Literature search methods

For the purpose of the present review, partly because of the relative dearth of higher-level RCT evidence, we have appraised all forms of published research, even lower-level evidence sources, to ensure that the scientific and clinical implications of the available evidence might be synthesised into useful treatise, which accurately reflects what is known about this area, and what remains to be elucidated. The review was carried out in two stages. First, we identified all relevant published articles from human and animal studies which reported on the relationship between MP (lutein and/or zeaxanthin and/or meso-zeaxanthin) and diabetes (type 1 and type 2), up until the year 2019. The second part of the search involved identifying publications which investigated the relationship between the metabolic perturbations typically associated with diabetes, and type 2 diabetes in particular (for example, adiposity/dyslipidaemia) and MP. Pre-selected keywords including ‘lutein’, ‘zeaxanthin’, ‘macular pigment’, ‘diabetes’, ‘diabetes AND MP’, ‘diabetes AND lutein/zeaxanthin’, ‘BMI’, ‘body fat AND diabetes’, ‘adipose’, ‘high density lipoprotein AND diabetes’, ‘triglycerides AND diabetes’, ‘oxidative stress’, ‘inflammation’, ‘hypertension AND MP’, ‘insulin resistance’ and ‘hyperinsulinemia’ were entered into academic databases and search engines including PubMed, Google Scholar, Mendeley, Scopus, Cochrane Library and the ISRCTN registry to define our search of the literature relating to MP and diabetes up until 2019. Although many supporting publications were retrieved, in total, only eight animal studies (Table 1) and eleven human studies (Table 2) were included to help clarify our current understanding of the links between diabetes and MP. A total of thirteen papers on adiposity and MP (Table 3) and seven papers on MP and dyslipidaemia (Table 4) were included to help analyse the relationship between MPOD and the metabolic correlates of type 2 diabetes (adiposity and dyslipidaemia). The overall findings suggest that MP is lower in diabetes, type 2 diabetes in particular, and that supplementation with macular carotenoids and/or other antioxidants may confer protection against diabetic eye disease. To elucidate the causal mechanisms and metabolic perturbations which might possibly explain the lower MP levels observed in diabetes, we will first explore the condition diabetes mellitus itself, including its association with oxidative stress and inflammation; and subsequently present the evidence linking adiposity and dyslipidaemia with type 2 (or poorly controlled type 1) diabetes and MPOD.

Table 3. Summary of studies linking macular pigment optical density (MPOD) and body fat

MP, macular pigment; CAREDS, Carotenoids in Age-Related Eye Disease Study; TILDA, The Irish Longitudinal Study of Ageing; IQ, intelligence quotient; DXA, dual-energy X-ray absorptiometry.

Table 4. Summary of studies linking macular pigment optical density (MPOD) with serum lipids

HFP, heterochromatic flicker photometry; CAREDS, Carotenoids in Age-Related Eye Disease Study; MP, macular pigment; AMD, age-related macular degeneration; LDL-C, LDL-cholesterol; HDL-C, HDL-cholesterol; DiVFuSS, Diabetes Visual Function Supplement Study; RCT, randomised controlled trial; DR, diabetic retinopathy; NPDR, non-proliferative diabetic retinopathy.

Diabetes mellitus

Diabetes mellitus is a group of metabolic disorders caused by the complex interaction of genetics, environmental factors and lifestyle choices⁽Reference Polonsky³⁷⁾. Diabetes is characterised by a deficiency of insulin and/or systemic insulin resistance. Over the past number of decades, the number of individuals with diabetes, particularly type 2 diabetes, has increased dramatically, making it a critical and universal public health challenge⁽Reference Chen, Magliano and Zimmet³⁸⁾. Type 1 diabetes usually develops in normal-weight children, teenagers and younger adults, and is an autoimmune condition involving the selective destruction of pancreatic β-cells, ultimately resulting in complete deficiency of insulin⁽Reference Wilkin³⁹⁾. Conversely, type 2 diabetes is linked to a sedentary lifestyle and being overweight, and is characterised by systemic insulin resistance and subsequent pancreatic endocrine dysfunction, which results in attenuated insulin synthesis as well as inhibition of its cellular effects⁽Reference Polonsky³⁷⁾. Whilst diabetes is characterised by having higher than normal blood glucose levels, the pathogenesis and development of type 1 and type 2 diabetes differ; therefore, the relationship between MPOD and the different forms of diabetes should not be generalised. At presentation, type 2 diabetes is most often accompanied by other co-morbidities including overweight/obesity, insulin resistance, hypertension and dyslipidaemia, features which are less common in type 1 diabetes at diagnosis, but which may occur as complications of type 1 diabetes later in the course of the disease. Oxidative stress and inflammation are implicated in both type 1 and type 2 diabetes; however, research has shown that these metabolic disturbances are very pronounced in type 2 diabetes⁽Reference Kilhovd, Giardino and Torjesen^40–Reference Butkowski and Jelinek⁴⁴⁾. The oxidative stress, inflammation, adiposity and dyslipidaemia which characterise diabetes may have independent relationships with MPOD, and therefore constitute plausible causal mechanisms in diabetic eye disease. These pathological mechanisms merit further exploration and will be discussed in more detail herein.

Oxidative stress and diabetes

Chronic hyperglycaemia induces oxidative stress in patients with diabetes⁽Reference Robertson, Harmon and Tran⁴⁵⁾. Oxidative stress, defined as the excessive production of ROS, results in oxidative injury when the redox balance is upset; i.e. where the level of oxidative species exceeds the capacity of the antioxidant defence system to neutralise them⁽Reference McCord^46, Reference Kowluru and Chan⁴⁷⁾. ROS can include free radicals, which are partially reduced oxygen species containing one or more unpaired electrons (for example, superoxide anion or hydroxyl radical), and species with their full complement of electrons in an unstable or reactive state (for example, singlet oxygen or H₂O₂)⁽Reference Rösen, Nawroth and King⁴⁸⁾. The body’s natural defence against oxidative damage is neutralisation by endogenous antioxidants⁽Reference Sies⁴⁹⁾, which include enzymic antioxidants such as superoxide oxidase (SOD), catalase (CAT) and GPx, and non-enzymic antioxidants such as GSH. Our endogenous antioxidant defence system, however, is incomplete without exogenous antioxidant nutrients including vitamin C, vitamin E, carotenoids (β-carotene, lutein, zeaxanthin and meso-zeaxanthin) and polyphenols⁽Reference Bouayed and Bohn⁵⁰⁾. Exogenous antioxidants, such as those available in dietary fruit and vegetables, are therefore necessary to balance redox status. Through normal physiological processes, antioxidants inhibit or quench free radical reactions and can delay or inhibit cellular damage⁽Reference McCord⁴⁶⁾. When these processes are overwhelmed, however, as in type 2 diabetes, ROS will readily react with lipids, proteins and nucleic acids, resulting in impaired cell function or cell death⁽Reference Young and Woodside^51, Reference Ceriello⁵²⁾.

