Can Curcumin Counteract Cognitive Decline? Clinical Trial Evidence and Rationale for Combining ω-3 Fatty Acids with Curcumin

Abstract

The rate of cognitive decline in the elderly is highly variable. One potential factor contributing to accelerated cognitive decline is chronic systemic inflammation, because it has been linked to cognitive impairment and increased dementia risk. Certain lifestyle factors, such as excess body weight and sedentary behavior, can exacerbate a proinflammatory state in older adults, resulting in chronic low-grade inflammation. Supplementing the diet with curcumin, an anti-inflammatory polyphenolic compound from the curry spice turmeric, is a potential approach to prevent accelerated cognitive decline by counteracting chronic inflammatory processes. Although the anti-inflammatory effects of curcumin are well established, the potential cognitive benefits of curcumin were discovered more recently. Several animal and epidemiologic studies on the effect of curcumin supplementation on cognition showed promising results; however, randomized controlled trials in humans are limited. In this review, we identified 5 randomized controlled trials, of which only 2 observed a beneficial effect of curcumin supplementation on cognition by improving working memory. By critically examining the methodologies of those studies, we identified some limitations, one of which is that none of the studies explored the possibility that anti-inflammatory mechanisms were mediating cognitive benefits (i.e., no study tested participants with low-grade inflammation or measured inflammatory biomarkers). Other factors influencing the likelihood of conclusive outcomes include choice of study population (cognitively unimpaired compared with impaired), study duration, curcumin dose and its bioavailability, and neurocognitive test battery. On the basis of these findings, we offer recommendations for future studies to examine the potential cognitive benefits of curcumin in humans, which include evaluating its effects on cerebral endothelial vasodilator function and boosting its cognitive effects by combining it with long-chain omega-3 (n–3) fatty acids.

Introduction

Aging sees a decline in cognitive abilities, such as conceptual reasoning, memory, processing speed, and executive function, whereas other abilities, such as language, are preserved (1). The rate of cognitive decline, however, is highly variable between individuals. Hence, it is important to distinguish between normal age-related cognitive decline and more rapid disease-related cognitive decline (1, 2). One possible factor contributing to the latter is chronic systemic inflammation. Increasing evidence shows a link between proinflammatory profiles and cognitive impairment and risk of dementia (3). Aging is associated with heightened and prolonged inflammation (3), which can be exacerbated by certain lifestyle factors such as sedentary behavior and excess body weight, particularly visceral adiposity, leading to chronic low-grade inflammation (4). Adipocytes are not only energy storage depots but also act as an endocrine organ, expressing proinflammatory proteins and recruiting macrophages in the absence of any acute injury (5). Overweight and obese adults have been shown to exhibit impairments in executive function and short-term memory compared with adults with a normal BMI, independent of age (6). Furthermore, the risk of developing dementia is increased by 35% in individuals who are overweight and by 74% in individuals who are obese at midlife (40–45 y) (7). Sedentary behavior also contributes to the inflammation milieu through increased release of proinflammatory proteins such as IL-6 and C-reactive protein (CRP) (4), thus further heightening the inflammatory state of sedentary, obese older adults.

