Effect of macular pigment carotenoids on cognitive functions

Physiology & Behavior 254 (2022) 113891 Available online 23 June 2022 0031-9384/Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Review Effect of macular pigment carotenoids on cognitive functions: A systematic review Marta-C. García-Romera a , * , María-Carmen Silva-Viguera a , Inmaculada L´ opez-Izquierdo a , Alfredo L´ opez-Mu˜ noz a , Raúl Capote-Puente a , Beatriz Gargallo-Martínez a , b a Vision Science Research Group (CIVIUS), Department of Physics of Condensed Matter - Optics Area, Av. de la Reina Mercedes, S/N, 41012 Sevilla, University of Seville, Spain b Faculty of Optics and Optometry, Campus Universitario de Espinardo, Edificio 35, 30100 Espinardo, Murcia, University of Murcia, Spain ARTICLE INFO Keywords: Brain activity Cognitive ability Cognitive function Lutein and Zeaxanthin Macular pigment optical density ABSTRACT Lutein and zeaxanthin—xanthophyll carotenoids with antioxidant and anti-inflammatory characteristics—are present in the retina and the brain. High concentrations of these carotenoids have been positively related to cognitive performance. Therefore, this systematic review analyses the relationship between macular pigment density and cognitive functions. Most relevant databases were scoured for studies on healthy people relating cognitive functions to macular pigment optical density (MPOD). There were no age, sex, or race limitations. PROSPERO registration: CRD42021254833. Nineteen studies were included, seven randomized controlled trials (RCT) and eleven observational studies. The general aim of the studies was to examine the association between carotenoids (lutein, meso‑zeaxanthin and zeaxanthin) and cognitive function. Most observational studies correlates MPOD levels with cognitive function or brain activity. Besides, RCTs compared the cognitive function and/or brain activity after increasing lutein and zeaxanthin intake though dietary supplementation or avocado consumption. Dietary lutein and zeaxanthin intake increased MPOD in six of the seven clinical trials and significantly improved most of the cognitive functions studied. A wide variety of test and methodologies for measuring cognitive functions were observed. Memory, processing speed, attention and reasoning were the cognitive function significantly related to MPOD levels in adults. Brain activity also was related to MPOD, but the results were inconsistent. Only four of the eleven observational studies were based on young people and all studies showed a significant relationship between MPOD and cognitive functions. This systematic review showed a direct relationship among cognitive functions, macular pigment and the intake of lutein and zeaxanthin. 1. Introduction Carotenoids are natural plant pigments commonly found in human diet, mainly in green leafy vegetables, bright-coloured fruits and eggs [1]. Lutein (L) and zeaxanthin (Z) are oxygenated carotenoids or xanthophylls, which are accumulated in the macula lutea, a 5–6mm-diameter area of the retina responsible for central vision [2]. Unlike L and Z, which are entirely of dietary origin, meso‑zeaxanthin (MZ) is formed in the macula from L in a 1:1 ratio [3]. L, Z, and MZ are collectively referred to as macular pigment (MP) and protect the retina from short-wavelength blue light and oxidative stress [4]. Higher intake of these nutrients, either through diet or Abbreviations: AD, Alzheimer’s disease; AMD, age-related macular degeneration; BIA, brief intellectual ability; BVRT, Benton Visual Retention Test; CANTAB, Cambridge Neuropsychological Test Automated Battery; CF, cognitive functions; CNS, central nervous system; CNSVS, CNS Vital Signs; FCSRT, Free and Cued Selective Reminding Test; IQ, Intelligent quotient; IST 15, Isaacs Set Test; KBIT-2, Kaufman Brief Intelligence Test Second Edition; K-TEA, Kaufman Test of Educational Achievement; L, Lutein; MP, macular pigment; MPD, macular pigment density; MPOD, macular pigment optical density; MZ, meso-zeaxanthin; PS, Processing Speed; RCI, Reliable Change Index; RCT, Randomised Controlled Trials; WJ-III, Woodcock-Johnson III; WTAR, Wechsler Test of Adult Reading; Z, zeaxanthin. * Corresponding author at: Department of Physics of Condensed Matter - Optics Area, Av. de la Reina Mercedes, S/N, 41012 Sevilla, University of Seville, Spain. E-mail address: [email protected] (M.