.J.
      • McDonald K.
      • Caldarella S.M.
      • Chung H.Y.
      • Troen A.M.
      • Snodderly D.M.
      Cognitive findings of an exploratory trial of docosahexaenoic acid and lutein supplementation in older women.
      ,
      • Nolan J.M.
      • Mulcahy R.
      • Power R.
      • Moran R.
      • Howard A.N.
      Nutritional intervention to prevent Alzheimer's disease: potential benefits of xanthophyll carotenoids and omega-3 fatty acids combined.
      ]. Doses were provided via two oval-sized capsules. Active and placebo capsules were identical in colour and size. Each active capsule contained equal quantities of fish oil, carotenoids and vitamin E. Carotenoid and vitamin E concentrations were manufactured by Industrial Orgánica (Monterrey, Mexico), while fish oil concentrations were manufactured by Epax (Ålesund, Norway; product number: EPAX1050 TG/N non-tuna). Participants were instructed to consume two capsules per day and in one sitting with a meal. Frequent phone calls were made to ensure compliance. Tablet counting was also performed at each follow-up visit to determine the overall level of compliance for both active and placebo groups. For each participant, the total number of capsules remaining at the end of the trial (i.e. the amount of capsules remaining after 12 months plus the amount remaining after 24 months) was divided by the total number of capsules issued for the trial. From this, a percentage was calculated. Study visits occurred at baseline, 12- and 24-months at a single site (Nutrition Research Centre Ireland). The trial commenced in March 2016 and concluded in June 2019 (i.e. last 24-month subject visit).

      2.4 Standard protocol approvals, registrations and patient consents

      Written informed consent was obtained from all participants prior to enrolment. Ethical approval was granted by the Waterford Institute of Technology and University Hospital Waterford research ethics committees in Waterford, Ireland, in December 2015. CARES (trial registration number: ISRCTN10431469) adhered to the tenets of the Helsinki Declaration (as revised in 2013) and followed the full code of ethics with respect to recruitment, testing and general data protection regulations as set out by the European Parliament and Council of the European Union.

      2.5 Sample size calculations and outcome measures

      Sample sizes of 30 per group were determined from power analysis to be suitable in this study. Subjects were randomly allocated between the active and placebo intervention groups and a 5% level of significance was chosen (i.e. a 95% confidence level). Calculations were based on repeated measures analysis of variance (rANOVA) analysis between two time points (i.e. baseline and end of study). All tests were assumed to be two-sided. RBANS across all five cognitive domains (i.e. RBANS total scale score) was the primary outcome measure for CARES. As all RBANS domains were considered to be of equal significance, the average scores of the 5 domains were used for power analysis. Based on data provided from baseline, the mean RBANS score was 106 and mean standard deviation (SD) was 12. Assuming a correlation of 0.70 for within-subject RBANS scores between baseline and end of study, a statistical power of approximately 96% was estimated for an effect size of 10.60 (10% of baseline RBANS score) and 79% for an effect size of 7.95 (7.5% of RBANS score). Secondary outcome measures included change in the following variables: working memory, attention, episodic memory, macular pigment optical volume (MPOV), skin carotenoid score (SCS), plasma ω-3FA concentrations, and serum concentrations of xanthophyll carotenoids and vitamin E.

      2.6 Measurements

      2.6.1 Cognitive function

      Global cognition was assessed using the MoCA version 7.1 [
      • Nasreddine Z.S.
      • Phillips N.A.
      • Bedirian V.
      • Charbonneau S.
      • Whitehead V.
      • Collin I.
      • et al.
      The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment.
      ] and the RBANS Record form A [
      • Randolph C.
      • Tierney M.C.
      • Mohr E.
      • Chase T.N.
      The repeatable battery for the assessment of neuropsychological status (RBANS): preliminary clinical validity.
      ] at screening and at 12- and 24-month follow-up (performed by RP). The MoCA is a short (10-min) 30-item cognitive screening questionnaire used to detect cognitive impairment. It assesses multiple cognitive domains including visuospatial abilities, executive function, phonemic fluency, attention, immediate and delayed recall, language and orientation. From this a composite score is generated. A score ≥26 out of 30 was desirable for enrolment. The RBANS is a core diagnostic tool for detecting cognitive decline or improvement. It takes approximately 30 min to administer and assesses immediate memory, visuospatial ability, language, attention and delayed memory using 12 sub-tests. Scores from each domain are summed to determine a total index/scale score. The RBANS yields index standard scores that are based on the raw scores of each subtest. RBANS index scores are metrically scaled, with a mean of 100 and a SD of 15 for each age group. A score of 100 on any of these measures equates to the average performance of individuals of similar age. Scores of 85 and 115 correspond to 1 SD below and above the mean, respectively, while scores of 70 and 130 are 2 SDs below and above the mean. Approximately 68% of all examinees score between 85 and 115 and circa 95% score in the 70 to 130 range [
      • Randolph C.
      RBANS Update : Repeatable Battery for the Assessment of Neuropsychological Status Manual. U.S.A.: Bloomington, Minn.
      ]. In the present study, a score of ≥78 was desirable for enrolment as it is above the defined cut-off score (1.5 SD below the mean) for cognitive impairment based on RBANS population-based norms.
      Additional assessments of specific cognitive domains were performed using the Cambridge neuropsychological test automated battery (CANTAB) Connect Research software (Cambridge Cognition, Cambridge, UK) [
      • Cognition Cambridge
      CANTAB Connect Research: Admin Application User Guide v1.6.
      ]. This computerised software program was performed on an iPad and required a finger-operated response. The CANTAB battery included the motor control task (MOT) to determine comprehension, the spatial working memory task (SWM) to measure working memory, the reaction time task (RTI) to assess attention and the paired associated learning task (PAL) to assess episodic memory [
      • Cognition Cambridge
      Product Overview: CANTAB Connect Research v11.10.
      ]. A description of each cognitive task and associated outcomes measures is provided in Supplementary eTable 1.

      2.6.2 Tissue carotenoid concentrations

      2.6.2.1 Macular pigment

      The xanthophyll carotenoids lutein, zeaxanthin and meso-zeaxanthin selectively accumulate in the central retina where they are collectively referred to as macular pigment (MP). Given that retinal concentrations (i.e. MP) correlate with brain concentrations of lutein and zeaxanthin [
      • Vishwanathan R.
      • Schalch W.
      • Johnson E.J.
      Macular pigment carotenoids in the retina and occipital cortex are related in humans.
      ], and higher MP levels are associated with better cognitive performance [
      • Feeney J.
      • Finucane C.
      • Savva G.M.
      • Cronin H.
      • Beatty S.
      • Nolan J.M.
      • et al.
      Low macular pigment optical density is associated with lower cognitive performance in a large, population-based sample of older adults.
      ,
      • Ajana S.
      • Weber D.
      • Helmer C.
      • Merle B.M.
      • Stuetz W.
      • Dartigues J.F.
      • et al.
      Plasma concentrations of lutein and zeaxanthin, macular pigment optical density, and their associations with cognitive performances among older adults.
      ], MP can be used as a non-invasive biomarker of brain nutrition and cognitive health. MP was measured by dual wavelength autofluorescence (AF) using the Spectralis HRA+OCT MultiColor (Heidelberg Engineering GmbH, Heidelberg, Germany). Pupillary dilation of one eye (using a drop of 0.5% proxymetacaine hydrochloride followed by a drop of 1% tropicamide) was performed prior to measurement and patient details were entered into the Heidelberg Eye Explorer (HEYEX version 1.7.1.0) software. Dual-wavelength AF in this device uses two excitation wavelengths; one that is well absorbed by MP (486 nm, blue) and one that is not (518 nm, green) [
      • Trieschmann M.
      • Heimes B.
      • Hense H.W.
      • Pauleikhoff D.
      Macular pigment optical density measurement in autofluorescence imaging: comparison of one- and two-wavelength methods.
      ]. The following acquisition parameters were used: high speed scan resolution, 2 s cyclic buffer size, internal fixation, 30-s movie and manual brightness control. Alignment, focus and illumination were first adjusted in infrared mode. Once the image was evenly illuminated, the laser mode was switched from infrared to blue plus green laser light AF. Using the HEYEX software, the movie images were aligned and averaged, and a MP density map was created. MPOV calculated as MP average times the area under the curve out to 7° eccentricity is reported here [
      • Roche W.
      • Green-Gomez M.
      • Moran R.
      • Nolan J.
      The physics of using the Heidelberg Spectralis dual-wavelength autofluorescence method for the measurement of macular pigment volume [Abstract].
      ] and has been previously validated as an accurate and reliable assessment of MP [
      • Green-Gomez M.
      • Bernstein P.S.
      • Curcio C.A.
      • Moran R.
      • Roche W.
      • Nolan J.
      Standardizing the assessment of macular pigment using a dual-wavelength Autofluorescence technique.
      ]. A higher MPOV score was indicative of greater MP.