Chronic systemic inflammation and aging exaggerate endothelial dysfunction, including increased blood-brain barrier (BBB) permeability, resulting in extravasation of peripheral monocytes into the perivascular spaces, as shown in recent studies involving high-fat-diet–fed obese rat models (8). This, in turn, stimulates microglial activity. Upon activation, these innate immune cells of the central nervous system undergo significant proliferation, which can aggressively phagocytose synaptic components, thus affecting synaptic and neuronal plasticity (9). Systemic inflammation accelerates the aging of microglias, which can increase their hyperactive response (10). Cumulative hyperactivation or dysfunction of the microglia can accelerate neuronal damage through extensive synaptic remodeling and release of neurotoxic molecules, such as proinflammatory factors, reactive oxygen species, and proteinases, contributing to further BBB breakdown (11, 12). The hippocampus and basal ganglia—which are important structures for memory, attention, emotion, and perception—contain more enzymes involved in inflammatory processes; therefore, these structures are at increased risk of damage from neuroinflammation (3). It should be noted that systemic inflammation not only disrupts the BBB itself but also alters the function of certain BBB transporters, which can disrupt, for instance, the clearance of amyloid-β (Aβ) protein, promoting its accumulation and formation of amyloid plaques, which is one of the pathological hallmarks of Alzheimer disease (AD) (13, 14). Amyloid plaques can further enhance microglial activation and induce secretion of proinflammatory cytokines, exacerbating the neuroinflammatory response [reviewed in more detail elsewhere (15)]. Furthermore, chronic systemic inflammation can also impair endothelium-mediated vasodilatation of cerebral vessels, resulting in a chronic state of hypoperfusion, which has been linked to the severity of cognitive impairment in the elderly (16, 17). Thus, neuroinflammation can have far-reaching deleterious consequences for neuronal, cerebrovascular, and consequently, cognitive function.

Epidemiologic studies suggest that mild to moderate use of nonsteroidal anti-inflammatory drugs might delay the onset of dementia (3). However, chronic use of nonsteroidal anti-inflammatory drugs has also been shown to increase the risk of acute renal failure, stroke, and myocardial infarction in older adults (18). Hence, an alternative approach to counteract chronic inflammatory processes is urgently required in order to prevent accelerated disease-related cognitive decline. One potential approach is to increase the consumption of anti-inflammatory dietary bioactives, such as curcumin, a polyphenolic compound that is concentrated in the curry spice turmeric (Curcuma longa). Curcumin (diferuloylmethane) has been used for centuries in Asia as an herbal medicine due to its ability to reduce pain associated with inflammatory conditions such as mild rheumatism and joint pain as well as skin disorders such as psoriasis (19). There is substantial evidence for the anti-inflammatory effects of curcumin from preclinical and clinical studies (20). In humans, curcumin has been shown to significantly decrease the concentration of proinflammatory cytokines such as TNF-α, IL-6, TGF-β, and monocyte chemoattractant protein 1 (MCP-1) (21, 22). Furthermore, Panahi et al. (23) showed that supplementing the diet with curcumin (1 g/d for 8 wk) improved antioxidant capacity, decreased lipid peroxidation, and reduced CRP concentrations in 117 adults with metabolic syndrome.

In animal models, curcumin has been shown to not only reduce systemic inflammatory markers but also to improve cognitive function. In a mouse model of midlife obesity, curcumin supplementation (1000 mg ⋅ kg⁻¹ ⋅ d⁻¹) for 12 wk resulted in reduced CRP concentrations as well as an improved executive function, suggesting a possible link (24). However, a direct link between improvements in inflammatory status and cognitive benefits due to curcumin still needs to be established. Nevertheless, several animal studies have shown curcumin’s cognitive benefits. For instance, curcumin increased memory acquisition ability and enhanced nonspatial memory in aged mice and rats (25, 26) and improved working and cued memory functions in a more advanced stage of cognitive impairment (mouse model of AD) (27). In humans, epidemiologic studies also suggest a link between curcumin consumption and cognitive benefits. Ng et al. (28) observed that subjects with higher amounts of curry consumption had 6% higher Mini-Mental State Examination (MMSE) scores compared with subjects who never or rarely consume curry.

Although the results from animal and epidemiologic studies are promising, the number of human intervention trials is still very limited, because the use of curcumin for human cognitive benefits is a relatively new concept. The aim of this review was as follows: 1) to provide an overview of the existing intervention trials examining the effects of curcumin on cognitive function in humans, 2) to critically appraise the studies’ methodologies and findings, 3) to provide a potential approach to enhance curcumin's cognitive effects, and 4) to make recommendations for future intervention studies.

Can Curcumin Counteract Cognitive Decline?

Current evidence from clinical trials

We identified 5 parallel-design randomized controlled trials examining the effect of curcumin on cognition in humans, published between 2008 and 2016. Only 2 of the 5 studies found an effect on cognition. We also identified 1 case-control study (29); however, we will report no further on this study because it focuses on the effects of curcumin on behavioral and psychological symptoms of dementia in 3 patients with AD rather than on cognitive function itself.

