For nearly 200 years, human life expectancy has been increasing at approximately 2 years per decade and shows little indication of reaching a plateau( Reference Kirkwood 1 ). Much of the early benefit was derived from reduced childhood mortality from infectious diseases, but, since the mid-twentieth century, the gains have been due mainly to reduced mortality in later life. The latter can be attributed not only to better health care but also to the effects of economic growth and to public health policies including those that have restricted tobacco use and improved road traffic safety( Reference Mackenbach, Karanikolos and McKee 2 ). Despite these improvements, globally, there are large inequalities with countries in sub-Saharan Africa, notably Swaziland, Chad, Guinea-Bissau and South Africa, having life expectancy at birth which is little over half of that in the most favoured countries including Monaco, Macau, Japan and Singapore( 3 ). Within Europe, improvements in life expectancy in Western Europe over the past four decades have not been achieved in Eastern Europe because of political and macro-economic issues and of a failure to implement effective health policies( Reference Mackenbach, Karanikolos and McKee 2 ). Similar socio-economic gradients in life expectancy have also been observed within countries, and Marmot( Reference Marmot 4 ) has estimated that 2·5 million years of life are potentially lost annually to premature mortality in England through the effects of inequalities in health.
Ageing is the major risk factor for most common chronic diseases( Reference Rae, Butler and Campisi 5 ) including cancers( Reference DePinho 6 ), CVD and stroke( Reference Feigin, Lawes and Bennett 7 ) and dementia( Reference Hebert, Weuve and Scherr 8 ). As a consequence, the gain in years of life over the past few decades has been accompanied by additional years of chronic poor health so that the greatest proportion of total expenditure on health care is now concentrated in the last few years of life. Interventions that reduce morbidity in later life are likely to have significant financial and social benefits, in addition to gains in individual well-being.
Biology of ageing
Ageing is characterised by progressive, time-dependent loss of function and increased likelihood of death. This loss of function includes widely recognised, relatively rapid processes such as the loss of female fertility following the menopause and much more insidious declines in brain volume and in skeletal muscle mass that can lead to cognitive and physical frailty, respectively. There is no evidence that ageing per se is genetically determined (or programmed)( Reference Kirkwood 9 ), and the ageing phenotype appears to result from the gradual accumulation of damage to all the cell's macromolecules, such as DNA, proteins and lipids( Reference Kirkwood 1 ). Damage to DNA is not confined to the nuclear genome and there is good evidence of increasing burden of mitochondrial DNA mutations with increasing age in humans (Fig. 1)( Reference Greaves, Barron and Plusa 10 , Reference Greaves, Elson and Nooteboom 11 ). Damaged (unfolded) proteins also accumulate with age( Reference Lopez-Otin, Blasco and Partridge 12 ), and it seems likely that long-lived proteins, i.e. those that turnover slowly, if at all, may be at the greatest risk of acquiring damage with age( Reference Toyama and Hetzer 13 ). For example, the accumulation of damaged crystallin in the human lens with age leads to cataract( Reference Toyama and Hetzer 13 ), which is the most common cause of blindness. The structure and function of cells and of organelles depend on the composition and integrity of membrane phospholipids. These phospholipids are rich in PUFA that are susceptible to oxidative damage, and peroxidation of membrane phospholipid acyl chains is hypothesised to play a causal role in the ageing process( Reference Pamplona 14 ). As one would expect, age-related macromolecular damage has profound effects on cell function, and such dysfunction is likely to be most pervasive when it affects stem cells that are essential for the normal maintenance of mitotic tissues and for tissue repair following disease, surgery or other trauma. During ageing, stem cells may (1) become depleted through loss of capacity for self-renewal, (2) lose their ability to generate appropriately differentiated cell lineages, (3) become moribund because of senescence or (4) acquire genetic and epigenetic damage that initiates tumorigenesis( Reference Liu and Rando 15 ).