      2.6.2.2 Skin carotenoid score

      Carotenoid concentrations were also measured using the Pharmanex® BioPhotonic Scanner (Salt Lake City, UT, USA). This scanner measures carotenoid levels in human tissue at the skin surface using optical signals (resonant Raman spectroscopy) [
      • Lademann J.
      • Meinke M.C.
      • Sterry W.
      • Darvin M.E.
      Carotenoids in human skin.
      ]. These signals identify the unique molecular structure of carotenoids, allowing their measurement without interference by other molecular substances. Participants placed a specific point (between the maximal and distal palmar creases, directly below the fifth finger) of their right hand (previously cleaned with hand sanitizer) in front of the scanner's low-energy blue light for 30 s. Following this, a SCS was generated, which provided an indication of the participants' overall carotenoid levels (ranging from zero to 90,000). A higher score was indicative of greater carotenoid intake. This technology has been previously validated for its safety and accuracy in measuring carotenoid status [
      • Zidichouski J.A.
      • Mastaloudis A.
      • Poole S.J.
      • Reading J.C.
      • Smidt C.R.
      Clinical validation of a noninvasive, Raman spectroscopic method to assess carotenoid nutritional status in humans.
      ,
      • Janse VAN Rensburg A.
      • Wenhold F.
      Validity and reliability of field resonance Raman spectroscopy for assessing carotenoid status.
      ].

      2.6.3 Biochemical analysis

      Non-fasting blood samples were collected at each study visit by standard venepuncture techniques. We used the same methodology as the one employed in a previous study [
      • Power R.
      • Nolan J.M.
      • Prado-Cabrero A.
      • Coen R.
      • Roche W.
      • Power T.
      • et al.
      Targeted nutritional intervention for patients with mild cognitive impairment: the cognitive impAiRmEnt study (CARES) trial 1.
      ] to extract lutein, zeaxanthin, meso-zeaxanthin and d-⍺-tocopherol from serum, and DHA and EPA from plasma; to quantify serum carotenoid and d-⍺-tocopherol concentrations and analyse amounts using high performance liquid chromatography (HPLC); and to quantify plasma concentrations of DHA and EPA and analyse amounts by gas chromatography coupled to flame ionisation detector (GC-FID).

      2.6.4 Demographic, health and lifestyle data

      Demographic, health and lifestyle data, medical history and current medication use were recorded via questionnaire. Height and weight measurements were recorded to calculate body mass index (kg/m2). Colour fundus photographs were taken to assess the presence of ocular pathology (Zeiss Visucam 200, Carl Zeiss Meditec AG, Jena, Germany).

      2.7 Statistical analysis

      The statistical packages IBM SPSS version 25 and Minitab 19.2 were used, and the 5% significance level applied for all analyses. An adjustment for multiple comparisons was not carried out for this study as initial sample sizes were determined according to a 5% level of significance and 80% power for the primary outcome measure and pre-planned comparisons. For all analyses, point estimates and 95% confidence intervals were provided. Results were expressed as means ± SD for numeric data. Categorical data were expressed as percentages. Between-group differences were analysed using Independent Samples t-tests or Chi-square tests, as appropriate. rANOVA was used to assess Time and Time–Group interaction effects across 3 time points (i.e. baseline, 12- and 24-month follow-up) between both intervention groups for cognition and nutrition variables. Time effects examine whether or not a response variable is different at the time points of interest. Time–Group effects examine whether or not the time effect differs between the active and placebo groups. In cases where rANOVA showed interesting results, further statistical analyses were conducted using paired samples t-tests to examine statistical difference within groups at baseline and 24 months for both groups. A general linear model was used to assess (for dependent variables tissue carotenoid concentrations, serum carotenoid and plasma ω-3FA concentrations) the potential impact of sex, smoking habits, and alcohol consumption on Time and Time–Group effects. Effect size interpretations were based on parameters set by Cohen in 1988 [
      • Lakens D.
      Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs.
      ,
      • Cohen J.
      Statistical power analysis for the behavioral sciences.
      ] (i.e. 0.01, 0.06 and 0.14 for small, medium and large effect sizes, respectively for rANOVA analysis, and 0.20, 0.50 and 0.80 for small, medium and large effect sizes, respectively for paired samples t-test analysis. Finally, Spearman's rank correlation coefficient was used to investigate potential relationships between the observed changes in cognitive function variables and the observed changes in tissue carotenoid concentrations, serum carotenoid concentrations, and plasma ω-3FA concentrations. Effect size interpretations for Spearman's rank correlation coefficient were based on parameters set by Cohen in 1988 [
      • Cohen J.
      Statistical power analysis for the behavioral sciences.
      ] (i.e. 0.20, 0.50, and 0.80 for small, medium and large effect sizes, respectively).

      2.8 Data availability

      Research protocols and anonymised data from CARES may be shared by written request from any qualified researchers for the purpose of replicating procedures and results.