-C. García-Romera). Contents lists available at ScienceDirect Physiology & Behavior journal homepage: www.elsevier.com/locate/physbeh https://doi.org/10.1016/j.physbeh.2022.113891 Received 15 March 2022; Received in revised form 18 June 2022; Accepted 20 June 2022 Physiology & Behavior 254 (2022) 113891 2 supplementation, has been associated with an increase in MP [5]. There is evidence that they may improve vision in individuals with the early form of age-related macular degeneration (AMD) [6]—a leading cause of vision-loss among older adults—and retard the progression of this disease [7]. IBased on our understanding of how L and Z benefit neural health of the retina, these carotenoids may prevent age-related cognitive decline, mainly with the areas of spatial memory, reasoning, complex attention and attributable also to young and pediatric people through their strong antioxidant and anti-inflammatory properties [8-10]. The retina is a component of the central nervous system (CNS), where carotenoids are found in the hippocampus, the occipital lobe, and the frontal cortex of the brain [11]. This knowledge has stimulated research on their potential cognitive benefits. A growing body of literature indicates that in older adults, MP positively correlate with a range of cognitive functions (CF), including visuospatial skills, learning, memory, language abilities, executive functions, processing speed, and global cognition [12,13], which are affected by dietary habits and nutrition [14]. There is also a corpus of evidence suggesting that good nutrition is important for optimal cognition [15], and it is also associated with a reduced risk of Alzheimer’s disease (AD)—the most common form of dementia—in later life [16]. To assess cognitive performance in individuals, a variety of cognitive batteries or assessments are employed, such as the Cambridge Neuropsychological Test Automated Battery (CANTAB) [17]—one of the most widely used— or the Woodcock-Johnson III (WJ-III), which is a set of standardised tests of cognitive skills, such as short intellectual ability, verbal ability, cognitive efficiency, processing speed, and decisional performance processes [18]. At least two of the three macular carotenoids are found in the brain [19] in concentrations proportional to that in the retina [20]. Therefore, macular pigment density (MPD) likely reflects L and Z brain concentrations. Given that concentration levels of these carotenoids relate positively to cognitive performance in both cognitively impaired [21] and cognitively intact individuals [22], the potential role of these dietary compounds (L and Z) in optimizing and maintaining cognition warrants study [23]. This may explain the epidemiological evidence of a significant relationship between MPOD and cognitive health in two cross-sectional studies [21,24]. So, MPOD could likely be just a biomarker for the effects of L-Z in brain. While it is true that the underlying mechanisms in vivo of L-Z in the brain are unclear, there are different possibilities about the mechanism that may influence latent physio-neurological mechanisms in cognitive or visual functions. A presumed indicator is increased brain perfusion in prefrontal regions, improving the potential neuronal mechanism over the course of cognitive performance. Another could be related to pathologies that may affect the brain such as oxidative, inflammatory or cardiometabolic stress [25]. This study analyses the relationship between MPD and cognitive capacity in healthy people, based on the scientific evidence presented in previous research, employing a systematic review. 2. Experimental section This systematic review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement recommendations [26] and was registered with the PROSPERO International Prospective Registry (record CRD42021254833). 2.1. Search strategy A systematic search based on population, intervention, comparison, and outcome strategy was performed between 1 December 2020 and 1 February 2021. The study population consisted of healthy individuals (non-AMD), without restriction by age, gender, or weight. Intervention with L and/or Z through diet and/or dietary supplementation was evaluated and compared to placebo groups. The relationship between macular pigment optical density (MPOD) and CF was investigated. Three databases namely, PubMed, Scopus, and Web of Science were scoured for articles published between 1 November 2015 and 1 January 2021. The following search terms were used: macular pigment, lutein, zeaxanthin, xanthophyll, carotenoids, cognition function, neural efficiency, memory, cognition diseases, intellectual ability, cognitive performance, and neurocognitive functioning. Considering the format differences for each database, the combination of keywords was the same in all of them: (Macular pigment OR MPOD OR Lutein OR Zeaxanthin OR carotenoids OR xanthophyll) AND (cognition function OR Neural Efficiency OR Memory OR cognition diseases OR intellectual ability OR cognitive performance OR Neurocognitive Functioning). 