Fig. 1 Accumulation with age of mutations in mitochondrial DNA shown as % crypts with respiratory chain defects in biopsies of colorectal mucosa from healthy humans. From Greaves et al. (2010)( Reference Toyama and Hetzer 13 ). R 2= 0·951.
In a recent review, Lopez-Otin et al. ( Reference Lopez-Otin, Blasco and Partridge 12 ) proposed nine tentative hallmarks of ageing that are common to many organisms, including mammals. These hallmarks include the following factors: genomic instability; telomere attrition; epigenetic alterations; loss of proteostasis; deregulated nutrient sensing; mitochondrial dysfunction; cellular senescence; stem cell exhaustion; altered intercellular communication (Fig. 2)( Reference Lopez-Otin, Blasco and Partridge 12 ). In selecting these hallmarks, Lopez-Otin et al. ( Reference Lopez-Otin, Blasco and Partridge 12 ) searched for factors that met the following criteria: (1) it occurs during ageing; (2) experimental enhancement of the factor accelerates ageing; (3) experimental amelioration of the factor slows ageing and so increases healthy lifespan. Lopez-Otin et al. ( Reference Lopez-Otin, Blasco and Partridge 12 ) recognised that evidence for the third criterion was often limited.
Fig. 2 Hallmarks of ageing proposed by Lopez-Otin et al. (2013)( Reference Lopez-Otin, Blasco and Partridge 12 ).
Although there is no ‘death gene’, population genetics studies have shown that about 25 % of the inter-individual variation in lifespan can be accounted for by heredity, and it seems likely that the underlying genetic factors influence ageing through effects on somatic maintenance( Reference Kirkwood 1 ). The remaining 75 % of the variation is due to environmental factors and stochastic events. In other words, the ageing process is plastic and interventions that increase longevity, and which reduce susceptibility to age-related diseases, are those that reduce macromolecular damage and/or which enhance the cell's ability to repair such damage. The ubiquitous causes of molecular damage include inflammation, metabolic stress and oxidative stress/redox changes. Some inter-individual differences in capacity to cope with macromolecular damage are genetically encoded. For example, the 50-fold difference in lifespan between the Atlantic bay scallop (Argopecten irradians) and the sessile giant clam (Tridacna derasa) appears to be due, at least in part, to more effective somatic maintenance in the giant clam( Reference Ungvari, Csiszar and Sosnowska 16 ). In humans, diseases characterised by accelerated ageing such as Werner's syndrome, Cockayne's syndrome and xeroderma pigmentosum are caused by inherited defects in the machinery responsible for DNA repair.
Nutritional modulation of the ageing process
Since the ageing phenotype results from the accumulation of damage to the cell's macromolecules, nutritional interventions that reduce ageing must do so because they reduce the amount of damage sustained by the cell and/or because they enhance the capacity of the cell, tissue or organism to repair, or to cope with, that damage. Genetics-based studies( Reference Kenyon 17 , Reference Bishop, Lu and Yankner 18 ), and intervention studies with pharmaceutical agents such as rapamycin( Reference Harrison, Strong and Sharp 19 ), have strengthened the evidence for the plasticity of the ageing process, and it is apparent that modulation of a relatively small number of pathways central to energy and nutrient sensing and to cellular defence is key to the ageing process. Investigation of these same processes is likely to reveal how nutrition influences ageing.