      3. Results

      3.1 Baseline data

      Demographic, health and lifestyle, nutrition and cognitive function data were statistically comparable between active and placebo groups at baseline (Table 1A, Table 1BA and 1B ).
      Table 1ABaseline demographic, health and lifestyle data of active and placebo intervention groups.
      VariableActive (n = 30)Placebo (n = 30)
      Demographic data
       Age (years)69.03 ± 4.4169.77 ± 3.74
       Sex ([n] [% female])17 (56.7%)21 (70.0%)
       Education (years)16.47 ± 1.6117.41 ± 2.69
      Health and lifestyle data
       Medications3.07 ± 2.553.27 ± 2.97
       Exercise (min/week)288.17 ± 308.23313.67 ± 282.12
       Smoking ([n]; [%])
      Never12 (40.0%)20 (66.7%)
      Past15 (50.0%)9 (30.0%)
      Current3 (10.0%)1 (3.3%)
       BMI (kg/m2)28.92 ± 5.1027.11 ± 4.18
      Nutrition data
       MP (MPOV)5325 ± 22065575 ± 2002
       SCS30,593 ± 831732,167 ± 11,354
       Serum L0.158 ± 0.0600.203 ± 0.156
       Serum Z0.052 ± 0.0140.059 ± 0.027
       Serum MZ00
       Serum vitamin E29.272 ± 5.57628.280 ± 5.693
       Plasma DHA191.692 ± 82.881214.406 ± 51.511
       Plasma EPA128.572 ± 68.841127.327 ± 39.650
      Data are presented as are mean ± standard deviation for numeric data and actual number and percentages for categorical data; Education: age (years) completed formal education; Medications: the number of prescribed medications consumed; Smoking status: never (smoked <100 cigarettes in lifetime), past smoker (smoked ≥100 cigarettes in lifetime and none in the past year) or current smoker (smoked ≥100 cigarettes in lifetime and at least 1 cigarette in the last year). BMI: body mass index; MP: macular pigment; MPOV: macular pigment optical volume, calculated as MP average times the area under the curve out to 7° eccentricity; SCS: skin carotenoid score; Serum lutein, zeaxanthin, meso-zeaxanthin, and plasma docosahexaenoic acid and eicosapentaenoic acid concentrations are expressed in μmol/L. Data missing in the active intervention group for the following variables: education (n = 1); MPOV (n = 1); SCS (n = 3); serum xanthophyll carotenoid and vitamin E concentrations (n = 4); plasma DHA and EPA (n = 3). Data missing in the placebo group for the following variables: serum xanthophyll carotenoid and vitamin E concentrations (n = 1).
      Table 1BBaseline cognition data of active and placebo intervention groups.
      VariableActive (n = 30)Placebo (n = 30)
      Global cognition
       MoCA27.53 ± 1.7627.03 ± 1.16
       RBANS immediate memory108.37 ± 13.30107.0 ± 13.38
       RBANS visuospatial112.73 ± 13.20115.40 ± 14.16
       RBANS language102.30 ± 10.28101.40 ± 9.18
       RBANS attention100.83 ± 9.7599.30 ± 15.24
       RBANS delayed memory107.30 ± 9.91106.13 ± 10.94
       RBANS total scale107.70 ± 11.41108.03 ± 11.34
      Comprehension (MOT)
      Latency (millisecond)941.81 ± 219.171013.22 ± 234.54
      Working memory (SWM)
      Between errors stage 40.70 ± 1.471.13 ± 1.81
      Between errors stage 63.87 ± 3.565.27 ± 3.77
      Between errors stage 812.0 ± 4.8111.07 ± 3.18
      Between errors all stages16.07 ± 7.6716.63 ± 6.67
      Total errors stage 40.87 ± 2.271.20 ± 2.09
      Total errors stage 64.0 ± 3.635.57 ± 4.20
      Total errors stage 812.50 ± 5.2311.28 ± 3.13
      Total errors all stages16.86 ± 8.0217.07 ± 6.68
      Strategy8.45 ± 2.289.27 ± 2.29
      Reaction time (RTI)
       Simple reaction time (millisecond)371.62 ± 52.74371.52 ± 46.22
       Simple error score (millisecond)0.97 ± 1.430.93 ± 1.02
       5-choice reaction time (millisecond)416.89 ± 45.37425.91 ± 48.61
       5-choice error score (millisecond)0.23 ± 0.570.60 ± 1.0
      Episodic memory (PAL)
       First attempt memory score10.40 ± 3.8810.07 ± 3.26
       Total errors adjusted stage 20.27 ± 0.690.13 ± 0.51
       Total errors adjusted stage 41.33 ± 2.261.87 ± 2.53
       Total errors adjusted stage 66.77 ± 5.486.80 ± 4.38
       Total errors adjusted stage 817.0 ± 9.5614.90 ± 9.38
       Total errors adjusted all stages25.37 ± 15.4223.70 ± 13.05
      Data are presented as are mean ± standard deviation. MoCA: Montreal Cognitive Assessment; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status; MOT: motor control task; SWM: spatial working memory; RTI: Reaction time; PAL: paired associated learning. Data missing in the active intervention group for the following variables: SWM between errors stage 8 score (n = 2); SWM between errors total score (n = 1); SWM total errors stage 8 score (n = 2); SWM total errors across all stages (n = 1); SWM strategy score (n = 1). Data missing in the placebo group for the following variables: SWM between errors stage 8 score (n = 1); SWM total errors stage 8 score (n = 1).

      3.2 Level of compliance

      On average, the level of compliance to the intervention was 87% among individuals in the active group (n = 27) and 91% in the placebo group (n = 21). Level of compliance was statistically comparable between both groups.

      3.3 Observed change in cognitive function

      Table 2 summarises the observed change over time in cognitive function and nutrition variables for active and placebo intervention groups, based on rANOVA analysis.
      Table 2Observed changes over time in cognitive function and nutrition variables for active and placebo intervention groups using repeated measures analysis of variance.
      VariableBaseline12 months24 months
      ActivePlaceboActivePlaceboActivePlacebo
      nM ± SDnM ± SDM ± SD%Δ; OutcomeM ± SD%Δ; OutcomeM ± SD%Δ; OutcomeM ± SD%Δ; Outcomeη2 (CI)
      Cognition
      SWM TE82211.32 ± 4.671711.06 ± 3.2511.45 ± 5.31+1; Declined11.76 ± 4.37+6; Declined7.05 ± 4.13−38; Improved11.18 ± 4.83+1; Declined0.090 (0.005, 0.189)
      SWM TE all stages2216.77 ± 7.831814.33 ± 6.3716.59 ± 7.91−1; Improved16.06 ± 6.85+12; Declined12.45 ± 6.06−26; Improved16.39 ± 5.38+14; Declined0.105 (0.012, 0.207)
      RBANS immediate28109.64 ± 12.3721106.57 ± 14.89110.46 ± 13.11+1; Improved110.67 ± 11.66+4; Improved113.07 ± 12.23+3; Improved119.10 ± 10.66+12; Improved0.083 (0.008, 0.169)
      RBANS language28102.86 ± 10.2721101.33 ± 9.42103.11 ± 9.80+0.2; Improved105.14 ± 9.22+4; Improved110.82 ± 10.82+8; Improved105.52 ± 10.56+4; Improved0.094 (0.014, 0.184)
      Nutrition
      MPOV265154 ± 2221215399 ± 16687338 ± 2704+42; Improved5403 ± 1847+0.01, Improved8505 ± 2972+65; Improved5063 ± 1808−6; Declined0.589 (0.472, 0.660)
      SCS2430,458 ± 85522033,750 ± 12,61541,125 ± 11,468+35; Improved32,250 ± 11,170−4; Declined38,542 ± 12,420+27; Improved33,650 ± 12,861−0.3; Declined0.253 (0.118, 0.361)
      Lutein230.157 ± 0.064190.207 ± 0.1900.689 ± 0.346+339; Improved0.204 ± 0.153−1; Declined0.550 ± 0.361+250; Improved0.218 ± 0.146+5; Improved0.392 (0.245, 0.494)
      Zeaxanthin230.051 ± 0.014190.064 ± 0.0310.085 ± 0.035+67; Improved0.068 ± 0.042+6; Declined0.075 ± 0.033+47; Improved0.069 ± 0.032+8; Improved0.167 (0.050, 0.276)
      MZ2301900.052 ± 0.032-; Improved00; Unchanged0.035 ± 0.031-; Improved00; Unchanged0.420 (0.274, 0.519)
      Vitamin E2329.060 ± 5.7151928.646 ± 5.91230.251 ± 5.557+4; Improved29.922 ± 6.810+4; Improved28.803 ± 5.399−1; Disimproved30.231 ± 7.217+6; Improved0.024 (0, 0.085)
      DHA24190.991 ± 85.89419207.415 ± 50.085304.303 ± 95.382+59; Improved204.695 ± 61.975−1; Declined319.740 ± 111.854+67; Improved227.305 ± 58.274+10; Improved0.256 (0.120, 0.366)
      EPA24125.704 ± 67.67919116.687 ± 32.506142.532 ± 54.126+13, Improved105.883 ± 41.300−9; Dieclined166.272 ± 77.310+32; Improved118.095 ± 36.092+1; Improved0.078 (0.003, 0.169)
      Data are presented as are mean ± standard deviation; %Δ at 12 months: 12-month visit minus baseline visit expressed as a percentage; %Δ at 24 months: 24-month visit minus baseline visit expressed as a percentage; Outcome: interpretation of direction of result (i.e. improved, declined or remained unchanged over time); η2: effect size; CI: 90% confidence interval (lower limit, upper limit); SWM TE8: spatial working memory total errors at stage 8, the number of times a box is selected that is certain not to contain a token and therefore should not have been visited by the individual at stage 8 of the assessment; SWM TE all stages: spatial working memory total errors across all stages, the number of times a box is selected that is certain not to contain a token and therefore should not have been visited by the individual, calculated across all stages of the assessment; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status; RBANS immediate: immediate memory domain of the RBANS; MPOV: macular pigment optical volume; SCS: skin carotenoid score; MZ: meso-zeaxanthin; DHA: docosahexaenoic acid; EPA: eicosapentaenoic acid.