2.2. Study selection The studies were evaluated by two independent reviewers and any disagreement was resolved by a third reviewer. First, studies searched were imported to Mendeley and duplicates were removed. Later, titles and abstracts were evaluated according to inclusion and exclusion criteria. Finally, full texts identified were screened. 2.3. Inclusion and exclusion criteria Inclusion criteria: (1) scientific articles with experimental studies (clinical trials) and observational studies (cross-sectional, case-control, cohort study designs) published in peer-review journals; (2) human studies; (3) MPOD and CF should be studied; and (4) dietary supplementation with L and/or Z as intervention. Exclusion criteria: (1) studies without MP or cognitive function analysis; (2) studies of psychosis and mental illness; (3) studies of possibility or risk of developing AD and dementia; and (4) animal trials. 2.4. Data analysis and study quality Two reviewers performed data analysis and study quality assessment independently. A third reviewer was consulted when disagreements arose. Recorded data included: (1) study design (type, purpose, duration); (2) inclusion and exclusion criteria; (3) sample characteristics, including age, race, and body mass index (BMI); (4) sample size (number of patients, percentage of males, groups); (5) MPOD assessment (basal value, change post-treatment); (6) L and Z concentrations in blood serum (basal value, change post-treatment); (7) CF evaluation (cognitive test battery, task for calculating the index score); and (8) relationship between MPOD and CF. The data were expressed as mean±SD. Studies based on children and pre-adolescents were analysed separately from studies on adults (that is, studies on people more than 18 years of age). A qualitative synthesis of the results was performed to determine the relationship between MPOD and CF. The impact of carotenoid supplementation was evaluated first on MPOD and then on brain activity and/or cognitive ability. We considered p <0.05 as statistically significant. The quality of studies was evaluated using the National Institutes of Health (NIH)’s tool for ‘Quality Assessment of Controlled Intervention Studies’ and the ‘Quality Assessment Tool for Observational Cohort and Cross-sectional Studies’ [27]. Studies with low, medium and high risks of bias were considered of good, fair and poor qualities, respectively. 3. Results 3.1. Flow diagram Fig. 1 shows the selection process in this systematic review and Table 1 presents the characteristics of both the samples and the studies included. M.-C. García-Romera et al. Physiology & Behavior 254 (2022) 113891 3 3.2. Characteristics of the studies included and their sample Nineteen studies were included in this systematic review (Table 1). A total of 1371 subjects (19.39% [35] −57% [29] male) were analysed, of whom, 420 were participants in clinical trials (active groups: n =264; control group: n =156) [11,23,28–32]. The age range of the sample was from 7 to 86 years. Four studies included only children, aged between 7 and 13 years [18,33–35]; two studies analysed young adults, that is, those aged between 18 and 30 years [28,29] ; five studies selected adults aged between 25 and 45 years [31,36–39]; and six included older adults [11,24,32,40,41] Race was reported in seven studies, with 88.19% being Caucasian; 3.77% Asian; 0.86% Hispanic; 3.99% African–American or Black; and 3.15% of other races [18,24,29,32,39,41,42]. The BMI ≥25.0 kg/m 2 or obesity was the inclusion criterion in four studies [31,36,37,39]. The general objective of the studies analysed was to examine the association between carotenoids (L, MZ, and Z) and CF (Table 1). All studies correlated MP levels with CF [11,18,23,24,28–31,33–40] or brain activity [32,41,42]. Clinical trials compared CF after increasing carotenoids intake through dietary supplementation [11,23,28,29] or avocado consumption [30,31]. In the RCT conducted by Scott et al. [28] subjects in the treatment group ingested 0.40 ±0.01 mg/100 g (edible portion) of lutein respect to <0.1 mg in the control group. In the RCT conducted by Edwards et al. [29] the lutein/zeaxanthin content was 3 times higher in the intervention meals as it was in the control meals (treatment group Female: 561 mcg, Male: 701mcg; control group Female: 164 mcg, Male: 205mcg). 