The retina is already at an increased risk of oxidative stress and damage, given its high oxygen demand and consumption, its exposure to visible light irradiation and its high concentration of PUFA in the photoreceptor outer segments⁽Reference Schalch⁵³⁾. These risks increase when the generation of free radicals and ROS is stimulated by environmental factors⁽Reference Machlin and Bendich^54, Reference Borish, Pryor and Venugopal⁵⁵⁾, short wavelength light exposure⁽Reference Algvere, Marshall and Seregard⁵⁶⁾, and other internal stresses including inflammation⁽Reference Zhang, Liu and Rojas⁵⁷⁾ and hyperglycaemia⁽Reference Ceriello^52, Reference Cohen, Riahi and Alpert^58, Reference Esposito, Nappo and Marfella⁵⁹⁾. While PUFA in the photoreceptor outer segment are thought to protect against oxidative and inflammatory damage, they can potentially act as a substrate for the propagation of free radicals as they provide a readily available source of electrons⁽Reference Beatty, Koh and Phil^9, Reference Machlin and Bendich⁵⁴⁾. With increasing age, systemic antioxidant levels decline and ROS levels increase in most tissues, and this is associated with a number of neurodegenerative diseases, including AMD and diabetes⁽Reference Beatty, Koh and Phil^9, Reference Davies and Morland^26, Reference Halliwell⁶⁰⁾.

Hyperglycaemia, oxidative stress and changes in redox homeostasis are fundamental events in the pathogenesis of diabetic retinopathy. Various mechanisms have been suggested to contribute to the formation of reactive oxygen-free radicals. Hyperglycaemia causes tissue damage through five major pathways, namely: increased flux of glucose and other sugars through the polyol pathway; increased intracellular formation of advanced glycation endproducts (AGE); increased expression of the receptor for AGE (RAGE) and its activating ligands; activation of protein kinase-C isoforms; and activity of the hexosamine pathway⁽Reference Brownlee⁶¹⁾. Over-activation of these pathways can lead to damage of cellular organelles and enzymes, increased lipid peroxidation and the development of insulin resistance⁽Reference Kahn^62, Reference Ceriello and Motz⁶³⁾. Several lines of evidence indicate that all five mechanisms are activated by a single upstream event: mitochondrial overproduction of ROS⁽Reference Brownlee⁶⁴⁾. Locally, the occurrence of these metabolic changes results in defective angiogenesis and the activation of a number of pro-inflammatory pathways which stimulate an influx of leucocytes which can alter vascular permeability in the retina⁽Reference Tang, Zhang and Jiang²¹⁾. The role of hyperglycaemia in diabetic microvascular complications in the eye has been established by large-scale prospective studies in type 2 diabetes (United Kingdom Prospective Diabetes Study)⁽⁶⁵⁾. Beyond elevations in blood glucose, however, ROS formation is also increased by elevations in NEFA (a common feature of type 2 diabetes), through their direct effect on mitochondria⁽Reference Nishikawa, Edelstein and Du⁶⁶⁾.

While oxidative stress is increased in diabetes, the activities of antioxidant defence enzymes responsible for scavenging free radicals and maintaining redox homeostasis (SOD, CAT, GPx) are concomitantly diminished in the retina of both type 1 and type 2 diabetes⁽Reference Kowluru and Chan^47, Reference Sies^49, Reference Matough, Budin and Hamid⁶⁷⁾. The activity of non-enzymic antioxidants is also depressed during hyperglycaemia-induced oxidative stress⁽Reference Matough, Budin and Hamid⁶⁷⁾. Redox balance is important, as its disruption can cause antioxidants to become pro-oxidative, potentially precipitating damage to the retina. It cannot be assumed, however, that a single reductive or antioxidant treatment would be sufficient to reverse the underlying pathophysiological mechanisms of diabetic retinopathy. Antioxidant protection in the retina is provided mainly from vitamins C and E, carotenoids, GPx, SOD and CAT enzymes, and GSH compound⁽Reference Cabrera and Chihuailaf⁶⁸⁾, and many of these antioxidants are aided by micronutrients (for example, Cu, Zn, Se) which act as cofactors⁽Reference Evans and Halliwell⁶⁹⁾. The carotenoids lutein and zeaxanthin are very effective quenchers of singlet molecular oxygen and lipid peroxy radicals; however, in the process, these carotenoids are themselves oxidised to their corresponding radical cations⁽Reference Roberts and Dennison⁷⁰⁾. These cations must be reduced to regenerate the original carotenoid, thereby allowing their reuse as an antioxidant. Vitamin E (α-tocopherol) reduces oxidised carotenoids, which leaves tocopherol oxidised⁽Reference Evans and Halliwell⁶⁹⁾. However, oxidised tocopherol can be reduced and regenerated by vitamin C (ascorbic acid) and GSH, which are themselves further reduced by Cu and Zn⁽Reference Chew, Clemons and SanGiovanni^15, Reference Sies, Stahl and Sundquist⁷¹⁾. Dietary lutein, zeaxanthin and vitamins E and C, as well as endogenous antioxidant enzymes, therefore act in concert with one another to lower the levels of ROS (via ROS detoxification) and to recycle oxidised lipid antioxidants (via re-reduction).