      3.3.1 Global cognition

      The RBANS total scale score (i.e. the primary outcome measure) improved in both groups after 24 months (+3% versus +6% for active and placebo groups, respectively). The Time effect was statistically significant (η2 = 0.135, 90% confidence interval [CI] [0.037, 0.232]) however, no Time–Group effect was noted (η2 = 0.054, 90% CI [0, 0.131]). Individuals in the active group performed better over time in comparison to individuals in the placebo group in the RBANS language domain. A statistically significant Time–Group interaction effect was also observed for the RBANS immediate memory domain, but this improvement was seen in the placebo group (Table 2).

      3.3.2 Working memory

      Medium to large Time and Time–Group effect sizes were recorded by individuals receiving the active intervention for working memory tasks, with small effect sizes noted among individuals in the placebo group. Following the 24-month intervention period, individuals in the active group significantly reduced the number of total errors made at stage 8 of the SWM task by 38%, while individuals receiving placebo declined by 1% after 24 months (η2 = 0.090, 90% CI [0.005,0.189]). Additionally, individuals consuming the active intervention recorded 26% fewer total errors post intervention in comparison to individuals receiving placebo where the number of errors increased by 14% after 24 months (η2 = 0.105, 90% CI [0.012, 0.207]) (see Fig. 2A and B, and Table 2).
      Fig. 2
      Fig. 2Line graphs illustrating change in spatial working memory errors over 24 months.
      Errors bars: +/− 1 standard error. Lower score is indicative of better performance. Spatial working memory total errors: the number of times a box is selected that is certain not to contain a token and therefore should not have been visited by the individual. Stage 8 involves 8 boxes, and all stages is the sum of stages 4, 6 and 8. Statistical significance was observed between active and placebo groups for both SWM total errors at stage 8 and SWM total errors across all stages.
      Of note, the number of total errors made at stage 4 of the SWM task were statistically comparable (η2 = 0.005, 90% CI [0, 0.032]) for active (mean ± SD baseline 0.93 ± 2.34; final visit 0.86 ± 1.46) and placebo (baseline 0.80 ± 2.29; final visit 0.30 ± 0.98) groups. Interestingly, as the cognitive load of the task increased (i.e. from 4 to 6 tokens, and from 6 to 8 tokens), individuals in the active intervention (stage 6: baseline 4.50 ± 3.67, final visit 2.27 ± 2.27; stage 8: baseline 11.32 ± 4.67, final visit 7.05 ± 4.13) outperformed individuals receiving placebo (stage 6: baseline 4.26 ± 4.27, final visit 4.74 ± 3.05; stage 8: baseline 11.06 ± 3.25, final visit 11.18 ± 4.83), with η2 = 0.059, 90% CI (0, −0.141) for stage 6 and η2 = 0.090, 90% CI (0.005, 0.189) for stage 8.

      3.3.3 Post hoc analysis

      As outlined previously, paired samples t-tests were conducted in cases where rANOVA showed interesting results (see Supplementary eTable 2). Medium to large effect sizes were observed over time among individuals in the active intervention group for working memory (between errors at stage 8 and total errors at stage 6) and reaction time, with small effect sizes recorded among individuals consuming placebo. The nutritional intervention had a small effect on the RBANS composite score (d = −0.282, 95% CI [−0.573, −0.002]). In contrast, a large effect size was noted in the placebo group for this global cognition assessment (d = −0.578, 95% CI [−0.914, −0.274]).

      3.4 Observed change in nutritional status

      Large Time–Group effect sizes were observed for individuals receiving the active intervention for carotenoid concentrations in tissue, with mean percentage increases of +65 and +27 after 24 months recorded for MPOV and SCS variables, respectively. Medium to large Time–Group effect sizes were recorded for serum concentrations of lutein, zeaxanthin and meso-zeaxanthin (mean percentage increases of +250 for lutein and +47 for zeaxanthin) and plasma concentrations of DHA and EPA (mean percentage increases of +67 and +32, respectively) after the 24-month intervention period. There was no evidence to suggest statistical significance for a Time or Time–Group effect for serum d-α-tocopherol concentrations in either group (Table 2). Of note, the observed increases in blood concentrations of xanthophyll carotenoids and ω-3FAs were all independently related to the observed increases in tissue carotenoid concentrations (MPOV and SCS), with the exception of EPA and SCS (Supplementary eTable 3).

      3.5 Effects of demographic and other lifestyle variables

      The possibility of an interaction effect for sex, education, BMI, smoking status and alcohol consumption on the statistically significant Time–Group effects observed for nutrition variables was examined using a general linear model. The dependent variables in these analyses included change in: MPOV; SCS; serum concentrations of lutein, zeaxanthin, meso-zeaxanthin and d-α-tocopherol; and plasma DHA and EPA concentrations. No statistically significant interactions were found. Thus, for example, changes in MPOV did not differ by sex, BMI, smoking status nor alcohol consumption.

      3.6 Relationships between change in nutrition status and change in cognitive function

      Spearman's rank correlation coefficient was used to investigate whether or not the observed changes in cognitive scores were related to the observed changes in tissue and serum concentrations of xanthophyll carotenoids and plasma concentrations of ω-3FAs (Table 3). Overall, medium to large-strength relationships were recorded, with the exception of relationships between reaction time and some nutritional variables. Individuals with higher concentrations of MPOV, lutein, meso-zeaxanthin, DHA or EPA after 24 months recorded fewer errors in the working memory task in comparison to individuals with lower changes in serum and tissue concentrations of xanthophyll carotenoids and lower changes in plasma ω-3FA concentrations.
      Table 3Relationships between observed changes in nutritional status and observed changes in cognitive function using Spearman's rank correlation coefficient.
      Observed change in nutritional statusObserved change in SWM total errors at stage 8Observed change in SWM total errors across all stagesObserved change in simple reaction time
      rCInrCInrCIn
      MPOV−0.452−0.699, −0.16445−0.458−0.671, −0.17545−0.353−0.585, −0.06848
      Serum Lutein−0.375−0.627, −0.05138−0.332−0.592, −0.01039−0.209−0.482, 0.10043
      Serum meso-zeaxanthin−0.388−0.637, −0.06638−0.352−0.607, −0.03139−0.220−0.491, 0.09043
      Plasma DHA−0.446−0.679, −0.13138−0.408−0.649, −0.09339−0.156−0.438, 0.15343
      Plasma EPA−0.310−0.578, 0.01838−0.317−0.580, 0.00739−0.365−0.606, −0.06343
      Observed change: exit visit data minus baseline visit data; CI: 95% confidence interval (lower limit, upper limit); MPOV: macular pigment optical volume; DHA: docosahexaenoic acid; EPA: eicosapentaenoic acid; Total errors: the number of times a box is selected that is certain not to contain a token and therefore should not have been visited by the individual; Simple reaction time: the duration between the onset of the stimulus and the time at which the individual released the button. Calculated for correct trials, where the stimulus could appear in one location only.