3.3. MPOD MPOD was measured by heterochromatic flicker photometry at 0.5◦ of retinal eccentricity in most of the analysed studies [11,18,24,28–31, 33–37,39,41]. Only two studies measured MP by dual-wavelength auto-fluorescence method [23,40]. Ajana et al. measured MPOD at 0.5◦and 1◦eccentricity [40], while Power et al. showed MP volume out to 7◦eccentricity [23]. Table 2 shows MPOD data. The MPOD at 0.5◦eccentricity [11,18,36,37,39–41,24,28–31, 33–35] ranged from 0.34 [24] to 0.70 [40]. The studies with pediatric population showed a mean MPOD of 0.598 ±0.08 (ranging from 0.48 to 0.66) [18,33–35], and the studies with adult population showed a mean MPOD of 0.472 ±0.08 (ranging from 0.38 to 0.70) [11,28–31,36–42]. The effect of dietary xanthophyll supplementation on MPOD was analysed in five studies [11,28–31]. An active supplement containing 10 mg L and 2 mg Z (DSM Nutritional Product, Kaiseraugst, Switzerland) was used in 2 studies for 12- months [11,29]. The authors found a significant increase in MPOD in both older [11] and younger [29] adults. Stringham et al. used two amounts of daily macular Xanthophyll supplements (a 13 mg supplement containing 10.86 mg L and 2.27 mg Z isomers versus a 27 mg supplement containing 22.33 mg L and 4.70 mg Z) for 6-months in Fig. 1. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) flow diagram for this systematic review [26]. M.-C. García-Romera et al. Physiology & Behavior 254 (2022) 113891 4 Table 1 Types of included studies and characteristics of their samples. First author, year Study Desing, Duration (months) Inclusion Criteria Exclusion Criteria Purpose of study Sample size (n (% male)) Groups: n Age (mean ± SD) Race (n) BMI (kg/ m2) (mean ± SD) Saint, 2018 Observational study (Crosssectional) 7–13 years of age NR Relationship between MPOD to CF 51 (51%) 9.14 (±1,86) White (Non- Hispanic) (39) Hispanic (1) >1 Race Listed (11) NR Barnett, 2018 Observational study (Crosssectional) Normal VA or CVA ND, PD and psychoactive medication status. Relationship between L and Z intake, MPOD values to academic performance 56 (30%) 8.8 (±0,1) NR 18.7 ± 0.4 Hassevoort, 2017 Observational study (Crosssectional) Fitness Improves Thinking in Kids (FITKids) Trial (ClinicalTrials.gov: NCT01619826) ND or AD, PD and psychoactive medication status Relationship between MPOD, aerobic fitness, and central adiposity to memory 40 (37%) 8.8 (±0.11) NR 19.0 ± 0.50 Walk, Khan, et al., 2017 Observational study (Crosssectional) Completed two pretesting sessions ≤60% correct on flanking trials Relationship between MPOD to CF 49 (19.39%) 8.69 (±0.08) NA NA Stringham, 2019 Controlled-trial. Double-blind. Randomised. (6- month) Healthy students at the University of Georgia, non-smokers, with UCVA or CVA ≥20/20 and no history of OP Current or previous supplementation with L and/or Z Relationship between dietary carotenoids to CF 59 (45.8%) Placebo: 10 13 mg: 24 27 mg: 25 21,5 NR 18.5–27 Hammond, 2017 Controlled-trial. Double-blind. Randomised. (12-month) Good overall health; no xanthophyll supplementation in the previous 6 months; BCVA ≥20/40; no ND to impair CF; absence of GC to impair absorption of NS NR Relationship between L +Z supplementation to CF in older adults. 51 (41,18%) Active: 36 Placebo: 15 Total: 73.74 (± 8.20) Active:72.51 (±6.24) Placebo: 70.93 (±5.70) NR NR Renzi- Hammond, 2017 Controlled-trial. Double-blind. Randomised. (12-months) Good overall and ocular health, BCVA≥20:40 and no NS use in the previous six months NA Relationship between L +Z supplementation to CF in young, healthy adults 51 (57%) Active: 37 Placebo: 14 21.21 (±2.52) Active:NR Placebo: NR White (40), Black (6) Pan-Asian (3) Latino (2) NR Scott, 2017 Controlled-trial. Randomised. (6- month) Low intakes of luteinrich foods History of liver, kidney or pancreatic disease; anemia; active bowel disease or resection; bleeding disorders; some food allergies; Some medications Relationship between intake of avocado to cognition 40 (37.5%) Active: 20 Placebo: 20 Active: 63.3 (±11.1) Placebo: 62.5 (±9.2) NR Active: 24.15 ± 3.86 Placebo: 23.03 ± 3.98 Power, 2018 Controlled-trial. Double-blind. Randomised. (12-months) Low MP; ≥18 years; BCVA ≥6/6; ≤5 D of SE; no DM; no ocular pathology; and no previous consumption of NS with L and/or Z and/or MZ NR Relationship between L +Z supplementation to CF in healthy individuals with low MP levels 91 (51.6%) Active: 45 Placebo: 46 45.42 (± 12.40) NR Active: 27.34 ± 4.72 Placebo: 26.22 ± 4.67 Edwards, 2020 Controlled-trial. Randomised. (12-week) Aged 25–45 years with a BMI ≥25.0 kg/m2 BMI <25 kg/m2, pregnancy or lactation, history of ND or MD, non-normal VA and food allergies or intolerances Relationship between avocado consumption to CF in overweight and obesity adults 84 (43%) Active: 47 Placebo: 37 Placebo: 34.0 (±6.2) Active: 34.6 (±5.7) NR Control goup: 31.31 ± 5.49 Avocado group: 32.49 ± 5.83 Lindbergh, 2018 Controlled-trial. Randomised. (12-months) Initial phone screening, more thorough medical history review Left-handedness, TBI, AMD, GD to interfere with supplement absorption, CVA 〈20/ 40, MRI incompatibility, GDS total score 〉19, or ND Relationship between L +Z to brain activity 44 (40.9%) Active: 30 Placebo: 14 Placebo: 70.43 (±5.43) Active: 72.43 (±6.48) Caucasian (44) NR Vishwanathan, 2014 Observational study (Crosssectional) NR NR Relationships between serum L+Z and MPOD with 108 (49.1%) 77.6 (±0,3) Black (33) 26.9 ± 0.4 (continued on next page) M.