Social modulation of ageing: mechanisms, ecology, evolution

Abstract

Human life expectancy increases, but the disease-free part of lifespan (healthspan) and the quality of life in old people may not show the same development. The situation poses considerable challenges to healthcare systems and economies, and calls for new strategies to increase healthspan and for sustainable future approaches to elder care. This call has motivated innovative research on the role of social relationships during ageing. Correlative data from clinical surveys indicate that social contact promotes healthy ageing, and it is time to reveal the causal mechanisms through experimental research. The fruit fly Drosophila melanogaster is a prolific model animal, but insects with more developed social behaviour can be equally instrumental for this research. Here, we discuss the role of social contact in ageing, and identify lines of study where diverse insect models can help uncover the mechanisms that are involved.

This article is part of the theme issue ‘Ageing and sociality: why, when and how does sociality change ageing patterns?’

1. Background

Our social environment can have a significant impact on our cognitive and physical ageing [1,2]. Clinical studies find that peoples' health and longevity correlate with a variety of social factors such as being married, living with extended family, social network size, social activities, social satisfaction and different social stressors [1,3–5]. These studies cover many health-related questions, but cognitive decline (i.e. dementia), physical disability, inflammation and nutrition are topics of particular interest [6–11]. Resolving the role of social environment in these four areas of health could lead to significant improvements in future strategies of elder care [12].

Cognitive decline is a symptom of many age-related neurodegenerative disorders such as dementia, Alzheimer's disease and Parkinson's disease. Dementia is not one disease but a syndrome of reduced cognitive functioning that goes beyond what is expected in successful ageing. Elderly that report being married, having many satisfying social relationships, and large social networks are less likely to be affected [1,3,13,14]. Social connectivity, however, does not protect against physical brain pathologies during ageing [13]. Social contact may thus enable cognitive functions to persist although the brain suffers from neurodegeneration. However, it is also possible that sociable people innately have brains that can endure more neurodegenerative damage independent of how much social contact they experience [15]. These competing explanations could be tested if peoples' private lives were manipulated so that social experiences no longer correlated with personal social inclinations. Such studies, however, would be difficult to perform—both technically and ethically.

Physical disability correlates with being alone or feeling alone during ageing [16,17]. Likewise, participation in a robust social network can have a protective effect on the prevalence and incidence of disability [18]. Perhaps psychosocial factors directly interfere with the physical health of old people [19,20], or perhaps lonely human beings are less active and as a result become frail [21–23]. These alternatives can be solved by surveying the activities of people that report different levels of loneliness. Physical activity, however, is usually self-reported and reporting can be biased by emotions like loneliness. Trained personnel can observe or standardize peoples’ physical activities, but imposed situations may have emotional side effects that also interfere with data interpretation [19]. Manipulating peoples' loneliness to study effects on health and activity is also generally unfeasible and unethical.

Pro-inflammatory mediators such as tumour necrosis factor (TNF)-α, interleukin (IL)-1β, IL-6, cyclooxygenase-2 and inducible nitric oxide synthase are elevated in many diseases, and levels tend to increase during ageing [24,25]. Social stress is also a predictor of heightened inflammatory activity in middle-aged and old people [6,26], while positive social ties might attenuate inflammation-driven health problems [27,28]. Still, it is unclear how these correlations arise. Perhaps social relationships modulate peoples’ inflammatory status, or perhaps individuals with low pro-inflammatory activity are more sociable. One can resolve these alternatives by manipulative research that decouples peoples' social desires from their actual social experiences, but again, such experiments remain technically and ethically unfeasible.

Central obesity can increase the risk of cardiovascular disease, diabetes 2 and physical disability over the majority of the human lifespan [29,30]. Social relationships influence food intake, body image and weight: people adjust their food intake to mimic or please a social companion [31,32], eat more in larger groups [33], and gain weight when friends become heavier [34]. The regulation of these behaviours is incompletely understood [31], and there is limited attention on whether such phenomena of ‘behavioural modelling’ or ‘copying’ can be tapped into to improve nutrition in the elderly.