      4. Discussion

      4.1 Summary of findings

      Following 24-month supplementation, individuals in the active intervention exhibited improvements in working memory. Improvements in attention and global cognition were also recorded. The observed improvements in cognition are consistent with previous observational [
      • Bowman G.L.
      • Silbert L.C.
      • Howieson D.
      • Dodge H.H.
      • Traber M.G.
      • Frei B.
      • et al.
      Nutrient biomarker patterns, cognitive function, and MRI measures of brain aging.
      ,
      • Ajana S.
      • Weber D.
      • Helmer C.
      • Merle B.M.
      • Stuetz W.
      • Dartigues J.F.
      • et al.
      Plasma concentrations of lutein and zeaxanthin, macular pigment optical density, and their associations with cognitive performances among older adults.
      ,
      • Ubeda N.
      • Achon M.
      • Varela-Moreiras G.
      Omega 3 fatty acids in the elderly.
      ] and interventional [
      • Kulzow N.
      • Witte A.V.
      • Kerti L.
      • Grittner U.
      • Schuchardt J.P.
      • Hahn A.
      • et al.
      Impact of omega-3 fatty acid supplementation on memory functions in healthy older adults.
      ,
      • Power R.
      • Coen R.F.
      • Beatty S.
      • Mulcahy R.
      • Moran R.
      • Stack J.
      • et al.
      Supplemental retinal carotenoids enhance memory in healthy individuals with low levels of macular pigment in A randomized, double-blind, placebo-controlled clinical trial.
      ] studies.

      4.2 Working memory

      With specific reference to working memory, individuals in the active group made significantly fewer errors in the final and combined stages of the SWM task in comparison to individuals receiving placebo. Of note, the observed changes in tissue and serum carotenoid concentrations, and in plasma ω-3FAs concentrations were directly related to the observed improvements in this working memory task. Previous studies have also reported a relationship between higher carotenoid and ω-3FA intake and better executive function [
      • Johnson E.J.
      • Vishwanathan R.
      • Johnson M.A.
      • Hausman D.B.
      • Davey A.
      • Scott T.M.
      • et al.
      Relationship between serum and brain carotenoids,-tocopherol, and retinol concentrations and cognitive performance in the oldest old from the Georgia Centenarian Study.
      ,
      • Feeney J.
      • O'Leary N.
      • Moran R.
      • O'Halloran A.M.
      • Nolan J.M.
      • Beatty S.
      • et al.
      Plasma lutein and zeaxanthin are associated with better cognitive function across multiple domains in a large population-based sample of older adults: findings from the Irish longitudinal study on aging.
      ,
      • Dangour A.D.
      • Allen E.
      • Elbourne D.
      • Fletcher A.
      • Richards M.
      • Uauy R.
      Fish consumption and cognitive function among older people in the UK: baseline data from the OPAL study.
      ]. Working memory is a key component of executive function that is responsible for the temporary holding of information for later access and application (e.g. holding a person's telephone number or address in mind, listening and responding to information spoken during a conversation). Brain regions involved in working memory include the prefrontal cortex, parietal regions, and the hippocampus. More specifically, working memory involves the encoding of stimuli (e.g. words, pictures) and can involve attending to just one feature of a stimulus (i.e. selective attention) (e.g. tuning out the various sounds in a busy restaurant to listen to your friend tell a story), or attending to multiple features of a stimulus (i.e. divided attention or multi-tasking) (e.g. singing along to a song on the radio while driving). While the span of our working memory is quite short (10–15 s) [
      • Goldstein E.B.
      Cognitive Psychology: Connecting Mind, Research and Everyday Experience.
      ], it is vital for learning, retaining and responding to information. In the present study, the encoding and retrieval of information was comparable between active and placebo groups during a working memory task with few stimuli (i.e. stage 4 of the SWM task where the individual had to locate 4 tokens). Importantly, as the cognitive load increased (i.e. from 4 to 6 tokens, and from 6 to 8 tokens) individuals in the active intervention out-performed individuals in the placebo group, with better performance in stage 8 and summed stages where the cognitive load was at its highest. This suggests that the working memory capacity of individuals in the active group was favourably altered over time and that these positive changes may be attributed to the enrichment of ω-3FAs and carotenoids, given that the magnitude of change in cognition was related to the magnitude of change in nutrition levels and given that these nutrients have been previously shown to be neuroprotective [
      • Power R.
      • Prado-Cabrero A.
      • Mulcahy R.
      • Howard A.
      • Nolan J.M.
      The role of nutrition for the aging population: implications for cognition and Alzheimer's disease.
      ]. In terms of clinical significance, the observed improvements in working memory can translate into practical benefits for day-to-day function. An improved working memory can enhance the capacity to retain information and prioritise the steps needed to make decisions and solve problems. Enhancing working memory can also aid individuals in focusing on the task at hand such as planning and prioritising tasks for the day ahead or remembering key information (e.g. keeping appointment).

      4.3 Carotenoid and omega-3 fatty acid synergy

      An additional and interesting findings from this work relates to the positive and significant relationships observed between blood concentrations of ω-3FAs and carotenoids, and tissue carotenoid concentrations. Previous carotenoid intervention studies [
      • Trieschmann M.
      • Beatty S.
      • Nolan J.M.
      • Hense H.W.
      • Heimes B.
      • Austermann U.
      • et al.
      Changes in macular pigment optical density and serum concentrations of its constituent carotenoids following supplemental lutein and zeaxanthin: the LUNA study.
      ,
      • Nolan J.
      • Loughman J.
      • Akkali M.C.
      • Stack J.
      • Scanlon G.
      • Davison P.
      • et al.
      The impact of macular pigment augmentation on visual performance in normal subjects: COMPASS.
      ] have shown that increases in carotenoid concentrations in serum/plasma do not necessarily lead to a response in tissue (i.e. MP). Many researchers have hypothesised why the carotenoid response in tissue is less reliable than the response in blood, with some suggesting that genetics [
      • Borel P.
      • de Edelenyi F.S.
      • Vincent-Baudry S.
      • Malezet-Desmoulin C.
      • Margotat A.
      • Lyan B.
      • et al.
      Genetic variants in BCMO1 and CD36 are associated with plasma lutein concentrations and macular pigment optical density in humans.
      ] or lifestyle factors [
      • Nolan J.
      • O'Donovan O.
      • Kavanagh H.
      • Stack J.
      • Harrison M.
      • Muldoon A.
      • et al.
      Macular pigment and percentage of body fat.
      ] may explain the variation in MP augmentation. Previous work has illustrated improvements in serum DHA and lutein concentrations following 4-month supplementation with lutein-only, DHA-only, and lutein plus DHA in comparison to individuals consuming placebo [
      • Johnson E.J.
      • Chung H.Y.
      • Caldarella S.M.
      • Snodderly D.M.
      The influence of supplemental lutein and docosahexaenoic acid on serum, lipoproteins, and macular pigmentation.
      ]. However, only individuals consuming lutein exhibited statistically significant improvements in MP. In the present study, all individuals in the active intervention group exhibited an increase in MP in comparison to individuals consuming placebo (with the exception of 1 patient where acquisition of MP was poor and therefore questionable). We suggest that the consistency in tissue response is due to the presence of ω-3 FAs. While this conclusion cannot be tested directly due to the lack of carotenoid- and omega-only groups, this hypothesis is supported by the positive and significant relationships between ω-3FA and carotenoid concentrations in blood and tissue carotenoid concentrations (Supplementary eTable 1). Interestingly, and in support of this hypothesis, it has been shown that carotenoid density at the centre of the macula is directly associated with ω-3FA index [
      • Rutledge G.A.
      • Pratt S.G.
      • Richer S.P.
      • Huntjens B.
      • Perry C.B.
      • Pratt G.
      • et al.
      Foveal macular pigment dip in offspring of age-related macular degeneration patients is inversely associated with omega-3 index.
      ] and plasma concentrations of docosapentaenoic acid [
      • Merle B.M.J.
      • Buaud B.
      • Korobelnik J.F.
      • Bron A.
      • Delyfer M.N.
      • Rougier M.B.
      • et al.
      Plasma long-chain omega-3 polyunsaturated fatty acids and macular pigment in subjects with family history of age-related macular degeneration: the Limpia Study.
      ]. Given this likely relationship, the suggestion that DHA facilitated a more consistent response in tissue carotenoid concentrations warrants further investigation in the future work.