-C. García-Romera et al. Physiology & Behavior 254 (2022) 113891 5 young adults. Both supplements induced a significant increase in the MPOD, but there are no differences between supplement types [28]. All studies were controlled by a placebo group that showed no significant change in MPOD after follow-up. Power et al. also evaluated the impact of daily dietary supplementation containing 10 mg L, 10 mg MZ and 2 mg Z for 12-months in adults, but the effect was measured on MP volume. The authors found a significant increase of MP volume in active supplementation group (+64%) while placebo group showed no significant change [23]. Xanthophyll supplementation through avocado daily consumption was evaluated in two studies with contrary results [30,31]. While Scott et al., found a significant increase in MPOD after 6-months in healthy adults [30], Edwards et al., did not present significant changes in MPOD after 3-months in overweight adults [31]. 3.4. Blood serum L and Z concentrations in blood serum were evaluated in eleven studies [11,23,42,24,28–31,36,40,41]. All authors used the same analysis method (high performance liquid chromatography), but the units of measurement presented were different. Table 2 shows the data obtained in each study. All controlled-trial found a significant increase in serum L and Z after 6–12 months of xanthophyll supplementation [11,23,28,29], or after 3–6 months of avocado consumption [30,31]. However, Scott et al. also observed a significant increase in serum L in their placebo group [30]. Only six studies analysed the correlation between MP and carotenoids in blood serum [11,28–31,36]. A significant direct relationship between both parameters was observed in all of them. Table 1 (continued) First author, year Study Desing, Duration (months) Inclusion Criteria Exclusion Criteria Purpose of study Sample size (n (% male)) Groups: n Age (mean ± SD) Race (n) BMI (kg/ m2) (mean ± SD) cognition in healthy older adults Ajana, 2018 Observational study (Crosssectional) ≥65 years from 3 French cities Developed late AMD; incomplete data for tests and missing information Relationship between MPOD, L +Z plasma concentrations to CF 184 (31,5%) 82,3 (±4,3) NR NA Cannavale, 2019 Observational study (Crosssectional) Aged 25–45 years, BMI ≥25 kg/m2, and no history of metabolic or gastrointestinal disease, or ND or CD Use of tobacco products Relationship between carotenoids in diet, serum, and the MPOD to memory in overweight or obese adults 94 (54.46%) 34.9 (±6.1) NR 33.3 ± 6.6 Khan, 2018 Observational study (Crosssectional) Aged 25–45 years with BMI >25.0 kg/m2 Pregnant, history of ND, used anti-psychotic or anti-anxiety medication, history of MD, or had non-normal VA Relationship between MPOD to IQ, in overweight and obesity adults 114 (39%) 34.6 (±6.1) Caucasian (91) Khan, 2018 Edwards, 2019 Observational study (Crosssectional) Aged 25–45 years with BMI >25.0 kg/m2 BMI <25 kg/m2, pregnancy, history ND, history of MD or nonnormal VA Relationship between MPOD to CF in overweight and obesity adults 101 (30.7%) 34.98 (±5.85) NR 32.78 ± 5.46 Walk, Edwards, et al., 2017 Observational study (Crosssectional) Aged 25 to 45 years, complete KBIT, a measure of IQ, readable EEG recording, normal VA or CVA, no ND Not complete all relevant aspects of testing, not in the selected age range, pregnant or nursing, or taking certain medications Relationship between age, MPOD, and neurocognitive indices. 60 (29%) 33.8 (±5.7) NA NA Mewborn, 2018 Observational study (Crosssectional) Community-dwelling older adults Left-handedness, history of OP, CVA < 20/40, AMD in either eye, GD to interfere with L and Z absorption, MRI incompatibility, or history of TBI, dementia, or other ND Relationship between L +Z to brain activity 51 (41.2%) 71.75 (±6.16) Caucasian (51) NR Lindbergh, 2017 Observational study (Crosssectional) Community-dwelling older adults (65–86 years) Left-handedness, TBI, AMD in either eye, GD to interfere with L/Z absorption, CVA <20/ 40, MRI incompatibility, and/or evidence of dementia or other ND Relationship between L +Z to brain activity 43 (41.86%) 71.55 (±5.84) Caucasian (43) NA AD: Attentional Disorders; AMD: Age-related Macular Degeneration; BCVA: Best Corrected Visual Acuity; BMI: Body Mass Index;; CD: Cognitive Disorders; CF: Cognitive Function; CVA: Corrected Visual Acuity; DM: Diabetes Mellitus; ERP: Event related potential; GC: Gastric Conditions; GDS: Geriatric Depression Scale; IQ: Intelligence Quotient; KBIT: the Kaufman Brief Intelligence Test; L: Lutein; MD: Metabolic Diseases; MP: Macular Pigment; MPOD: Macular Pigment Optical Density; ND: Neurological Disorders; NS: Nutritional Supplements; OP: Ocular Pathology; PD: Physical Disabilities; SE: Spherical Equivalence; TBI: Traumatic Brain Injury; UCVA: Uncorrected Visual Acuity; VA: Visual Acuity; Z: Zeaxanthin;. M.