The presence of oxidation products of the macular carotenoids in human and primate retinal tissue indicates that lutein, zeaxanthin and meso-zeaxanthin play an important antioxidant role there⁽Reference Khachik, Bernstein and Garland⁷²⁾. Zeaxanthin appears to be a more potent antioxidant than lutein⁽Reference Cantrell, McGarvey and Truscott⁷³⁾, with meso-zeaxanthin even more efficacious, but only in conjunction with its binding protein⁽Reference Bhosale and Bernstein⁷⁴⁾. Li et al.⁽Reference Li, Ahmed and Bernstein⁷⁵⁾ demonstrated that a 1:1:1 ratio mixture of lutein, zeaxanthin and meso-zeaxanthin quenches more singlet oxygen than any of these carotenoids individually at the same total concentration⁽Reference Li, Ahmed and Bernstein^75, Reference Akuffo, Nolan and Howard⁷⁶⁾, suggesting that they are most effective when acting synergistically with one another. Diet is a major lifestyle factor which can greatly influence the incidence and progression of type 2 diabetes. Some dietary and lifestyle modifications associated with enhanced antioxidant supply could therefore be an effective prophylactic means to prevent or limit oxidative stress in diabetic eye disease. Type 2 diabetes is associated with a sedentary lifestyle and being overweight; therefore, patients with type 2 diabetes may not only be exposed to diets which are high in fat and low in antioxidants⁽Reference Hendrychova, Vytrisalova and Alwarafi⁷⁷⁾, and to the increased oxidative stress imposed by their diabetes⁽Reference Lima, Rosen and Maia^16, Reference Davies and Morland^26, Reference Scanlon, Connell and Ratzlaff²⁸⁾, but may also be subject to greater chronic oxidative stress as a result of their obesity itself⁽Reference Manna and Jain⁷⁸⁾. The co-occurrence of these phenomena consequently demands an even greater supply of exogenous antioxidants (vitamins C and E, carotenoids lutein, zeaxanthin and meso-zeaxanthin) to balance redox status and to combat this state of chronic oxidative stress to which the diabetic retina is exposed. While the control of blood glucose is critical in patients with diabetes, neuroprotective strategies and novel antioxidant treatments should also be considered to target the specific aetiologies and underlying pathophysiological mechanisms of the condition, and to ultimately protect against the vascular pathologies of diabetes. It is logical that natural antioxidants (lutein, zeaxanthin and meso-zeaxanthin) may have beneficial implications for diabetic retinopathy in this context. However, larger trials involving intensive assessment of oxidative stress, antioxidant defence activities and their effects on one another in the diabetic state are necessary before empirical supplementation of these carotenoids can be recommended clinically.

Inflammation and diabetes

Diabetic retinopathy has traditionally been considered a disease of the retinal microvasculature, associated with oxidative stress induced by hyperglycaemia. More recently, however, inflammation has been cited as a deleterious factor in the development of diabetic complications⁽Reference Wellen and Hotamisligil⁷⁹⁾, and is now also implicated in the pathogenesis of many ocular diseases including AMD⁽Reference Loane, Kelliher and Beatty⁸⁰⁾ and diabetic retinopathy⁽Reference Kowluru⁸¹⁾. Numerous interplays exist between inflammation and oxidative stress and vice versa⁽Reference Ishibashi, Matsui and Maeda⁸²⁾. Hyperglycaemia increases the levels of pro-inflammatory proteins⁽Reference Wellen and Hotamisligil⁷⁹⁾, and it is now believed that inflammatory processes underlie many of the functional changes in retinal vasculature observed histologically in early diabetic retinopathy, such as pericyte loss, microaneurysms, and occluded and degenerated capillaries⁽Reference Semeraro, Cancarini and Rezzola⁸³⁾.

While inflammation is normally a protective response to tissue stress, hyperglycaemia leads to a dysregulation of inflammation, which ultimately becomes chronic. Oxidative stress triggers inflammatory responses, and inflammation also enhances the production of ROS, which creates a self-perpetuating cascade towards oxidative tissue damage⁽Reference Hollyfield, Bonilha and Rayborn⁸⁴⁾. The increased expression of many inflammatory proteins in diabetic retinopathy is regulated through the activation of pro-inflammatory transcription factors including NF-κB⁽Reference Chen, Chen and Wang⁸⁵⁾. Activation of NF-κB leads to the synthesis of pro-inflammatory molecules such as IL-1, IL-6, TNF-α, c-Jun N-terminal kinase-1 (JNK-1), Iκ-B kinase-β (IKK-β) and growth factors such as VEGF⁽Reference Tang, Zhang and Jiang^21, Reference Chen, Chen and Wang⁸⁵⁾.

In the retina, the microvascular endothelium forms an effective barrier to control the movement of blood, fluid and proteins across the vessel wall. This barrier is formed by tight junctions, which consist of over forty different proteins and various inflammatory mediators including VEGF, TNF-α, protein kinase-C, IL-1β and IL-6⁽Reference Semeraro, Cancarini and Rezzola^83, Reference Bhagat, Grigorian and Tutela⁸⁶⁾. Dysfunction of these proteins is highly correlated with the pathogenesis of blood retinal barrier breakdown and increased vascular permeability, which leads to diabetic macular oedema, a major cause of vision loss in diabetic retinopathy⁽Reference Zhang, Zeng and Bao⁸⁷⁾. Inflammatory processes can induce oxidative stress, and similarly, oxidative stress can induce inflammation through the activation of multiple pathways. The coincident over-activation of inflammation and oxidative stress can therefore deplete cellular antioxidant capacity and reduce MP levels in the eye⁽Reference Beatty, Koh and Phil^9, Reference Lima, Rosen and Farah⁸⁸⁾.

The expression of various pro-inflammatory cytokines, such as TNF-α, IL-1 and IL-6, is also increased from visceral adipose tissue (more common in type 2 diabetes), and this has been linked to systemic inflammation and accompanying insulin resistance⁽Reference McArdle, Finucane and Connaughton⁸⁹⁾. The inhibition of signalling downstream of the insulin receptor is a primary mechanism through which inflammatory signalling leads to insulin resistance⁽Reference Wellen and Hotamisligil⁷⁹⁾. Several cross-sectional studies have shown that insulin resistance and type 2 diabetes are associated with higher levels of C-reactive protein (CRP), IL-6 and TNF-α, all markers of subclinical inflammation⁽Reference Tangvarasittichai, Pongthaisong and Tangvarasittichai^90–Reference Rodrigues, Pietrani and Bosco⁹²⁾. Various longitudinal studies have also shown that elevated levels of CRP and IL-6 are linked to the later development of type 2 diabetes⁽Reference Barzilay, Abraham and Heckbert^93, Reference Freeman, Norrie and Caslake⁹⁴⁾.