      4.4 Null and unexpected findings

      While individuals receiving the active intervention responded positively to ω-3FA and carotenoid supplementation, improvements in d-⍺-tocopherol were not observed. Reasons underlying the poor vitamin E response following supplementation remain unclear. While in accordance with international recommended dietary allowances, it is possible that the daily dosage of vitamin E used in the present study was too low (in comparison to other interventional studies [
      • Dysken M.W.
      • Sano M.
      • Asthana S.
      • Vertrees J.E.
      • Pallaki M.
      • Llorente M.
      • et al.
      Effect of vitamin E and memantine on functional decline in Alzheimer disease: the TEAM-AD VA cooperative randomized trial.
      ]) to have any meaningful effect.
      In addition to working memory, small-scale improvements in attention, language and global cognition were also recorded. These findings are also consistent with the literature [
      • Bowman G.L.
      • Silbert L.C.
      • Howieson D.
      • Dodge H.H.
      • Traber M.G.
      • Frei B.
      • et al.
      Nutrient biomarker patterns, cognitive function, and MRI measures of brain aging.
      ,
      • Ajana S.
      • Weber D.
      • Helmer C.
      • Merle B.M.
      • Stuetz W.
      • Dartigues J.F.
      • et al.
      Plasma concentrations of lutein and zeaxanthin, macular pigment optical density, and their associations with cognitive performances among older adults.
      ,
      • Feeney J.
      • O'Leary N.
      • Moran R.
      • O'Halloran A.M.
      • Nolan J.M.
      • Beatty S.
      • et al.
      Plasma lutein and zeaxanthin are associated with better cognitive function across multiple domains in a large population-based sample of older adults: findings from the Irish longitudinal study on aging.
      ,
      • Hammond Jr., B.R.
      • Miller L.S.
      • Bello M.O.
      • Lindbergh C.A.
      • Mewborn C.
      • Renzi-Hammond L.M.
      Effects of lutein/zeaxanthin supplementation on the cognitive function of community dwelling older adults: a randomized, double-masked, placebo-controlled trial.
      ]. An unusual finding from our research included statistically significant improvements in the RBANS immediate memory domain in the placebo group. This was unexpected and may be a true result due to a learning/practice effect for this cognitive domain or driven by the poor performance of a small number of subjects in the active intervention group at their final study visit. Given that ω-3FAs [
      • Yurko-Mauro K.
      • McCarthy D.
      • Rom D.
      • Nelson E.B.
      • Ryan A.S.
      • Blackwell A.
      • et al.
      Beneficial effects of docosahexaenoic acid on cognition in age-related cognitive decline.
      ] and carotenoids [
      • Power R.
      • Coen R.F.
      • Beatty S.
      • Mulcahy R.
      • Moran R.
      • Stack J.
      • et al.
      Supplemental retinal carotenoids enhance memory in healthy individuals with low levels of macular pigment in A randomized, double-blind, placebo-controlled clinical trial.
      ,
      • Kelly D.
      • Coen R.F.
      • Akuffo K.O.
      • Beatty S.
      • Dennison J.
      • Moran R.
      • et al.
      Cognitive function and its relationship with macular pigment optical density and serum concentrations of its constituent carotenoids.
      ] have been shown to positively impact episodic memory, the null findings for episodic memory in the present study were surprising. This may be due to a lack of statistical power or due to the age of the sample (combined aged of 69.40 ± 4.07 years). Given that age-related changes in episodic memory accelerate after age 60 and that changes in the relevant neuro-circuitry may have already occurred, nutritional supplementation may have been too late to exhibit an effect.

      4.5 Strengths and limitations

      Strengths of CARES include its double-blind, placebo-controlled, randomised design, strict eligibility criteria which ensured a clean dataset to test hypotheses of interest (e.g. no previous consumption of carotenoid supplements), and the comprehensive assessment of cognition using sensitive and validated diagnostic measurement tools. However, it is important to acknowledge that the results of this trial are not necessarily generalisable to the overall population and may be subject to selection bias, given that data were collected at a single-site (Nutrition Research Centre Ireland) and involved a study sample that was primarily recruited from the same geographical area. Despite these limitations, this study found improvements in cognitive performance, xanthophyll carotenoid concentrations in tissue and serum, and plasma concentrations of ω-3FAs following a 24-month nutritional intervention trial, with the observed improvements in cognition related to the observed increases in the nutrients of interest.

      5. Conclusion

      In conclusion, this research has shown improvements in working memory following 24-month supplementation with ω-3FAs, xanthophyll carotenoids and vitamin E in cognitively healthy older adults. These results support a biologically plausible rationale whereby these nutrients work synergistically, and in a dose-dependent manner, to improve cognitive performance. These findings illustrate the importance of nutritional enrichment in improving cognition and enabling older adults to continue to function independently, and highlight how a combination of ω-3FAs and xanthophyll carotenoids may prove beneficial in reducing cognitive decline and/or delaying Alzheimer's disease onset in later life.

      Funding statement

      This research was funded by the Howard Foundation UK (UK Charity Registration Number 285822 ). Dr Alan Howard, founder of the Howard Foundation UK, was involved in the conceptualisation and study design of CARES.

      Author contributions

      Rebecca Power: Data curation, formal analysis, investigation, project administration, visualisation, roles/writing – original draft, writing – review & editing.
      John M. Nolan: Formal analysis, methodology, resources, supervision, visualisation, roles/writing – original draft, Writing – review & editing.
      Alfonso Prado-Cabrero: Methodology, resources, supervision, validation, roles/writing – original draft, Writing – review & editing.
      Warren Roche: Data curation, formal analysis, software, writing – review & editing.
      Robert Coen: Methodology, validation, visualisation, writing – review & editing.
      Tommy Power: Validation, writing – review & editing.
      Ríona Mulcahy: Methodology, resources, supervision, validation, visualisation, roles/writing – original draft, writing – review & editing.

      Conflict of Interest

      Rebecca Power: RP has performed consultancy work for MacuHealth LLC™ (Birmingham, MI, USA). RP is funded in part by the Howard Foundation (registered with the Charity Commission of England & Wales #285822), hereafter “Howard Foundation”. These organisations have an interest in commercially available supplements containing the macular carotenoids. RP is also funded by a joint research centre grant from Science Foundation Ireland (SFI) and the Department of Agriculture, Food, and Marine on behalf of the government of Ireland under grant #16/RC/3835 —VistaMilk to develop commercial dairy products enriched in carotenoids. John M. Nolan does consultancy work as a Director of NOW Science Consultancy Ltd. for companies with an interest in food supplements. Alfonso Prado-Cabrero: APC has performed consultancy work for MacuHealth LLC™ and the Howard Foundation. APC has also been involved in a Commercialisation Fund Programme from Enterprise Ireland to develop a biotechnological process to produce carotenoids and the fatty acids EPA and DHA. APC is currently supported by grant #16/RC/3835 — VistaMilk . Robert Coen, Warren Roche and Tommy Power declare no conflicts of interest. Ríona Mulcahy does consultancy work on behalf of the Howard Foundation.

      Acknowledgements

      The authors wish to acknowledge the late Dr Alan Howard, nutritionist, philanthropist and founder of the Howard Foundation UK (a charitable trust in the UK that supports scientific research into nutriceuticals) for his inception of CARES and dedication to the research conducted at the Nutrition Research Centre Ireland. The authors also thank Dr Michael Kirby (University Hospital Waterford, Waterford, Ireland) for his contribution to the consensus panel; Catherine Kelly and Lisa O'Brien (UPMC Whitfield Pharmacy, Waterford, Ireland) for the management of the trial supplements; Dr Jim Stack for the statistical work performed during the planning and initiation stages of the trial; and Heidelberg Engineering (Heidelberg, Germany) for their support with the measurement of macular pigment.