The first study was conducted in 2008 by Baum et al. (30) in 27 older adults with AD who were given 1000 mg or 4000 mg curcumin/d (encapsulated or in powder form, percentage of curcumin not reported) or placebo for 6 mo (Table 1). The participants’ MMSE score did not significantly change when comparing either of the treatment groups with placebo.

TABLE 1.

Parallel-design randomized controlled trials examining the effects of curcumin on cognitive function¹

Study (ref),		Sample
year (country)	Participants	size, n	Dose	Duration	Main findings2
Baum et al. (30), 2008 (China)	Adults with AD; mean age: 78 y	27	1000 mg or 4000 mg curcumin/d (as capsule or powder), or placebo	6 mo	No significant change in global cognitive function (MMSE score)
Ringman et al. (31), 2012 (USA)	Adults with mild to moderate AD; mean age: 74 y	30	2000 mg or 4000 mg curcumin/d (curcumin C3 complex), or placebo	6 mo	No significant change in global cognitive function (MMSE) and ADAS-cog
Lee et al. (32), 2014 (Taiwan)	Adults with untreated prediabetes; mean age: 73 y	48	1 g turmeric, 2 g cinnamon, combination of both, or placebo	Single dose, 6 h	Improved working memory (P = 0.050) after turmeric consumption
Cox et al. (33), 2015 (Australia)	Healthy adults; mean age: 68 y	60	400 mg curcumin/d [Longvida (Verdure Science)] or placebo	Single dose, 1 h and 3 h (acute) and 4 wk (chronic)	Acute: trend in improvement in sustained attention (P = 0.168) and significant improvement in working memory (P = 0.033); chronic: significant improvement in working memory (P = 0.044)
Rainey-Smith et al. (34), 2016 (Australia)	Healthy adults; mean age: 66 y	96	1500 mg curcumin/d [Biocurcumax (Sabinsa Corporation)] or placebo	12 mo (interim assessment at 6 mo)	No significant difference in cognitive performance (RAVLT, COWAT, WDSS, and CogState battery of assessments); no significant time × treatment group interaction for global cognitive function (MoCA test) by using baseline and 12-mo performance

¹AD, Alzheimer disease; ADAS-cog, Alzheimer's Disease Assessment Scale, cognitive subportion; COWAT, Controlled Oral Word Association Test; MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; RAVLT, Rey Auditory Verbal Learning Test; ref, reference; WDSS, Wechsler Digit Symbol Scale.

²Compared with placebo.

Another study in patients with AD was conducted by Ringman et al. (31) in 2012. Thirty older adults with mild to moderate AD consumed 2000 mg or 4000 mg curcumin C3 complex/d (70–80% curcumin) or placebo for 6 mo. In addition to the MMSE, this study also included the Alzheimer Disease Assessment Scale, cognitive subportion (ADAS-cog), but no significant effect of curcumin on cognition was observed.

The study by Lee et al. (32) was performed in 48 adults with untreated prediabetes who were supplemented with a single dose of either 1 g turmeric (percentage of curcumin not reported), 2 g cinnamon, a combination of turmeric and cinnamon, or placebo. At baseline and 6 h after supplementation, working memory was assessed with the N-Back test. Individuals supplemented with turmeric were shown to have a significantly improved working memory.

The study by Cox et al. (33) was the first study, to our knowledge, to investigate the effect of curcumin on cognitive performance in healthy older adults, 60 of whom were supplemented with 400 mg Longvida/d (80 mg curcumin/d, Verdure Science (Noblesville, IN, USA)) or placebo. A single dose of curcumin tended to enhance sustained attention and significantly improved working memory. The significant improvement in working memory was sustained after chronic supplementation with curcumin for 4 wk.