Inhibition of different inflammatory mediators, including NF-κB, IL-1β, VEGF and cyclo-oxygenase-2 (COX-2), has been shown to limit the degeneration of retinal capillaries characteristic of early-stage diabetic retinopathy⁽Reference Tang and Kern⁹⁵⁾. There is evidence to suggest that lutein and zeaxanthin may prevent the development of diabetic retinopathy by suppressing ROS induced by inflammation⁽Reference Izumi-Nagai, Nagai and Ohgami¹⁴⁾. Li et al.⁽Reference Li, Fung and Fu¹¹⁾ investigated the anti-inflammatory effects of lutein on Muller cells in a murine model. Treatment with lutein led to lower production of the pro-inflammatory factors NF-κB, IL-1β and COX-2, suggesting an anti-inflammatory role for lutein in protection against retinal ischaemic/hypoxic injury⁽Reference Li, Fung and Fu¹¹⁾. In another study, the protective effects of lutein and zeaxanthin against photo-oxidative damage to RPE cells were investigated. Supplementation with carotenoids modulated the inflammatory responses in these cultured RPE cells in response to photo-oxidation⁽Reference Bian, Gao and Zhou⁹⁶⁾.

The effects of zeaxanthin on diabetes-induced retinal oxidative damage and inflammation have also been investigated in diabetic rats⁽Reference Kowluru, Menon and Gierhart¹⁹⁾. It was found that zeaxanthin significantly inhibits diabetes-induced retinal oxidative damage and elevations in VEGF and its adhesion molecule, abnormalities which are commonly associated with the pathogenesis of diabetic retinopathy⁽Reference Kowluru, Menon and Gierhart¹⁹⁾. Clinical findings have highlighted the inverse association between lutein and IL-6 in coronary artery disease patients⁽Reference Chung, Leanderson and Lundberg⁹⁷⁾. A recent novel RCT, the Diabetes Visual Function Supplement Study (DiVFuSS)⁽Reference Chous, Richer and Gerson³¹⁾, explored the effects of a nutritional supplement (which included lutein and zeaxanthin) on patients with type 1 and type 2 diabetes, and suggested that the preparation used mitigated the damaging effects of systemic inflammation on ocular function, and that these effects may have been mediated by enhancements in MPOD. Chronic low-grade inflammation and oxidative stress co-exist in diabetes, particularly type 2 diabetes but also poorly controlled type 1 diabetes. Therefore, findings from DiVFuSS, which achieved significant increases in MPOD (27 % mean increase in DiVFuSS group v. 2 % mean decrease in placebo group), as well as reductions in serum high-sensitivity CRP (60 % mean decrease in DiVFuSS group v. 11 % mean decrease in placebo group), may be due in part to the inclusion of compounds which, in combination, specifically target both inflammation and oxidative stress, and therefore, which block some of the oxidative and/or inflammatory pathways responsible for the progression of diabetic retinopathy⁽Reference Chous, Richer and Gerson³¹⁾. The inclusion of many compounds in this test formula and the fact that both types of diabetic patients (type 1 and 2) were examined renders interpretation of the outcome more difficult; however, these findings are promising as they clearly show that MPOD can be augmented in patients with type 2 diabetes⁽Reference Chous, Richer and Gerson³¹⁾.

While it is clear that oxidative stress and inflammation are implicated in the initiation and development of diabetic retinopathy, it remains unclear whether the pathological effects of this oxidative stress/inflammation are mediated through MPOD depletion in type 2 diabetes. The current body of evidence which directly links lower MP levels with other metabolic correlates of diabetes, namely adiposity and dyslipidaemia, will now be explored to elucidate the possible mechanisms by which these metabolic factors mediate their effects on MPOD. It is important to note that both increased adiposity and dyslipidaemia (primarily reduced HDL) may adversely affect MP by compromising the availability⁽Reference Johnson, Hammond and Yeum^98–Reference Nolan, O’Donovan and Kavanagh¹⁰¹⁾, and transport⁽Reference Connor, Duell and Kean^102, Reference Wang, Connor and Johnson¹⁰³⁾, of dietary carotenoids to the retina. It follows that a condition such as type 2 diabetes, which unfavourably affects these parameters, may have deleterious effects on MPOD, with clinically relevant outcome effects. An analysis of the literature is presented describing these putative mechanisms: (a) evidence in relation to MPOD and body fat/adiposity (n 13; Table 3); and (b) evidence concerning MPOD and dyslipidaemia (n 7; Table 4).

Type 2 diabetes, macular pigment optical density and adiposity: the evidence

Diabetes mellitus is a chronic disorder which can alter carbohydrate, protein and fat metabolism. Weight gain, following excessive energy intake, results in the body becoming markedly resistant to the action of insulin. Despite initial impairment in insulin action, glucose tolerance can remain normal for some time because there is a compensatory increase in insulin secretion by pancreatic β-cells, which results in a well-compensated metabolic state, a condition known as hyperinsulinaemia. There is now mounting evidence to indicate that persistently elevated plasma insulin levels can contribute to the development of hypertension, plasma lipid abnormalities and atherosclerosis⁽Reference Boden¹⁰⁴⁾. The escalating rates of insulin secretion as a result of advancing obesity, however, cannot be maintained and type 2 diabetes ensues.