-C. García-Romera et al. Physiology & Behavior 254 (2022) 113891 6 Table 2 Studies reporting xanthophyll carotenoids in retina and/or blood serum. First author, year Population Retinal concentrations Blood serum concentrations Age (mean ±SD) Sex (% female) MPOD Basal value (mean ± SD) Change post-treatment (time, mean change, p) L basal value (mean ±SD) Z basal value (mean ±SD) L+Z basal value (mean ±SD) Change post-treatment (time, mean change, p) Saint, 2018 10(±4.26) 45.5% 0.476 ±0.167 NA NR NR NR NA Barnett, 2018 8.8 (±0.1) 70% 0.640 ±0.030 NA NR NR NR NA Hassevoort, 2017 8.8 (±0.11) 63% 0.660 ±0.030 NA NR NR NR NA Walk, Khan, et al., 2017 8.69 (±0,08) 80.61% 0.610 ±0.030 NA NR NR NR NA Stringham, 2019 21.5 54.20% NR 6-months Active groups 13 mg/day: þ0.106* (p<0.05) 27 mg/day: þ0.120* (p<0.05) Placebo: +0.026 (NR) Active groups: 0.210 ±0.111 μ g/mL Placebo: 0.237 ±0.141 μ g/mL Active groups: 0.210 ±0.111 μ g/mL Placebo: 0.052 ±0.034 μ g/Ml NR 6-months Active groups: Lþ1.040* μ g/ mL (p<0.05) Zþ0.152* μ g/ mL (p<0.05) Placebo: L −0.015 μ g/ mL (NR) Z −0.002 μ g/ mL (NR) Hammond, 2017 Total: 73.74 (±8.20) Active:72.51 (±6.24) Placebo: 70.93 (±5.70) 58.82% Active: 52.77% Placebo: 73.33% Active group 0.520 ±0.19 Placebo: 0.420 ±0.16 12-months: Active group þ0.070 (p<0.03) Placebo: +0.050 (NR) Active group 0.150 ±0.080 ng/ μ L Placebo: 0.150 ±0.060 ng/ μ L Active group 0.030 ±0.020 ng/ μ L Placebo: 0.030 ±0.020 ng/ μ L Active group: 0.180 ± 0.110 ng/ μ L Placebo: 0.180 ± 0.070 ng/ μ L 12-months: Active group: Lþ0.440* ng/ μ L (p<0.05) Zþ0.100* ng/ μ L (p<0.05) L-Zþ0.540* ng/ μ L (p<0.05) Placebo: L −0.001 ng/ μ L (p>0.05) Z −0.000 ng/ μ L (p>0.05) L-Z −0.001 ng/ μ L (p>0.05) Renzi-Hammond, 2017 21.21 (±2.52) 43% Active: 43.24% Placebo: 42.85% Active group 0.470 ±0.18 Placebo: 0.400 ±0.12 12-months: Active group þ0.090 (p<0.001) Placebo: +0.040 (p>0.05) Active group 0.110 ±0.070 ng/ μ L Placebo: 0.100 ±0.03 ng/ μ L Active group 0.030 ±0.020 ng/ μ L Placebo: 0.030 ±0.020 ng/ μ L Active group: 0.140 ± 0.080 ng/ μ L Placebo: 0.130 ± 0.040 ng/ μ L 12-months: Active group: L+0.270* ng/ μ L (NR) Z+0.050* ng/ μ L (NR) L-Z+0.320* ng/ μ L (NR) Placebo: L+0.090 ng/ μ L (NR) Z+0.010 ng/ μ L (NR) L-Z+0,100 ng/ μ L (NR) Scott, 2017 Active: 63.3 (±11.1) Placebo: 62.5 (±9.2) 62,5% Active: 70% Placebo: 55% Active groups: 0.393 ±0.142 Placebo: 0.380 ±0.141 6-months Active groups þ0.101 (p = 0.001) Placebo: +0.044 (p>0.05) Active groups: 0.330 ±0.139 nmol/L Placebo: 0.322 ±0.134 nmol/L Active groups: 0.067 ±0.017 nmol/L Placebo: 0.065 ±0.033 nmol/L NR 6-months Active groups: Lþ0.084 nmol/L (p =0.001) Z −0.060 nmol/L (p>0.05) Placebo: Lþ0.049 nmol/L (p =0.03) Z+0.013 nmol/L (p =0.004) Power, 2018 45.42 (±12.40) 48.4% Active: NR % Active groups: MP volume 3982 ±1337 12-months Active groups MP volume Active groups: 0.249 ±0.134 μ mol/L Active groups: 0.052 ±0.042 μ mol/L NR 12-months Active groups: Lþ0.647* (continued on next page) M.-C. García-Romera et al. Physiology & Behavior 254 (2022) 113891 7 3.5. CF and MPOD Table 3 summarises the relationship between MPOD and CF. 3.5.1. Cognitive functions in preadolescent children – childhood The CF analysed were: academic achievement (mathematics, written language composite standard scores and reading) [33]; accurate performance [35]; brief intellectual ability (BIA), cognitive efficiency, visual-auditory learning, spatial relations [18]; and relational memory [34]. Intelligent quotient (IQ) was studied by four authors using the WJ [33,34] or Kaufman Test of Educational Achievement (K-TEA) [33,35]. Furthermore, different cognitive abilities were analysed with the WJ-III Tests: BIA, Verbal Ability, Cognitive Efficiency, Processing Speed (PS), and Executive Processes [18] or cognitive test battery (memory task) [34] . Data are showed in Table 3. MPOD was significantly related (p <0.05) to executive processes, BIA, the spatial relations subtest [18], academic achievement, mathematics (math concepts and computation), written language composite standard scores (written expression and spelling) [33], memory [34], and accurate performance [35]. 