A more recent study was conducted in 2016 by Rainey-Smith et al. (34), who supplemented 96 healthy adults with 1500 mg Biocurcumax/d (containing 88% total curcuminoids, Sabinsa Corporation (Piscataway, NJ, USA)) or placebo for 12 mo. Cognitive function was measured at baseline and after 6 and 12 mo. This study did not find any significant changes in cognitive function, measured by (non)computerized composite scores, after curcumin supplementation compared with placebo. For the global cognitive test [Montreal Cognitive Assessment (MoCA)], a significant time × treatment group interaction was observed. However, this significant outcome was mainly due to a decrease in MoCA score in the placebo group after 6 mo, which was not observed in the treatment group. No difference in MoCA score was observed between treatment and placebo groups after 12 mo.

Lessons learned from clinical trials

Curcumin and its potential effects on cognitive function have only fairly recently (<15 y) gained attention. Therefore, the number of randomized controlled trials evaluating the effects of curcumin on cognition in humans is very limited (n = 5). On the basis of the existing evidence, acute and short-term (4 wk) curcumin supplementation seems to enhance cognitive function by improving working memory, but no effect on delaying cognitive decline has been observed in the longer-term studies (6–12 mo). The mechanisms underlying the effects of curcumin on cognition are mostly speculative and based on animal studies. All 5 studies mentioned the anti-inflammatory effects of curcumin and therefore its potential benefit for improving cognitive function, but none of the studies included participants who were likely to have chronic low-grade inflammation and only Cox et al. (33) actually measured inflammatory markers. However, because their cohort was healthy, most inflammatory markers at baseline were below the level of quantification. Thus, a major limitation is that none of the trials explored the mechanistic link proposed to mediate cognitive benefits of curcumin. Apart from this, there are several possible reasons for the lack of significant effects of curcumin on cognition in 3 of the 5 studies.

Study population

The choice of study population might be one possible reason why the majority of the studies did not find significant cognitive benefits after curcumin supplementation. The first 2 studies conducted by Baum et al. (30) and Ringman et al. (31), which looked at effects of curcumin on delaying cognitive decline, were exploratory and performed in patients with AD. Neither study observed a change in MMSE score after 6 mo of curcumin supplementation when compared with placebo. In both studies, most of the participants were taking AD medication, possibly already attenuating the progression of the disease, making it more difficult to see any potential effect of curcumin supplementation. Another possibility is that the underlying pathology of AD might have been too advanced for curcumin to be able to reverse damage to the BBB, Aβ accumulation, and impairments in vasodilator function (13, 14, 35). Therefore, future studies should consider evaluating curcumin in older adults who may experience some subjective memory deficits due to age-related cognitive decline but do not yet suffer from dementia.

Duration of the trial and choice of neurocognitive tests

Another major difference between the studies that observed a significant effect of curcumin supplementation on cognition and those that did not is the study aim and consequently the study duration. Lee et al. (32) and Cox et al. (33) aimed to examine if curcumin can improve cognitive function, by looking at acute (1–6 h) and short-term (4 wk) effects of curcumin supplementation, and observed improvements in working memory. The studies by Baum et al. (30), Ringman et al. (31), and Rainey-Smith et al. (34) were of longer duration (6–12 mo), because they aimed to see whether curcumin could delay cognitive decline. None of these studies found a significant effect of curcumin. However, the study period was probably insufficient to detect any change in cognitive decline, measured by the change in MMSE or MoCA score, because no cognitive decline was observed in the placebo group. Another choice of cognitive test or a more comprehensive cognitive test battery examining certain cognitive domains rather than global cognition alone might have been more sensitive to detect any cognitive decline and any possible effect of curcumin supplementation in the Baum et al. (30) and Ringman et al. (31) studies. Rainey-Smith et al. (34) did use a more comprehensive cognitive test battery, examining episodic verbal memory, executive function, working memory, attention, and verbal fluency in addition to global cognition. However, the lack of cognitive results in this study might be due to other reasons, such as the high drop-out rate and differences between groups at baseline. The placebo group performed significantly better at baseline in the MoCA, digit symbol task, and the noncomputerized tests (pooled together with the noncomputerized composite score), which might have masked any effects of curcumin supplementation on cognition.