For at least 80 years, it has been apparent that nutrition modulates the ageing process. The earliest unequivocal evidence for the impact of nutrition on ageing came from studies in rodents showing that dietary restriction (DR) – providing animals with less food than they would eat under ad libitum conditions – increased lifespan and reduced (or delayed) the development of age-related diseases. DR (usually energy restriction while ensuring adequate nutrient supply) is now a very well-established experimental paradigm that extends lifespan and healthspan in many species( Reference Nakagawa, Lagisz and Hector 20 , Reference Simons, Koch and Verhulst 21 ). However, note that the effects of DR on lifespan can be heterogeneous, even within the same species, and that there is evidence that DR can reduce lifespan in some inbred strains of mice( Reference Liao, Rikke and Johnson 22 ). The effect of DR in humans is not known, and both practical and ethical constraints mean that a direct test of the hypothesis is challenging; however, two recent studies have attempted to test the hypothesis in another primate – the rhesus monkey. One of these studies reported apparently greater longevity in rhesus monkeys exposed to 30 % DR throughout adulthood( Reference Colman, Anderson and Johnson 23 ). In contrast, the other study found no effect on lifespan regardless of the stage in the life course at which DR was introduced( Reference Mattison, Roth and Beasley 24 ). Several features of the studies, including differences in dietary regimen, might explain the divergent effects on lifespan, but none of these was conclusive( Reference Austad 25 ). However, in both studies, the dietary-restricted animals were leaner and had a lower burden of age-related diseases. Translating these outcomes to humans suggests that avoiding obesity may improve healthy ageing, and, indeed, there is strong evidence for greater risk of death when BMI is in the overweight and, especially, obese range( 26 ).
Obesity and ageing
The Prospective Studies Collaboration pooled data from fifty-seven human prospective studies involving approximately 900 000 adult participants to investigate the relationships between BMI at baseline and subsequent mortality( 26 ). Most studies were based in Western Europe or North America, and to minimise potential confounding due to current ill health, deaths in the first 5 years of follow-up were excluded. For most of the common causes of death, including CVD, there was strong evidence of progressively greater risk of death with increasing BMI above the minimum in the range of 22·5–25 kg/m2. For some cause-specific mortality, e.g. cancer, there was greater risk with BMI < 20 kg/m2( 26 ). In addition, there is evidence that increased adiposity in adulthood has adverse effects on long-term health. Compared with women who were lean (BMI 18·5–22·9 kg/m2) at age 18 years of age and who remained weight stable in adult life, those who gained weight in midlife had reduced likelihood of healthy survival after the age of 70 years( Reference Sun, Townsend and Okereke 27 ). In this case, ‘healthy survival’ was defined as having no history of major chronic diseases and no substantial cognitive, physical or mental limitation( Reference Sun, Townsend and Okereke 27 ). Women who were overweight at 18 years of age, regardless of subsequent weight change, had reduced healthy survival and those who were overweight at 18 years of age and gained at least 10 kg in midlife were five times less likely to enjoy healthy survival after 70 years of age( Reference Sun, Townsend and Okereke 27 ).
Being obese, or carrying the fat mass and obesity-associated protein encoding gene (FTO) risk allele for overweight and obesity, is associated with changes in brain structure, which are consistent with more rapid brain ageing( Reference Ho, Stein and Hua 28 ). Such results from imaging studies correlate positively with evidence from epidemiological studies showing that overweight and obesity in midlife are associated with a greater risk of dementia( Reference Anstey, Cherbuin and Budge 29 ). Obesity and sedentary behaviour (a potential causative factor) are associated with an increased risk of molecular and cellular damage and, if sustained, with an increased risk of all common age-related diseases due, at least in part, to the damaging effects of chronic, low-level inflammation( Reference Handschin and Spiegelman 30 ). Given that obesity-related behaviours (diet and physical activity) track from childhood into adulthood( Reference Craigie, Lake and Kelly 31 ), effective interventions that prevent obesity in childhood may have lifelong benefits.
Dietary patterns and human ageing
Nutrition is critical for health and well-being at all stages in the life course, and, indeed, the nutrition of one generation may influence ageing in the next generation. While nutrition has immediate effects on metabolism and health, nutritional exposures can have very long legacies. This is exemplified by the impact of maternal nutrition during pregnancy and of nutrition in early postnatal life on ageing (reviewed by Langie et al. ( Reference Langie, Lara and Mathers 32 )). For example, we have shown recently that a nutritional insult during pregnancy and lactation (maternal folate insufficiency), especially when followed by a second nutritional insult (high fat feeding from weaning), may reduce genomic defence mechanisms in the adult offspring brain, i.e. expression of base excision repair genes (a DNA repair system that corrects the lesions caused by oxidative damage)( Reference Langie, Achterfeldt and Gorniak 33 ).