      Appendix A. Supplementary data

      The following are the supplementary data to this article:

      References

        • Zwilling C.E.
        • Talukdar T.
        • Zamroziewicz M.K.
        • Barbey A.K.
        Nutrient biomarker patterns, cognitive function, and fMRI measures of network efficiency in the aging brain.
        Neuroimage. 2019; 188: 239-251
        • Loughrey D.G.
        • Lavecchia S.
        • Brennan S.
        • Lawlor B.A.
        • Kelly M.E.
        The impact of the mediterranean diet on the cognitive functioning of healthy older adults: a systematic review and meta-analysis.
        Adv Nutr. 2017; 8: 571-586
        • Jennings A.
        • Cunnane S.C.
        • Minihane A.M.
        Can nutrition support healthy cognitive ageing and reduce dementia risk?.
        BMJ. 2020; 369: m2269
        • Solfrizzi V.
        • Custodero C.
        • Lozupone M.
        • Imbimbo B.P.
        • Valiani V.
        • Agosti P.
        • et al.
        Relationships of dietary patterns, foods, and micro- and macronutrients with Alzheimer's disease and late-life cognitive disorders: a systematic review.
        J Alzheimers Dis. 2017; 59: 815-849
        • Yuan C.
        • Chen H.
        • Wang Y.
        • Schneider J.A.
        • Willett W.C.
        • Morris M.C.
        Dietary carotenoids related to risk of incident Alzheimer dementia (AD) and brain AD neuropathology: a community-based cohort of older adults.
        Am J Clin Nutr. 2020; 113: 200-208
        • Singh M.
        Essential fatty acids, DHA and human brain.
        Indian J Pediatr. 2005; 72: 239-242
        • Weiser M.J.
        • Butt C.M.
        • Mohajeri M.H.
        Docosahexaenoic acid and cognition throughout the lifespan.
        Nutrients. 2016; 8: 99
        • Craft N.E.
        • Haitema T.B.
        • Garnett K.M.
        • Fitch K.A.
        • Dorey C.K.
        Carotenoid, tocopherol, and retinol concentrations in elderly human brain.
        J Nutr Health Aging. 2004; 8: 156-162
        • Johnson E.J.
        • Vishwanathan R.
        • Johnson M.A.
        • Hausman D.B.
        • Davey A.
        • Scott T.M.
        • et al.
        Relationship between serum and brain carotenoids,-tocopherol, and retinol concentrations and cognitive performance in the oldest old from the Georgia Centenarian Study.
        J Aging Res. 2013; 2013
        • Lieblein-Boff J.C.
        • Johnson E.J.
        • Kennedy A.D.
        • Lai C.S.
        • Kuchan M.J.
        Exploratory metabolomic analyses reveal compounds correlated with lutein concentration in frontal cortex, Hippocampus, and occipital cortex of human infant brain.
        PLoS One. 2015; 10e0136904
        • Bowman G.L.
        • Silbert L.C.
        • Howieson D.
        • Dodge H.H.
        • Traber M.G.
        • Frei B.
        • et al.
        Nutrient biomarker patterns, cognitive function, and MRI measures of brain aging.
        Neurology. 2012; 78: 241-249
        • Kryscio R.J.
        • Abner E.L.
        • Caban-Holt A.
        • Lovell M.
        • Goodman P.
        • Darke A.K.
        • et al.
        Association of antioxidant supplement use and dementia in the prevention of Alzheimer's disease by vitamin E and selenium trial (PREADViSE).
        JAMA Neurol. 2017; 74: 567-573
        • Johnson E.J.
        • McDonald K.
        • Caldarella S.M.
        • Chung H.Y.
        • Troen A.M.
        • Snodderly D.M.
        Cognitive findings of an exploratory trial of docosahexaenoic acid and lutein supplementation in older women.
        Nutr Neurosci. 2008; 11: 75-83
        • Power R.
        • Nolan J.M.
        • Prado-Cabrero A.
        • Coen R.
        • Roche W.
        • Power T.
        • et al.
        Targeted nutritional intervention for patients with mild cognitive impairment: the cognitive impAiRmEnt study (CARES) trial 1.
        J Personalized Med. 2020; 10
        • Albert M.S.
        • DeKosky S.T.
        • Dickson D.
        • Dubois B.
        • Feldman H.H.
        • Fox N.C.
        • et al.
        The diagnosis of mild cognitive impairment due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease.
        AlzheimersDement. 2011; 7: 270-279
        • Dubois B.
        • Albert M.L.
        Amnestic MCI or prodromal Alzheimer's disease?.
        Lancet Neurol. 2004; 3: 246-248
        • Sabour-Pickett S.
        • Beatty S.
        • Connolly E.
        • Loughman J.
        • Stack J.
        • Howard A.
        • et al.
        Supplementation with three different macular carotenoid formulations in patients with early age-related macular degeneration.
        Retina. 2014; 34: 1757-1766
        • Meagher K.A.
        • Thurnham D.I.
        • Beatty S.
        • Howard A.N.
        • Connolly E.
        • Cummins W.
        • et al.
        Serum response to supplemental macular carotenoids in subjects with and without age-related macular degeneration.
        Br J Nutr. 2013; 110: 289-300
        • Shindou H.
        • Koso H.
        • Sasaki J.
        • Nakanishi H.
        • Sagara H.
        • Nakagawa K.M.
        • et al.
        Docosahexaenoic acid preserves visual function by maintaining correct disc morphology in retinal photoreceptor cells.
        J Biol Chem. 2017; 292: 12054-12064
        • Skinner E.R.
        • Watt C.
        • Besson J.A.
        • Best P.V.
        Differences in the fatty acid composition of the grey and white matter of different regions of the brains of patients with Alzheimer's disease and control subjects.
        Brain. 1993; 116: 717-725
        • Nolan J.M.
        • Mulcahy R.
        • Power R.
        • Moran R.
        • Howard A.N.
        Nutritional intervention to prevent Alzheimer's disease: potential benefits of xanthophyll carotenoids and omega-3 fatty acids combined.
        J Alzheimers Dis. 2018; 64: 367-378
        • Nasreddine Z.S.
        • Phillips N.A.
        • Bedirian V.
        • Charbonneau S.
        • Whitehead V.
        • Collin I.
        • et al.
        The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment.
        J Am Geriatr Soc. 2005; 53: 695-699
        • Randolph C.
        • Tierney M.C.
        • Mohr E.
        • Chase T.N.
        The repeatable battery for the assessment of neuropsychological status (RBANS): preliminary clinical validity.
        J Clin Exp Neuropsychol. 1998; 20: 310-319
        • Randolph C.
        RBANS Update : Repeatable Battery for the Assessment of Neuropsychological Status Manual. U.S.A.: Bloomington, Minn.
        NCS Pearson, PsychCorp2012
        • Cognition Cambridge
        CANTAB Connect Research: Admin Application User Guide v1.6.
        Cambridge Cognition Limited, Cambridge, UK2019
        • Cognition Cambridge
        Product Overview: CANTAB Connect Research v11.10.
        Cambridge Cognition Limited, Cambridge, UK2019
        • Vishwanathan R.
        • Schalch W.
        • Johnson E.J.
        Macular pigment carotenoids in the retina and occipital cortex are related in humans.
        Nutr Neurosci. 2016; 19: 95-101
        • Feeney J.
        • Finucane C.
        • Savva G.M.
        • Cronin H.
        • Beatty S.
        • Nolan J.M.
        • et al.
        Low macular pigment optical density is associated with lower cognitive performance in a large, population-based sample of older adults.
        Neurobiol Aging. 2013; 34: 2449-2456
        • Ajana S.
        • Weber D.
        • Helmer C.
        • Merle B.M.
        • Stuetz W.