Curcumin dose and attrition

In addition to the study duration, another major difference between the studies by Lee et al. (32) and Cox et al. (33) compared with the studies by Baum et al. (30), Ringman et al. (31), and Rainey-Smith et al. (34) was the curcumin dose. The former 2 studies used lower curcumin doses of 400 and 1000 mg/d, respectively, and it might be possible that a lower dose is more efficient in yielding cognitive benefits. Furthermore, higher doses of curcumin seem to cause more gastrointestinal problems, resulting in more withdrawals during the study. Cox et al. (33) did not have any withdrawals during the 4-wk trial after randomization, whereas Rainey-Smith et al. (34) had a 25% drop-out rate due to adverse events, mainly gastrointestinal. The high drop-out rate might have affected their statistical power and could have resulted in unbalanced groups, both of which might explain the lack of significant results.

Bioavailability of curcumin

The poor bioavailability of curcumin might have also contributed to the lack of significant results in the majority of the studies. It is known that curcumin is hydrophobic and has a low intestinal absorption, low tissue distribution, rapid metabolism, and a rapid systemic clearance (19). Another problem is glucuronidation of curcumin, because curcumin needs to be unconjugated to cross the BBB and elicit effects in the brain (36). Only 2 of the 5 studies included in this review actually measured curcumin concentrations in plasma. Baum et al. (30) detected no free curcumin in plasma, indicating that nearly all of the curcumin was glucuronidated. Ringman et al. (31) used a more sensitive method to measure curcumin and also found much higher glucuronidated curcumin concentrations compared with native curcumin concentrations in plasma 3 h after administration at week 24. Another human trial in which 6 volunteers took 2 g curcumin together with 5 mg piperine—an inhibitor of hepatic and intestinal glucuronidation—showed a doubling of curcumin absorption (37), indicating that glucuronidation is at least partly responsible for the low bioavailability of curcumin.

This raises the need for more efficient curcumin formulations. Cox et al. (33) used a new curcumin formula with enhanced bioavailability. Longvida curcumin uses the solid lipid curcumin particle technology to deliver curcumin to target tissues. This technology dissolves curcumin in a lipid particle, thereby increasing its stability by protecting it from degradation in gastric acid and increasing its solubility by enhancing the absorption of curcumin (38). Gota et al. (39) tested the bioavailability of Longvida curcumin by supplementing 11 patients with osteosarcoma with either Longvida curcumin or a 95% curcuminoid extract. They showed that 2 h after a 650-mg administration of Longvida curcumin, plasma concentrations increased to 20 ng/mL, whereas no curcumin was detected in plasma after 650 mg of the 95% curcuminoid extract.

Thus, in addition to considering study population, study duration, curcumin dose, and the choice of neurocognitive tests, future intervention trials should use an optimized curcumin formulation, such as Longvida, which can enhance bioavailability, in order to increase curcumin's potential to improve cognitive function.

Can Long-Chain n–3 PUFAs Enhance Effects of Curcumin?

Another potential approach to increase curcumin’s efficacy might be to combine curcumin with other bioactive nutrients. One such combination is with long-chain (LC) n–3 PUFAs, such as EPA or DHA, which are found in marine sources such as fish oil and seafood, because they also have anti-inflammatory properties and have been shown to improve certain cognitive domains (40, 41). This combination could exert additional or complementary effects, such as was shown by Kakoti et al. (42), who combined LC n–3 PUFAs with resveratrol and showed synergistic effects on inflammation. Both curcumin and LC n–3 PUFAs also have vasodilator effects, which might further improve cognitive performance. Interestingly, a study in rats has shown that curcumin can boost DHA content in the liver and brain by increasing the synthesis of DHA from the n–3 precursor α-linoleic acid (18:3n–3) (43). Thus, there are multiple mechanisms by which curcumin could act in combination with LC n–3 PUFAs or other bioactive nutrients to enhance cognitive performance.