Intra-abdominal fat appears to be an important determinant of insulin sensitivity compared with subcutaneous fat⁽Reference Nieves, Cnop and Retzlaff¹⁰⁵⁾. Visceral fat and subcutaneous fat differ in their phenotypic, physiological and functional aspects, as there are a greater number of macrophages, T lymphocytes and pro-inflammatory molecules in visceral v. subcutaneous fat in obese individuals. Abdominal adipose tissue produces a variety of adipocytokines including IL-1, IL-6, IL-8, TNF-α, leptin and resistin; therefore, the abdominal adipose mass acts as an important mediator of inflammation in diabetes⁽Reference Castro, Macedo-de la Concha and Pantoja-Meléndez¹⁰⁶⁾. Obesity also leads to an increased size of adipocytes. The death of adipocytes, which is rare in healthy individuals, is very common in obese individuals and has been linked to adipocyte hypoxia. The accumulation of macrophages in the adipose tissue of obese individuals not only eliminates the dead cells but also further increases their synthesis of inflammatory mediators IL-8, IL-6, IL-1 and TNF-α⁽Reference Castro, Macedo-de la Concha and Pantoja-Meléndez¹⁰⁶⁾. The resultant inflammation becomes chronic; therefore, obesity is considered a low-grade inflammatory condition which in turn leads to an increase in oxidative stress. Serum 8-hydroxy guanosine (8-OhdG), a known sensitive marker of oxidative DNA damage and of total systemic oxidative stress in vivo, has been shown to be positively correlated with BMI in individuals with type 2 diabetes mellitus⁽Reference Al-Aubaidy and Jelinek¹⁰⁷⁾. Over-expression of oxidative stress, and the concomitant reduction in antioxidant defence, damages cellular structures⁽Reference Savini, Catani and Evangelista¹⁰⁸⁾, often resulting in further inflammatory responses. Antioxidant defences may be lower in overweight/type 2 diabetic patients, due to their lower intake of antioxidant-rich foods (for example, fruits and vegetables), their increased utilisation of these molecules (for example, increased inflammation/ROS in insulin-dependent tissue such as the retina) and their impaired generation of other supportive antioxidants⁽Reference Savini, Catani and Evangelista¹⁰⁸⁾, factors which may collectively lead to MPOD depletion.

The diminished response to insulin in type 2 diabetes also leads to increased lipolysis in adipocytes. Enlarged adipocytes release NEFA and adipocytokines⁽Reference Matough, Budin and Hamid^67, Reference Boden¹⁰⁴⁾; therefore greater amounts of NEFA reach the liver via the portal vein, resulting in fatty liver infiltration. Mobilised NEFA are oxidised by vascular endothelium to generate ROS. Elevations of NEFA in serum also inhibit insulin suppression of hepatic glucose production, leading to an increase in glucose production by the liver⁽Reference Reaven, Hollenbeck and Jeng¹⁰⁹⁾. This, in turn, results in both insulin resistance and inflammation in the major insulin target tissues (skeletal muscle, liver and endothelial cells). Abnormal levels of lipids, fatty acids and various adipocytokines from adipose tissue therefore initiate a vicious cycle of fat damage, inflammation, worsening insulin resistance, impaired β-cell insulin secretion and ultimately type 2 diabetes⁽Reference Boden^104, Reference Karpe, Dickmann and Frayn¹¹⁰⁾. The chronic low-grade inflammation associated with obesity and type 2 diabetes leads to increased oxidative stress and further inflammation, putting a greater demand on antioxidant defences.

Apart from its pro-inflammatory and pro-oxidant effects, adipose tissue is also a major body store for carotenoids such as lutein and zeaxanthin. Higher levels of body fat have been shown to be related to lower levels of circulating carotenoids⁽Reference Burke, Curran-Celentano and Wenzel¹¹¹⁾, making these pigments less available to retinal tissue⁽Reference Gruber, Chappell and Millen¹¹²⁾. Furthermore, a randomised controlled weight loss trial found that a significant weight loss in the intervention group was related to increases in serum lutein⁽Reference Kirby, Beatty and Stack¹¹³⁾. Higher BMI⁽Reference Seddon, Cote and Davis¹¹⁴⁾ and higher body fat percentage⁽Reference Nolan, O’Donovan and Kavanagh¹⁰¹⁾ have both been linked with an increased risk of AMD. Studies have reported an inverse relationship between adiposity and BMI and MPOD⁽Reference Hammond, Ciulla and Snodderly^99, Reference Nolan, Feeney and Kenny¹¹⁵⁾, with the observed relationships largely attributable to participants with a BMI > 29 kg/m²⁽Reference Hammond, Ciulla and Snodderly⁹⁹⁾. One study found that higher body fat percentage, even within relatively healthy limits, was associated with lower retinal tissue lutein and zeaxanthin status⁽Reference Bovier, Lewis and Hammond¹¹⁶⁾, suggesting that adiposity may affect the nutrient status of the retina. Sex differences in MPOD have also been reported⁽Reference Johnson, Hammond and Yeum^98, Reference Nolan, O’Donovan and Kavanagh¹⁰¹⁾, which may imply differences in metabolism of MP, or reflect differences in adipose tissue storage of these dietary carotenoids between men and women. Competitive absorption by a larger overall fat mass (percentage fat mass is generally higher in women)⁽Reference Blaak¹¹⁷⁾, or specifically by greater abdominal fat mass (which is more common in men)⁽Reference Björntorp¹¹⁸⁾, may reduce lutein and zeaxanthin uptake by the retina. Concentrations of carotenoids differ according to body site, with levels demonstrably higher in abdominal fat⁽Reference Chung, Ferreira and Epstein¹¹⁹⁾, which may explain the differences in MPOD currently observed between men and women⁽Reference Johnson, Hammond and Yeum^98, Reference Nolan, O’Donovan and Kavanagh¹⁰¹⁾. Since adipose tissue may trap lutein and zeaxanthin⁽Reference Bovier, Lewis and Hammond¹¹⁶⁾, individuals who are overweight or obese (type 2 diabetes) may consequently have lower concentrations of the carotenoids lutein and zeaxanthin available at the macula.