3.5.2. Cognitive functions in adults Psychomotor and PS [28], verbal and visual memory, CF in the reasoning and executive processes [11], visual memory, reasoning ability and complex attention [29], spatial working memory and efficiency in approaching a problem [30], verbal and visual memory [23], verbal and visual memory, verbal abilities [24,40], PS, attentional Table 2 (continued) First author, year Population Retinal concentrations Blood serum concentrations Age (mean ±SD) Sex (% female) MPOD Basal value (mean ± SD) Change post-treatment (time, mean change, p) L basal value (mean ±SD) Z basal value (mean ±SD) L+Z basal value (mean ±SD) Change post-treatment (time, mean change, p) Placebo: NR% Placebo: 4026 ±1758 þ2558* (p<0.001) Placebo: MP volume −151 (NR) Placebo: 0.283 ±0.133 μ mol/L Placebo: 0.058 ±0.037 μ mol/L μ mol/L (p<0.001) Zþ0.034 μ mol/L (p =0.016) Placebo: L −0.011 μ mol/L (p>0.05) Z+0.005 μ mol/L (p>0.05) Edwards, 2020 Placebo: 34.0 (±6.2) Active: 34.6 (±5.7) 57% Active: 48.93% Placebo: 48.64% Active groups: 0.470 ±0.220 Placebo: 0.470 ±0.190 3-months Active groups +0.020 (p>0.05) Placebo: +0.020 (p>0.05) Active groups: 0.120±0.060 μ mol/L Placebo: 0.140 ±0.070 μ mol/L NR NR 3-months Active groups: Lþ0.040* μ mol/L (p<0.01) Placebo: L −0.005 μ mol/L (p>0.05) Lindbergh, 2018 Placebo: 70.43 (±5.43) Active: 72.43 (±6.48) 59.1% Active: 53.33% Placebo: 71.43% Active groups: 0.540 ±0.190 Placebo: 0.440 ±0.140 3-months Active groups þ0.070* (p = 0.016) Placebo: +0.000 (p>0.05) NR NR NR NA Vishwanathan, 2014 77.6 ±0,3 50.9% 0.343 (±0,018) NA NR NR 494 ±27 nmol/l. NA Ajana, 2018 82.3 (±4,3) 68.5% 0.700 ±0.200 NA NR NR 0.400 ± 0.200 μ M NA Cannavale, 2019 34.9 (±6.1) 45.54% 0.438 ±0.200 NA 0.129 ±0.06 μ mol/L NR NR NA Khan, 2018 34.6 (±6.1) 61% 0.460 ±0.210 NA NR NR NR NA Edwards, 2019 34.98 (±5,85) 69.30% 0.480 ±0.260 NA NR NR NR NA Walk, Edwards, et al., 2017 33.8 (±5.7) 71% 0.490 ±0.250 NA NR NR NR NA Mewborn, 2018 71.75 (±6.16) 58,8% Younger adults 0.426 ±0.160 Older adults 0.496 ±0.170 NA NR NR Younger adults 0.247 ± 0.120 μ mol/ L Older adults 0.308 ± 0.670 μ mol/ L NA Lindbergh, 2017 71.55 (±5,84) 58.14% 0.510 ±0.180 NA NR NR 0.310 ± 0.170 μ mol/ L NA MPOD Macular Pigment Optical Density; L =Lutein; Z =Zeaxanthin; NA =Not applicated; NR =Not reported. Bolded data show significant changes over time (p <0.05). * Data significantly different from placebo group (p <0.05). M.-C. García-Romera et al. Physiology & Behavior 254 (2022) 113891 8 Table 3 Relationship between MPOD and Cognitive Function considering different tests. First author, year Age (mean ±SD) Sex (% female) TEST Battery CF (task for calculating the index score) Relationship between MPOD and CF Saint, 2018 10(±4.26) 45.5% WJ-III Brief Intellectual Ability (BIA) Verbal Ability Cognitive Efficiency Processing Speed Executive Processes Visual-Auditory Learning Spatial Relations WJ-III Composite Scores: MPOD vs *BIA r =0.268 (p<0.05) Verbal Ability r =0.159 (p>0.05) *Cognitive Efficiency r =0.206 (p ≤ 0.10)) Processing Speed r =0.099 (p>0.05) Executive Processes r =0.288 (p <0.05) Select WJ-III Subtests considering sex differences): *Visual-Auditory Learning r =0.236 (p ≤0.10) *Spatial Relations r =0.299 (p<0.05)) Barnett, 2018 8.8 (±0,1) 70% WJ KTEA IQ (WJ) Academic performance: math concepts, letter and word recognition, reading comprehension, word recognition fluency, written expression, listening comprehension (KTEA) MPOD vs *academic achievement (R 2 = 0.10. P <0.01) *mathematics (R 2 =0.07. P=0.02) *math concepts (R 2 =0.05. P=0.04) *math computation (R 2 =0.09. P= 0.02) *written language composite standard scores (R 2 =0.15. P <0.01) * written expression (R 2 =0.11. P= 0.008) *spelling (R 2 =0.13. P=0.004) *reading or reading fluency improve (p<0.05) listening comprehension did not improve Hassevoort, 2017 8.8 (±0,11) 63% WJ IQ Memory performance (WJ-) MPOD improve IQ (p>0.05) *relational memory (p<0.05) Walk, Khan, et al., 2017 8.69 (±0,08) 80.61% K-TEA EEG IQ Educational achievement (K-TEA) Higher MPOD improve *accurate performance: (higher MPOD group: M =82.242) vs (lower MPOD group: M =75.545) -for congruent trials (F=0.566. p= 0.456. η p 2 =0.012). -*for incongruent trials (F=4.623. p= 0.037. η p 2 =0.091) Reaction times revealed no significant effects involving MPOD. Lower MPOD improve *P3 amplitudes: (higher MPOD: M =14.247) vs (lower MPOD: M =18.535) -(F (1. 46) =5.287. p=0.026. η p 2 = 0.103) Stringham, 2019 21.5 54.20% CNSVS Composite memory (N/R) Verbal memory (N/R) Visual memory (N/R) Working memory (N/R) Psychomotor Speed (N/R) Processing speed (N/R) Attention (N/R) Reasoning (N/R) Executive function (N/R) Cognitive flexibility (N/R) Social acuity (N/R) Reaction time (N/R) MPOD vs memory (p>0.05) *psychomotor speed (r =0.38; p = 0.003) *processing speed (r =0.35; p =0.007) MPOD vs the rest of cognitive functions N/R Hammond, 2017 Total: 73.74 (±8.20) Active:72.51 (±6.24) Placebo: 70.93 (±5.70) 58.82% Active: 52.77% Placebo: 73.33% CNSVS Verbal Memory (Verbal Memory Test) Visual Memory (Visual Memory Test) Psychomotor Speed (FTT, SDC) Attention (ST, SAT, CPT) Reasoning (NVRT) Executive