Potential mechanisms

Anti-inflammatory effects

Curcumin is a potent inhibitor of NF-κB signaling—a transcription factor involved in the upregulation of inflammatory mediators—by inhibiting IκB kinase β (IKKβ), thereby blocking further downstream activation (44, 45). This enables it to inhibit expression of cyclooxygenase 2 (COX-2) and proinflammatory cytokines such as IL-1 and IL-6. Moreover, inflammatory signaling is reduced due to curcumin inhibiting the binding of activator protein 1 to DNA. Curcumin has also been shown to lower oxidative stress and reduce the concentrations of oxidized proteins (23, 46). Furthermore, curcumin can affect Aβ protein plaques, a potential site of origin of chronic inflammatory processes in the brain. Aggregation of Aβ protein, forming amyloid plaques, starts decades before the first clinical symptoms of cognitive impairment appear. Curcumin has been shown to reduce insoluble Aβ concentrations, prevent Aβ aggregation, and reduce Aβ plaque burden in animal studies (47). However, the studies by Baum et al. (30) and Ringman et al. (31) measured Aβ concentrations in blood and cerebrospinal fluid of patients with AD and did not find any significant change after supplementation. One possibility might be that, in humans, curcumin can prevent Aβ aggregation at an early stage but is unable to reduce Aβ plaque burden in established AD; however, this needs to be further evaluated.

LC n–3 PUFAs can also inhibit NF-κB by decreasing phosphorylation of IκB, which is required for NF-κB to dissociate from IκB and translocate to the nucleus (46). However, the anti-inflammatory effects of LC n–3 PUFAs are mostly mediated by changes in FA composition of cell membranes, resulting in altered gene expression and patterns of lipid mediator production (40). An increased intake of LC n–3 PUFAs results in a decreased incorporation of n–6 arachidonic acid (AA) into cell membranes. Both AA and EPA produce eicosanoids, which are signaling molecules. However, eicosanoids produced from AA have a high proinflammatory profile, whereas those produced from EPA are far less potent and exhibit a low proinflammatory profile. Moreover, EPA and DHA give rise to resolvins, which are anti-inflammatory and have inflammation-resolving properties by preventing neutrophil infiltration at sites of inflammation and inhibiting the proinflammatory cytokines IL-1β and TNF-α. LC n–3 PUFAs can also activate PPAR-γ, which possesses potent anti-inflammatory properties (40).

Several studies have shown negative effects of proinflammatory cytokines on cognition, suggesting a link between chronic systematic inflammation and cognitive deficits. For instance, a 2-y observational study in elderly participants by Yaffe et al. (48) showed that those in the highest tertiles of IL-6 and CRP had significantly lower MMSE scores at baseline as well as at follow-up and had a 24% increased risk of cognitive decline compared with participants in the lowest tertiles. Reducing systemic inflammatory processes via nutrient supplementation might help to prevent or reverse damage to the BBB as well as inhibit activation of microglia, thereby preventing the release of neurotoxic substances leading to neuroinflammation and neurodegradation, and consequently counteracting accelerated cognitive decline. Both curcumin and LC n–3 PUFAs can inhibit inflammatory processes, but act via different mechanisms to resolve inflammation. Thus, their combination might help to alleviate inflammation more efficiently and thereby elicit a greater beneficial effect on cognitive function.

Effects on endothelial vasodilator function

Several studies have examined the effect of LC n–3 PUFA supplementation on endothelial vasodilator function, with the majority observing an increase in flow-mediated dilatation (FMD) (49). This effect is probably multifactorial, because LC n–3 PUFAs can increase the production of NO and prostacyclin, both endothelial vasodilators, and inhibit production of thromboxane A₂, a potent vasoconstrictor and facilitator of platelet aggregation (50, 51).

Studies on the effect of curcumin on vasodilation, although limited, also showed an increase in FMD after supplementation (52, 53). The study by Pu et al. (54) showed that, in rats, curcumin supplementation restored impaired cerebrovascular vasodilatation via the AMP kinase (AMPK) pathway. Curcumin was shown to promote AMPK phosphorylation, leading to increased endothelial NO synthase phosphorylation and ultimately resulting in increased NO production and bioavailability (55). Increased NO bioavailability was also seen in a human study performed by Santos-Parker et al. (56) in 2017, in which 39 healthy older adults were supplemented with 2000 mg Longvida/d (400 mg curcumin/d) for 12 wk. They showed a 36% increase in FMD in the treatment group, whereas the placebo group showed no change in FMD. Next to the increased NO bioavailability, the authors suggest that the increase in FMD is also due to the observed reduction in oxidative stress.