Overall, the research appears to suggest that obesity has a negative impact on MPOD; however, there is also some evidence to the contrary. One study found that mean MPOD values did not differ significantly between subjects with various types of obesity and ideal-weight subjects⁽Reference Gupta, Raman and Biswas¹²⁰⁾. Our previous study⁽Reference Scanlon, Connell and Ratzlaff²⁸⁾ is the only one to date which has examined the relationship between BMI and MPOD in a group with diabetes. This study found that MPOD was significantly lower and BMI was significantly higher in patients with type 2 diabetes, compared with type 1 and non-diabetic controls, highlighting the possibility that different relationships exist between MPOD status and type 1 and 2 diabetes. The higher BMI levels observed in type 2 diabetes are reflective of excess adiposity which acts as a store for lutein and zeaxanthin. However, as described previously, obesity is also independently associated with increased oxidative stress⁽Reference Savini, Catani and Evangelista^108, Reference Fernández-Sánchez, Madrigal-Santillan and Bautista^121, Reference Bondia-Pons, Ryan and Martinez¹²²⁾, and with inflammation⁽Reference Kwon and Pessin¹²³⁾ which exacerbates oxidative stress, and may therefore have multiple depletive effects on MP. Unfortunately, the cross-sectional nature of this study⁽Reference Scanlon, Connell and Ratzlaff²⁸⁾ means that it cannot elucidate the extent or nature of any causal relationships between BMI and MPOD in these type 2 subjects.

The different presenting features of type 1 and 2 diabetes, and the interplay of their disease components with MPOD status (for example, adiposity), may mean that the mechanistic relationship between these conditions and MPOD status is not uniform across both. The fact that much of the research to date includes both diabetic groups may obfuscate any relationships present. Although the evidence to date is somewhat conflicting, there appears to be a theoretical basis for a relationship between the characteristic body fatness of type 2 diabetes and lower levels of MPOD. A summary of the studies investigating the association between MPOD and adiposity are outlined in Table 3⁽Reference Scanlon, Connell and Ratzlaff^28, Reference Mares, LaRowe and Snodderly^29, Reference Johnson, Hammond and Yeum^98, Reference Hammond, Ciulla and Snodderly^99, Reference Nolan, O’Donovan and Kavanagh^101, Reference Kirby, Beatty and Stack^113, Reference Nolan, Feeney and Kenny^115, Reference Bovier, Lewis and Hammond^116, Reference Gupta, Raman and Biswas^120, Reference Broekmans, Berendschot and Klopping-Ketelaars^124–Reference Edwards, Walk and Cannavale¹²⁷⁾.

Type 2 diabetes, macular pigment optical density and dyslipidaemia: the evidence

Insulin resistance and obesity, metabolic abnormalities commonly associated with type 2 diabetes, may themselves lead to disturbances in the production and clearance of plasma lipoproteins⁽Reference Goldberg¹²⁸⁾. Dyslipidaemia is often characterised by low levels of HDL and hypertriacylglycerolaemia. In addition, LDL particles are converted to smaller, more atherogenic lipoproteins⁽Reference Krauss and Siri¹²⁹⁾. A number of factors are likely to be responsible for diabetic dyslipidaemia and these include insulin’s effects on liver apoprotein production, dysregulation of lipoprotein lipase (LPL), impaired action of cholesteryl ester transfer protein and the peripheral action of insulin on adipose and muscle⁽Reference Goldberg¹²⁸⁾. Hormone-sensitive lipase and LPL are two enzymes that regulate mobilisation and deposition of fatty acids in adipose tissue in a reciprocal manner⁽Reference Frayn, Coppack and Fielding¹³⁰⁾. Reduced insulin action leads to increased lipolysis in adipocytes with increased NEFA release, which can mediate many adverse metabolic effects, most notably further insulin resistance⁽Reference Karpe, Dickmann and Frayn¹¹⁰⁾ and dyslipidaemia⁽Reference Goldberg¹²⁸⁾. In diabetes greater amounts of fatty acids returning to the liver are reassembled into TAG and secreted in VLDL⁽Reference Ginsberg and Illingworth¹³¹⁾. In the presence of increased concentrations of VLDL in circulation, cholesteryl ester transfer protein will exchange VLDL TAG for cholesteryl ester in the core of HDL, which results in smaller HDL particles, which are more rapidly cleared from plasma⁽Reference Horowitz, Goldberg and Merab¹³²⁾. This explains the hypertriacylglycerolaemia and reduced HDL commonly observed in type 2 diabetes⁽Reference Goldberg¹²⁸⁾. Cholesteryl ester transfer protein also exchanges VLDL TAG for cholesteryl ester in the core of LDL and the net effect is a decrease in the size and an increase in the density of LDL particles⁽Reference Goldberg¹²⁸⁾. In vitro, small dense LDL can be oxidised more easily, and it binds to LDL receptors less avidly than normal LDL. As a result, diabetic patients are at an increased risk of CVD. Many clinicians measure LDL density and/or size to predict risk of CVD. An increase in waist circumference and waist:hip ratio has been found to be associated with small dense LDL particles⁽Reference Goldberg^128, Reference Campos, Willett and Peterson¹³³⁾. Therefore, intra-abdominal fat is a critical determinant of an atherogenic lipoprotein profile, including disruptions in the distribution of cholesterol in the different lipoprotein fractions⁽Reference Nieves, Cnop and Retzlaff¹⁰⁵⁾.

The characteristic lipid profile in an individual with type 2 diabetes (increased serum VLDL, increased TAG, decreased serum HDL and less commonly, an increase in LDL-cholesterol)⁽Reference Goldberg¹²⁸⁾ may have important implications for MP levels in the eye, as studies have shown that the carotenoids lutein and zeaxanthin are primarily transported by HDL in plasma⁽Reference Connor, Duell and Kean^102, Reference Wang, Connor and Johnson¹⁰³⁾. HDL is a known ligand for scavenger receptor B type 1 (SR-B1), and this suggests a ‘piggy-back’ mechanism of lutein and zeaxanthin uptake into the retina via RPE cells which is mediated by SR-B1⁽Reference During, Doraiswamy and Harrison¹³⁴⁾. This hypothesis is supported by evidence from lutein deficiencies observed in the retina of chickens with a genetic HDL defect but not in other organs, implying a major role for HDL in the receptor-mediated transport of lutein to the eye⁽Reference Connor, Duell and Kean¹⁰²⁾. It has been argued that the subspecies of HDL containing apoE supplies lipids and lipid-soluble lutein and zeaxanthin to the retina, and that by increasing HDL, retinal lutein levels may be simultaneously augmented⁽Reference Thomson, Toyoda and Langner¹⁰⁰⁾.

n-3 Long-chain PUFA supplementation may also lead to increases in serum HDL⁽Reference Thomas, Smith and Donahue¹³⁵⁾. Recent epidemiological studies indicate that increased dietary long-chain-PUFA intakes enhance MPOD⁽Reference Delyfer, Buaud and Korobelnik¹³⁶⁾, and that DHA facilitates the accumulation of lutein in the blood and macula. The mechanism by which DHA increases MPOD may relate to its effect on the transport and uptake of lutein into the macula by mediating changes in lipoprotein levels⁽Reference Johnson, Chung and Caldarella¹³⁷⁾. Putatively, this effect could also be mediated by the anti-inflammatory effects of DHA, but this is conjectural and no direct evidence supports this effect within the retina.