It is probable that many dietary factors including total energy intake relative to energy needs (which determines the risk of obesity), specific nutrients and other non-nutrient bioactive constituents, individually and collectively, influence the accumulation of macromolecular damage within the cell and, therefore, the ageing process. For example, we have observed that the capacity for nucleotide excision repair (a DNA repair system that removes large DNA adducts) is reduced in those with a higher BMI, even among young, healthy adults( Reference Tyson, Caple and Spiers 34 ). More broadly, there is evidence that nutritional intake and nutritional status influences DNA repair( Reference Tyson and Mathers 35 ), and these nutritional factors may explain some of the substantial inter-individual variation in DNA repair capacity in humans( Reference Caple, Williams and Spiers 36 ). Investigation of such complex interrelationships is challenging especially in human epidemiology, with difficulties in exposure measurement( Reference Fave, Beckmann and Draper 37 , Reference Penn, Boeing and Boushey 38 ) and with substantial risks of confounding. The more recent focus on dietary patterns offers promise not only in identifying associations with healthy ageing but also in providing the evidence base for the development of public health interventions. The strongest evidence for links between a dietary pattern and ageing is that for the Mediterranean diet – a dietary pattern which is characterised by high consumption of plant-based foods, moderate-to-high consumption of fish and low intakes of dairy foods and meats and meat products( Reference Trichopoulou, Orfanos and Norat 39 ). In addition, meta-analysis has shown that adherence to the Mediterranean diet is associated with substantial reductions in the risk of several major age-related diseases including CVD, cancers and neurodegenerative diseases( Reference Sofi, Abbate and Gensini 40 ). Such observational evidence has been strengthened considerably by a recent human intervention study that has demonstrated primary prevention of CVD with Mediterranean diet-based interventions( Reference Estruch, Ros and Salas-Salvado 41 ).
Concluding comments
This is an exciting time for research on nutrition and ageing. It is now evident that the ageing phenotype is due to the accumulation of macromolecular damage and that the ageing trajectory is plastic and responds to dietary, and other, interventions. In addition, there is reason to hope that we will soon have a better understanding of the molecular mechanisms through which nutrition can enhance healthy ageing. At a public health level, we need effective interventions to improve dietary behaviours so that we can reap the rewards of greater health and well-being in later life( Reference Mathers 42 ). To date, research on such interventions has been hampered by the lack of appropriate outcome measures since the biological complexity of the ageing process means that there is no single, simple and reliable measure of how healthily someone is ageing. To help address this research gap, we have proposed, tentatively, a panel of outcome measures based on the concept of the ‘Healthy Ageing Phenotype’( Reference Franco, Karnik and Osborne 43 ), which could be deployed in community-based, lifestyle interventions on healthy ageing( Reference Lara, Godfrey and Evans 44 ).
Acknowledgements
The present study was supported by the Centre for Ageing & Vitality and the LiveWell Programme, both of which are funded through the Lifelong Health and Wellbeing cross-council initiative managed by the Medical Research Council (MRC) on behalf of the funders: Biotechnology and Biological Sciences Research Council; Engineering and Physical Sciences Research Council; Economic and Social Research Council; MRC; Chief Scientist Office of the Scottish Government Health Directorates; National Institute for Health Research (NIHR)/the Department of Health; the Health and Social Care Research & Development of the Public Health Agency (Northern Ireland); and the Wales Office of Research and Development for Health and Social Care and the Welsh Assembly Government. These funders had no role in the design and analysis of the study or in the writing of this article.
There are no conflicts of interest.