Inflammation might have not only a direct negative effect on cognition but also an indirect effect by promoting endothelial dysfunction, which has been suggested to contribute to an accelerated cognitive decline and progression of dementia (57–59). The link between endothelial dysfunction and cognitive decline might be mediated by an impaired cerebrovascular responsiveness to cognitive demands, leading to an insufficient blood supply to active brain regions (60). Poor cerebrovascular function is suggested to play an important role in the initiation and progression of dementia (16). A hallmark of endothelial dysfunction is reduced NO bioavailability, resulting in impaired vasodilator function (61). Curcumin and LC n–3 PUFAs have both been shown to improve FMD by increasing NO availability. However, both nutrients also have other pathways (modulation of thromboxane production or inhibition of oxidative stress) that can result in an improved endothelial vasodilator function, raising the possibility of an additive effect of the nutrient combination on vasodilator function.

Taken together, both inflammation and impaired vasodilator function seem to play an important role in the process of cognitive decline and progression to dementia. Curcumin and LC n–3 PUFAs both possess anti-inflammatory effects and are able to enhance systemic endothelial vasodilator function. The different pathways used by each nutrient to achieve these effects may lead to an additive benefit of their combination on cognitive function, especially in populations with chronic low-grade inflammation (e.g. sedentary, overweight or obese older adults).

Future Perspectives

Gaps in the literature and implications for future studies

A major gap in the literature is that, so far, no intervention trial has tried to link the anti-inflammatory effects of curcumin with the potential beneficial cognitive effects; therefore, future studies in this field are needed. Furthermore, the potential for curcumin to improve vasodilator function should be assessed in not only the systemic but also the cerebral circulation. If curcumin is able to improve endothelial function equally in cerebral and systemic circulations, this might be another potential mechanism explaining cognitive benefits. Such studies should carefully consider the study population, study duration, curcumin dose and formulation, and choice of cognitive test battery.

Participants most likely to benefit from curcumin supplementation might be those with chronic low-grade inflammation who are prone to accelerated cognitive decline, such as older overweight or obese and sedentary adults.

Depending on the study outcome (i.e., cognitive benefits or delaying cognitive decline), future trials should consider the optimal study duration. For cognitive benefits, studies still need to validate short-term cognitive benefits as well as investigate longer-term effects to see whether curcumin has benefits in other cognitive domains in addition to working memory. To see whether curcumin can delay cognitive decline, future studies with longer study periods (>2 y) are needed so that a slower rate of decline can be detected in the treatment group compared with the placebo group.

Dose-response studies are required to identify the optimal curcumin dose to elicit cognitive benefits. It appears that lower curcumin doses might be favorable, because high curcumin doses (>1.5 g/d) are more likely to cause adverse effects, predominantly gastrointestinal problems (30, 31, 34). A study of 3 mo duration established an upper limit of 12 g/d for curcumin; whether this is valid for a longer time period is unknown (62). Lower curcumin doses might be more effective and result in fewer adverse events and therefore lower drop-out rates during trials. Future studies might also consider giving curcumin 2 times/d, in order to maintain a stable curcumin concentration in the blood, because curcumin has a rapid systematic clearance. Moreover, due to the low bioavailability of curcumin, more optimized curcumin formulations should be used and the curcumin concentration in blood plasma should be monitored. In addition, future studies should examine the combination of curcumin with other bioactive nutrients, such as LC n–3 PUFAs, to see if a combination can boost the efficacy of curcumin to improve cognitive function.

Furthermore, a comprehensive battery of cognitive tests should be used to examine effects of curcumin in different cognitive domains such as working memory, attention, executive function, language, and verbal fluency.