One observational study found that the correlation between MPOD, serum lutein, zeaxanthin and lipids differed depending on the age profile of participants⁽Reference Olmedilla-Alonso, Beltrán-de-Miguel and Estévez-Santiago¹³⁸⁾. In younger adults, MPOD was found to be influenced by serum lutein, whereas in older adults, circulating lipids were an added determining factor⁽Reference Olmedilla-Alonso, Beltrán-de-Miguel and Estévez-Santiago¹³⁸⁾. Mares et al.⁽Reference Mares, LaRowe and Snodderly²⁹⁾ queried whether any association with serum lipids is truly strong enough to play a meaningful role in the pathogenesis of MPOD depletion. One study which used a diabetic cohort, involving a small group of forty-three participants (fourteen controls, seventeen with diabetes but without retinopathy and twelve with diabetes and with mild non-proliferative retinopathy), found no significant correlation between MPOD and serum lipid levels⁽Reference Lima, Rosen and Maia¹⁶⁾. Our own research group⁽Reference Scanlon, Connell and Ratzlaff²⁸⁾, however, reported lower levels of MPOD in type 2 v. type 1 diabetics and noted significantly lower HDL values within the type 2 group, raising the possibility that lower HDL levels may have contributed to this difference. More recently, the DiVFuSS study⁽Reference Chous, Richer and Gerson³¹⁾ has demonstrated significant improvements in serum LDL-cholesterol level (9 % mean decrease in the intervention group v. 1 % mean increase in placebo group), HDL-cholesterol (7 % mean increase in the intervention group v. 3 % decrease in placebo group) and TAG (8·6 % mean decrease in the intervention group v. 2 % mean increase in placebo group), suggesting a positive effect of the study formulation on serum lipids. Whilst this effect was probably attributable to other non-carotenoid components of the supplement, the synchronous increase in MPOD observed in the intervention group may have been partially mediated by these favourable changes in lipid profile and their consequential impact on carotenoid transport to the retina⁽Reference Connor, Duell and Kean^102, Reference Wang, Connor and Johnson^103, Reference Loane, Nolan and Beatty¹³⁹⁾. In the broader context, the evidence from this study is sufficiently robust to prompt further lines of investigation into the role of MP supplementation either in isolation or as a multi-component supplement in diabetes. The research to date, however, is not conclusive and it remains unclear from the limited amount of observational and RCT data whether serum lipids (primarily HDL) play a meaningful role in determining MPOD status. Given the theoretical basis for an association between lipid profile and MP status, and the fact that lipid profile is often altered deleteriously in diabetes, further investigation is warranted. The relationship between dyslipidaemia and lower levels of MP is outlined in Table 4⁽Reference Lima, Rosen and Maia^16, Reference Mares, LaRowe and Snodderly^29, Reference Chous, Richer and Gerson^31, Reference Olmedilla-Alonso, Beltrán-de-Miguel and Estévez-Santiago^138–Reference Nagai, Izumi-Nagai and Suzuki¹⁴¹⁾.

Causal mechanisms and metabolic perturbations associated with lower macular pigment in diabetes

The candidate causal mechanisms explored herein as contributors to lower MPOD levels in diabetes relate primarily to type 2 diabetes (or poorly controlled type 1). The complex inter-relationship between oxidative stress/inflammation and the metabolic correlates of diabetes (adiposity, dyslipidaemia, insulin resistance, hyperinsulinaemia and hyperglycaemia), and their associations with MPOD, both individually and as a whole, represent a significant scientific challenge. In summary, our analysis indicates that MPOD might be depleted through at least four causal mechanisms previously described, all of which are outlined in Fig. 2.

Fig. 2. Metabolic perturbations and supply deficiencies contributing to macular pigment depletion in diabetes. MPOD, macular pigment optical density; ROS, reactive oxygen species. For a colour figure, see the online version of the paper.

Dietary determinants of macular pigment optical density in diabetes

First and foremost, patients with type 2 diabetes may experience lower levels of MP as a result of poor dietary intake of antioxidant nutrients (lutein, zeaxanthin and meso-zeaxanthin). Lutein and zeaxanthin are not synthesised de novo in humans; therefore, a diet rich in fruit and vegetables is necessary to enhance MP levels in the eye⁽Reference Sommerburg, Keunen and Bird⁴⁾. While the exact cause of diabetes is still not fully understood, dietary factors have garnered particular attention over the last number of years. In fact, some research indicates that up to 74 % of type 2 diabetes cases are directly attributable to obesity⁽Reference Janssen¹⁴²⁾.