        • Dartigues J.F.
        • et al.
        Plasma concentrations of lutein and zeaxanthin, macular pigment optical density, and their associations with cognitive performances among older adults.
        Invest Ophthalmol Vis Sci. 2018; 59: 1828-1835
        • Trieschmann M.
        • Heimes B.
        • Hense H.W.
        • Pauleikhoff D.
        Macular pigment optical density measurement in autofluorescence imaging: comparison of one- and two-wavelength methods.
        in: Graefe's archive for clinical and experimental ophthalmology = Albrecht von Graefes Archiv fur klinische und experimentelle Ophthalmologie. 244. 2006: 1565-1574 (12)
        • Roche W.
        • Green-Gomez M.
        • Moran R.
        • Nolan J.
        The physics of using the Heidelberg Spectralis dual-wavelength autofluorescence method for the measurement of macular pigment volume [Abstract].
        J Alzheim Dis. 2018; 64: 1019-1048
        • Green-Gomez M.
        • Bernstein P.S.
        • Curcio C.A.
        • Moran R.
        • Roche W.
        • Nolan J.
        Standardizing the assessment of macular pigment using a dual-wavelength Autofluorescence technique.
        Transl Vision Sci Technol. 2019; 8: 41
        • Lademann J.
        • Meinke M.C.
        • Sterry W.
        • Darvin M.E.
        Carotenoids in human skin.
        Exp Dermatol. 2011; 20: 377-382
        • Zidichouski J.A.
        • Mastaloudis A.
        • Poole S.J.
        • Reading J.C.
        • Smidt C.R.
        Clinical validation of a noninvasive, Raman spectroscopic method to assess carotenoid nutritional status in humans.
        J Am Coll Nutr. 2009; 28: 687-693
        • Janse VAN Rensburg A.
        • Wenhold F.
        Validity and reliability of field resonance Raman spectroscopy for assessing carotenoid status.
        J Nutr Sci Vitaminol. 2016; 62: 317-321
        • Lakens D.
        Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs.
        Front Psychol. 2013; 4: 863
        • Cohen J.
        Statistical power analysis for the behavioral sciences.
        2nd ed. Lawrence Erlbaum Associates, Hillsdale, New Jersey1988 1988
        • Ubeda N.
        • Achon M.
        • Varela-Moreiras G.
        Omega 3 fatty acids in the elderly.
        Br J Nutr. 2012; 107: S137-S151
        • Kulzow N.
        • Witte A.V.
        • Kerti L.
        • Grittner U.
        • Schuchardt J.P.
        • Hahn A.
        • et al.
        Impact of omega-3 fatty acid supplementation on memory functions in healthy older adults.
        J Alzheimers Dis. 2016; 51: 713-725
        • Power R.
        • Coen R.F.
        • Beatty S.
        • Mulcahy R.
        • Moran R.
        • Stack J.
        • et al.
        Supplemental retinal carotenoids enhance memory in healthy individuals with low levels of macular pigment in A randomized, double-blind, placebo-controlled clinical trial.
        J Alzheimers Dis. 2018; 61: 947-961
        • Feeney J.
        • O'Leary N.
        • Moran R.
        • O'Halloran A.M.
        • Nolan J.M.
        • Beatty S.
        • et al.
        Plasma lutein and zeaxanthin are associated with better cognitive function across multiple domains in a large population-based sample of older adults: findings from the Irish longitudinal study on aging.
        J Gerontol A Biol Sci Med Sci. 2017; 72: 1431-1436
        • Dangour A.D.
        • Allen E.
        • Elbourne D.
        • Fletcher A.
        • Richards M.
        • Uauy R.
        Fish consumption and cognitive function among older people in the UK: baseline data from the OPAL study.
        J Nutr Health Aging. 2009; 13: 198-202
        • Goldstein E.B.
        Cognitive Psychology: Connecting Mind, Research and Everyday Experience.
        3rd ed. 3rd ed. Wadsworth CENGAGE Learning, Canada2011
        • Power R.
        • Prado-Cabrero A.
        • Mulcahy R.
        • Howard A.
        • Nolan J.M.
        The role of nutrition for the aging population: implications for cognition and Alzheimer's disease.
        Annu Rev Food Sci Technol. 2019; 10: 619-639
        • Trieschmann M.
        • Beatty S.
        • Nolan J.M.
        • Hense H.W.
        • Heimes B.
        • Austermann U.
        • et al.
        Changes in macular pigment optical density and serum concentrations of its constituent carotenoids following supplemental lutein and zeaxanthin: the LUNA study.
        Exp Eye Res. 2007; 84: 718-728
        • Nolan J.
        • Loughman J.
        • Akkali M.C.
        • Stack J.
        • Scanlon G.
        • Davison P.
        • et al.
        The impact of macular pigment augmentation on visual performance in normal subjects: COMPASS.
        Vis Res. 2011; 51: 459-469
        • Borel P.
        • de Edelenyi F.S.
        • Vincent-Baudry S.
        • Malezet-Desmoulin C.
        • Margotat A.
        • Lyan B.
        • et al.
        Genetic variants in BCMO1 and CD36 are associated with plasma lutein concentrations and macular pigment optical density in humans.
        Ann Med. 2011; 43: 47-59
        • Nolan J.
        • O'Donovan O.
        • Kavanagh H.
        • Stack J.
        • Harrison M.
        • Muldoon A.
        • et al.
        Macular pigment and percentage of body fat.
        Invest Ophthalmol Vis Sci. 2004; 45: 3940-3950
        • Johnson E.J.
        • Chung H.Y.
        • Caldarella S.M.
        • Snodderly D.M.
        The influence of supplemental lutein and docosahexaenoic acid on serum, lipoproteins, and macular pigmentation.
        Am J Clin Nutr. 2008; 87: 1521-1529
        • Rutledge G.A.
        • Pratt S.G.
        • Richer S.P.
        • Huntjens B.
        • Perry C.B.
        • Pratt G.
        • et al.
        Foveal macular pigment dip in offspring of age-related macular degeneration patients is inversely associated with omega-3 index.
        BMC Ophthalmol. 2020; 20: 473
        • Merle B.M.J.
        • Buaud B.
        • Korobelnik J.F.
        • Bron A.
        • Delyfer M.N.
        • Rougier M.B.
        • et al.
        Plasma long-chain omega-3 polyunsaturated fatty acids and macular pigment in subjects with family history of age-related macular degeneration: the Limpia Study.
        Acta Ophthalmol. 2017; 95: e763-e769
        • Dysken M.W.
        • Sano M.
        • Asthana S.
        • Vertrees J.E.
        • Pallaki M.
        • Llorente M.
        • et al.
        Effect of vitamin E and memantine on functional decline in Alzheimer disease: the TEAM-AD VA cooperative randomized trial.
        J Am Med Assoc. 2014; 311: 33-44
        • Hammond Jr., B.R.
        • Miller L.S.
        • Bello M.O.
        • Lindbergh C.A.
        • Mewborn C.
        • Renzi-Hammond L.M.
        Effects of lutein/zeaxanthin supplementation on the cognitive function of community dwelling older adults: a randomized, double-masked, placebo-controlled trial.
        Front Aging Neurosci. 2017; 9: 254
        • Yurko-Mauro K.
        • McCarthy D.
        • Rom D.
        • Nelson E.B.
        • Ryan A.S.
        • Blackwell A.
        • et al.
        Beneficial effects of docosahexaenoic acid on cognition in age-related cognitive decline.
        Alzheimer's Dementia. 2010; 6: 456-464
        • Kelly D.
        • Coen R.F.
        • Akuffo K.O.
        • Beatty S.
        • Dennison J.
        • Moran R.
        • et al.
        Cognitive function and its relationship with macular pigment optical density and serum concentrations of its constituent carotenoids.
        J Alzheimers Dis. 2015; 48: 261-277