function (SAT) Cognitive flexibility (SAT, ST) MPOD vs *Verbal Memory (r=0.31. p= 0.07) *Visual Memory (r=0.24. p=0.09) Psychomotor Speed (N/R) *Reasoning (r=0.45. p=0.04) *Executive function (errors of attention) (r = − 0.18. p=0.08) Cognitive flexibility (r=0.20. p=0.10) Neurocognitive Index (N/R) Renzi-Hammond, 2017 21.21 (±2.52) 43% Active: 43.24% Placebo: 42.85% CNSVS Verbal Memory (Verbal Memory Test) Visual Memory (Visual Memory Test) Psychomotor Speed (FTT, SDC) Attention (ST, SAT, CPT) Reasoning (NVRT) Executive function (SAT)) Cognitive flexibility (SAT, ST) MPOD improvements in: Verbal memory (p>0.05) *Visual memory (p <0.05) *Complex attention (p <0.04) *Reasoning ability (p <0.05) Executive function (p>0.05) Cognitive flexibility (p>0.05) (continued on next page) M.-C. García-Romera et al. Physiology & Behavior 254 (2022) 113891 9 Table 3 (continued) First author, year Age (mean ±SD) Sex (% female) TEST Battery CF (task for calculating the index score) Relationship between MPOD and CF Scott, 2017 Active: 63.3 (±11.1) Placebo: 62.5 (±9.2) 62,5% Active: 70% Placebo: 55% CANTAB Memory (PAL) Spatial Working Memory (SWM) Processing speed Attention Delayed Match to Sample (DMS) Spatial Span (SSP) Spatial Span Reverse (SSP-R) MPOD vs Memory (Avocado group): *Spatial Working Memory (r=0.46. p= 0.041) *Efficiency in approaching a problem (r=0.47. p=0.036) Processing speed N/R Attention N/R Delayed Match to Sample N/R Spatial Span N/R Spatial Span Reverse N/R Power, 2018 45.42 (±12.40) 48.4% Active: NR % Placebo: NR% CANTAB Verbal Memory (VRM) Visual memory (PAL) Processing speed =Comprehension (MOT) Phonemic fluency and semantic fluency (“FAS” and “Animal” tests) Executive function (AST) MPOD vs *Verbal Memory (reduction in intrusion errors)(r=–0.306; p=0.033) * Visual memory (reduction in total errors) (r=–0.342; p=0.005) Processing speed =Comprehension N/R Phonemic fluency and semantic fluency N/R Executive function N/R Edwards, 2020 Placebo: 34.0 (±6.2) Active: 34.6 (±5.7) 57% Active: 48.93% Placebo: 48.64% KBIT-2 EEG IQ Processing speed Attention (Flanker task, Oddball task) Inhibition (Nogo task) (EEG) MPOD vs cognition: No relationship (p>0.05) Lindbergh, 2018 Placebo: 70.43 (±5.43) Active: 72.43 (±6.48) 59.1% Active: 53.33% Placebo: 71.43% WTAR fMRI - adapted task Memory (WTAR) Neural mechanisms (fMRI) MPOD vs estimated intellectual functioning (memory) No relation (p>0.05) Vishwanathan, 2014 77.6 ±0,3 50.9% 3MS SRT: learn SRT: delayed recall Reaction time Verbal fluency Digit-symbol substitution task Box drawing task Pattern comparison task Attention, language and orientation (3MS) Verbal learning and memory (SRT) Reaction time Verbal fluency Speed and associative learning ability (Digitsymbol substitution task) Sensory motor speed (Box drawing task) Perceptual speed (Pattern comparison task) Relationship between cognitive function measures and MPOD: 3MS: 0.269 (P≤0.05) SRT: learn: 0.263 (P≤0.05) SRT: delayed recall: 0.220 (P≤0.05) Reaction time: −0.059 Verbal fluency: 0.249 (P≤0.05) Digit-symbol substitution task: 0.249 (P≤0.05) Box drawing task: 0.154 Pattern comparison task: 0.195 (P≤ 0.05) Ajana, 2018 82.3 (±4,3) 68.5% MMSE FCSRT BVRT IST15 Global cognitive functioning (MMSE) Memory performance and verbal learning (FCSRT) Visual memory and visual perception (BVRT) Verbal fluency abilities and speed of verbal production (IST15) MPOD improvements in : *Memory performance and verbal learning (β=0.15, p<0.05) * Visual memory and visual perception (β=0.39, p<0.05) * Verbal fluency abilities and speed of verbal production (β=1.16, p<0.05) Cannvavale, 2019 34.9 (±6.1) 45.54% KBIT-2 IQ (KBIT-2) Memory ability (Computerized spatial reconstruction task) MPOD and Memory: 0.438 ±0.20 (p> 0.05) Misplacement: −0.110 (p >0.05) Object-Location Binding: 0.083 (p > 0.05) Khan, 2018 34.6 (±6.1) 61% KBIT IQ (KBIT-2) Fluid and crystallized intelligence MPOD vs *IQ (β =0.20, p=0.04) * Fluid intelligence (β =0.20, p=0.03) Crystallized intelligence (β =0.11, p= 0.25) Edwards, 2019 34.98 (±5,85) 69.30% KBIT-2 EEG IQ (KBIT-2) Processing speed Attention (Oddball task) (EEG) MPOD improve IQ: r =0.18 (p>0.05) and *Processing speed (p<0.05) MPOD was inversely related to *Attention (attentional demands) (p<0.05) Walk, Edwards, et al., 2017 33.8 (±5.7) 71% KBIT EEG IQ (KBIT) Cognitive Tasks: Attentional inhibition Selective attention (oddball task) Inhibition (A go/no-go task) (EEG) MPOD vs *IQ: the response accuracy to standard stimuli (r=0.342. p=0.007) reaction time to target stimuli (r= 0.468. p≤0.001). Cognitive Tasks: *Attentional inhibition: incongruent trials (r =–0.306. p= 0.018) peak amplitude for incongruent trials (r=0.259. p=0.045) *Selective attention: (continued on next page) M.-C. García-Romera et al.