Many studies have reported a positive association between a high intake of sugars and the development of type 2 diabetes⁽Reference Gross, Li and Ford^143–Reference Elliott, Keim and Stern¹⁴⁵⁾. Research has also found an association between high fat intakes and type 2 diabetes and impaired glucose tolerance⁽Reference Marshall, Hamman and Baxter^146, Reference Saldana, Siega-Riz and Adair¹⁴⁷⁾, with some literature suggesting that high-fat, low-carbohydrate diets are associated with the onset of type 2 diabetes⁽Reference Marshall, Hamman and Baxter¹⁴⁶⁾. A global change in dietary intake has occurred over recent decades resulting from an increased use of sweeteners such as fructose and sucrose by the food industry⁽Reference Nseir, Nassar and Assy¹⁴⁸⁾. Recent evidence suggests a causal link between the high intake of sugar-sweetened soft drinks and obesity and diabetes. This is thought to arise from the large amount of fructose in these soft drinks, which impairs insulin sensitivity and raises blood glucose levels and BMI to dangerous levels⁽Reference Bray, Nielsen and Popkin^144, Reference Raben, Vasilaras and Møller¹⁴⁹⁾. Fructose consumption increases postprandial TAG concentrations within hours⁽Reference Swarbrick, Stanhope and Elliott¹⁵⁰⁾, which suggests that postprandial hypertriacylglycerolaemia is one of the earliest metabolic perturbations associated with fructose consumption. The most likely mechanism for this hypertriacylglycerolaemia is increased hepatic de novo lipogenesis, which in turn up-regulates VLDL production and secretion⁽Reference Stanhope and Havel¹⁵¹⁾, factors which have a knock-on suppressive effect on HDL, a known transporter of carotenoids to the eye⁽Reference Connor, Duell and Kean^102, Reference Wang, Connor and Johnson¹⁰³⁾.

Studies have also found an inverse correlation between the intake of green leafy vegetables and risk of developing type 2 diabetes⁽Reference Wang, Fang and Gao^152, Reference Li, Fan and Zhang¹⁵³⁾, and that the consumption of fruits and vegetables may protect against the development of type 2 diabetes, as they are rich in micronutrients, fibre and antioxidants⁽Reference Li, Fan and Zhang¹⁵³⁾, dietary components also known to enhance MPOD. A recent meta-analysis revealed that a higher intake of fruit, especially berries and green leafy vegetables, yellow vegetables and cruciferous vegetables or their fibre, is associated with a lower risk of type 2 diabetes⁽Reference Wang, Fang and Gao¹⁵²⁾. Health benefits have also been observed with the Mediterranean diet⁽Reference Tsoupras, Lordan and Zabetakis^154, Reference Romagnolo and Selmin¹⁵⁵⁾, a diet which contains an abundance of fruit and vegetables. While the beneficial effects of fruits and vegetables on diabetes risk are likely to be multiple, these foods are the primary dietary source of carotenoids including lutein and zeaxanthin, highlighting their consumption as a key determinant of MPOD. Lower levels of MPOD in diabetes might also be associated with high intakes of fried foods and fat⁽Reference Guallar-Castillón, Rodriguez-Artalejo and Fornes¹⁵⁶⁾, alcohol⁽Reference Traversy and Chaput¹⁵⁷⁾, red and processed meat⁽Reference Wang and Beydoun¹⁵⁸⁾, sugar-sweetened beverages⁽Reference Gross, Li and Ford^143, Reference Bray, Nielsen and Popkin¹⁴⁴⁾ and high-glycaemic index foods⁽Reference Ludwig, Majzoub and Al-Zahrani¹⁵⁹⁾, all of which have been linked with higher levels of visceral fat and increased waist circumference. Numerous studies have shown that obesity (in particular, central obesity) can increase the risk of type 2 diabetes⁽Reference Mokdad, Ford and Bowman¹⁶⁰⁾.

While the body’s natural defence against oxidative damage and inflammation is neutralisation by endogenous antioxidants (enzymic and non-enzymic), in association with exogenous antioxidants (vitamin C, vitamin E, carotenoids, β-carotene, lutein, zeaxanthin and meso-zeaxanthin)⁽Reference Bouayed and Bohn^50, Reference Cabrera and Chihuailaf⁶⁸⁾, these antioxidant defences may be compromised in diabetes. Hence, it is clear that both over-utilisation and under-supply of antioxidant nutrients may contribute to lower levels of MP in the diabetic retina.

Based on the outlined associations between nutritional and dietary factors and the components of the metabolic syndrome which may contribute to MPOD depletion, we outline dietary factors which: (1) may exacerbate the metabolic correlates associated with type 2 diabetes (Table 5 column a); and (2) may help to prevent or resolve the metabolic perturbations associated with type 2 diabetes (Table 5 column b).

Table 5. Dietary factors associated with the metabolic correlates of type 2 diabetes mellitus

FA, fatty acids.

Overall, Western dietary habits are a significant factor in the development of the metabolic syndrome. The continuous over-provision of energy via dietary carbohydrate and lipid, when unmatched by physical activity-induced energy expenditure, leads to a state of excess adiposity, chronic low-grade inflammation, and oxidative stress and dyslipidaemia; features associated with type 2 diabetes and which may lead to MPOD depletion.

The importance of diet cannot be overemphasised in diabetes management. Given the variety of mechanisms which may contribute to MPOD depletion and the broader ocular risks associated with such MPOD depletion in diabetes, a comprehensive dietary approach is therefore necessary to optimise ocular health and MPOD status. The generous provision of brightly coloured fruits and vegetables which are rich in a broad range of antioxidants including the target carotenoids lutein, zeaxanthin and meso-zeaxanthin represents an obvious recommendation. This, however, should be complemented by the inclusion of wholegrain-fortified breakfast cereals for their folate, vitamin D and anti-obesogenic fibre content; and with plenty of oily fish to provide further vitamin D and n-3 essential fatty acids to attenuate inflammation and correct or prevent hypertriacylglycerolaemia. These foods should be used to displace those which are high in fat, saturated fat, refined sugar and fructose; for example, oily fish should replace red or processed meats at several main meals each week, fruit or raw vegetables should be used to replace snack foods high in fat, saturated fat, trans-fat and refined sugar (for example, baked goods and confectionery), and low-fat fortified milk should replace sugar-sweetened soft drinks. Starchy carbohydrates should be evenly distributed over the day to achieve better glycaemic control, with sterol- or stanol-enriched products used to reduce serum LDL and consequently enhance LDL:HDL ratio. Finally, from a broader lifestyle perspective, these individuals should be advised to moderate their alcohol consumption to no more than two standard drinks per d with several alcohol-free days each week, and to engage in moderate-intensity exercise (for example, walking) on most days of the week to enhance their insulin sensitivity and optimise their blood glucose and lipoprotein profiles. Table 5 (column a) outlines the dietary factors causally associated with the metabolic correlates of type 2 diabetes, and Table 5 (column b) outlines the dietary factors associated with protection against the metabolic perturbations of type 2 diabetes.