Testing the combination of curcumin and LC n–3 PUFAs

We are currently conducting a 16-wk, double-blind, placebo-controlled dietary intervention trial to investigate the effects of curcumin supplementation or LC n–3 PUFA supplementation, or both, on inflammatory, circulatory and cerebrovascular and cognitive outcomes in overweight or obese and sedentary older adults without cognitive impairment (ACTRN12616000732482p, https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=370788&isReview=true). Participants will be allocated to receive either LC n–3 PUFAs (total of 2.4 g/d), Longvida curcumin (total of 160 mg/d), or a combination of LC n–3 PUFAs and curcumin or placebo and will be instructed to take the supplements 2 times/d. We chose this particular study population because these adults are very likely to have chronic low-grade inflammation and several studies have shown impairments in systemic endothelial function (63), regional cerebral blood flow in the prefrontal cortex (64), and cerebrovascular reactivity in obese adults (65). These impairments are still reversible, as opposed to patients with AD who display cerebral arterial stiffness and vasodilator impairment at such an advanced stage that the endothelium is likely to be beyond repair (35). In this study, we will test the hypothesis that the combination of curcumin and LC n–3 PUFAs can restore cerebral endothelial function, partially mediated by a reduction in systemic inflammation, which, in turn, can enhance cognitive function.

Cerebral circulatory function will be assessed by transcranial Doppler ultrasound measurement of blood flow velocity in the middle cerebral artery at rest and during hypercapnia and cognitive tests, reflecting endothelium-dependent vasodilator responsiveness of the cerebral microvasculature to these physiologic and mental stimuli (i.e., neurovascular coupling capacity).

Cognitive outcomes include a comprehensive battery of cognitive tests assessing global cognition (Addenbrooke's Cognitive Examination III), working and episodic memory, processing speed, executive function, cognitive flexibility, inhibitory control, attention, and language.

At baseline and at 16 wk, a blood sample will be taken to assess metabolic and inflammatory markers as well as plasma curcumin and erythrocyte LC n–3 PUFAs (Omega-3 index).

This is the first study, to our knowledge, with a comprehensive battery of assessments to examine the following: 1) the relation between improvements in cerebrovascular function and cognitive performance after curcumin supplementation, 2) possible additive effects of the combination of LC n–3 PUFAs and curcumin on these outcomes, and 3) anti-inflammatory effects of LC n–3 PUFAs and curcumin in a population expected to have chronic low-grade inflammation and their relation with cerebrovascular function and cognitive performance.

Conclusions

The evidence for curcumin affecting cognition in humans is very limited and future studies are urgently needed to further investigate curcumin’s potential cognitive benefits. On the basis of existing trials, curcumin appears to have an acute and short-term beneficial effect on working memory. Whether curcumin can enhance other cognitive domains and delay cognitive decline has to be further investigated in longer-term intervention trials. The major limitation of the existing intervention trials is that they have not examined the proposed mechanistic link (i.e., that curcumin's anti-inflammatory effects contribute to its cognitive benefits). Future studies should explore this link and examine if curcumin can improve not only systemic but also cerebral endothelial function. Learning from the limitations of the existing 5 intervention trials, future trials need to carefully consider the study population, study duration (depending on whether the study outcome is cognitive enhancement or cognitive decline), curcumin dose and formulation, and choice of cognitive test battery. Furthermore, future studies should test combinations of curcumin with other bioactive nutrients such as LC n–3 PUFAs, which might help to boost curcumin's efficacy through different mechanisms to resolve inflammation and increase endothelial vasodilator function, leading to a possible additive effect on cognitive performance.

Notes

The authors reported no funding received for this review.

Author disclosures: JCK, RHXW, and PRCH, no conflicts of interest.

Abbreviations used:

AA

arachidonic acid

Aβ

amyloid-β

AD

Alzheimer disease

BBB

blood-brain barrier

CRP

C-reactive protein

FMD

flow-mediated dilatation

IKKβ

IκB kinase β

LC

long-chain

MMSE

Mini-Mental State Examination

MoCA

Montreal Cognitive Assessment