While connections are proposed, it is largely unresolved how social influences impact the biology of ageing. Experiments that go beyond correlation are few or missing, and it is practically and ethically difficult to design manipulative experiments with human participants. Nevertheless, the challenges imposed by our increasingly aged global human population require we address these knowledge gaps. It is thus timely to ask if animal models can help reveal mechanisms that connect social relationships to ageing outcomes. Here, we present a selection of insect systems that can be employed in ageing research, we propose lines of investigation that can take advantage of these models, and we discuss the relevance of expanding these comparisons to broader questions about the ecology and evolution of ageing.

2. Insect models of social behaviour

(a). A solitary insect, Drosophila melanogaster

Though they lack advanced social behaviour, fruit flies respond adaptively to other individuals (including dead conspecifics) and can therefore model (some) social processes [35–37]. One such paradigm is social niche construction, where flies shape their own environment through behaviour [38]. Males optimize mating success by constructing social milieus that match their ability to win aggressive encounters: winners benefit from aggressive niches, while losers benefit from foregoing aggression. Flies may also use social information to make reproductive choices: sexually inexperienced females copy the male choices of experienced females [39], and also prefer the food substrates they associate with mated females and their eggs for own egg-laying [40]. For both sexes, social interactions are associated with reduced lifespan the underlying physiological mechanisms are nuanced and vary between sexes [41,42].

Complementary to these behavioural paradigms, the fruit fly system has an impressive toolkit for neurogenetic analysis of behaviour [35]. Transgene fly stocks, RNA interference (RNAi) lines and other mutant Drosophila strains provide genetic models for age-related disorders such as Alzheimer's and Parkinson's disease [43–46]. Transgenesis is required to model human Alzheimer's pathology in the flies because insects do not form neurodegenerative tangles [47]. The Drosophila lines that model Parkinson's pathology, moreover, lose dopaminergic neurons and develop motor behavioural defects that do not occur during normal ageing in flies [48]. Such models can be combined with behavioural paradigms to study connections between social behaviour and age-related disorders [49,50]. In any case, these experiments should factor in that normal fly behaviours may not interface correctly with pathologies from introduced or changed genetic elements. Such mismatches could obscure the cause-effect relationships we want to study.

(b). A gregarious insect, Schistocerca gregaria

The desert locust (Schistocerca gregaria) can reversibly transform between a solitary phenotype that avoids conspecifics and a gregarious phenotype that forms large groups. Gregarious behaviour (tendency to stay in groups) is triggered by crowding: touch, visual and olfactory contact with others [51]. The behavioural phenotypes of gregarization occur within a few hours, while changes in body and brain morphology, sensory and endocrine physiology, colour and life-history strategy occur over multiple generations. These transgenerational changes occur because (female) crowding induces the epigenetic programming of eggs so new individuals hatch with more gregarious phenotypes [52]. With this mechanism, S. gregaria might provide a suitable system for exploring transgenerational physiological effects of social interactions, a phenomenon also observed in humans [53]. Today, expressed sequence tag databases provide a resource for locust genomic studies while the full genome sequence is pending. RNAi-mediated gene knockdowns are established [54,55], and additional functional genomic tools are underway.

(c). Social insects

Social insects share resources and reproduce cooperatively, forming communities or colonies with individuals of the same species [56]. The most advanced form of insect social living, called eusociality, is defined by division of labour between reproductives and helpers (workers), cooperative care for offspring, and overlapping generations in the colony. Highly eusocial behaviour is found in all living termites, the majority of ants, in honeybees, bumblebees and stingless bees, and the ambrosia beetle Austroplatypus incompertus. Several different forms of social organization are found in other social insects, and this diversity is widely used to probe proximate and ultimate mechanisms of social behaviour [36].

Insect division of labour has led to amazing organizational achievements in colonies (figure 1), and can be based on flexible behavioural biases between similar individuals as well as permanent task allocations among morphological specialized individuals [56]. Behaviour biases include care behaviour such as cleaning, tending and feeding others (nursing), and defensive activities like patrolling and attacking intruders (guarding). Morphologically similar workers of honeybees, stingless bees, several ants and wasps show a flexible division of labour for these and other social tasks like nest hygiene, construction, foraging, food handling and thermoregulation. Workers may initially nurse and later forage or defend the colony, so they perform several tasks during their lifetimes [60]. Permanent morphological specializations are seen in highly eusocial insects where reproductives are distinct from workers that are either functionally or completely sterile. Further morphological differences between workers are found in termites, some ants and a bee, and include enlarged mandibles and head armour in termite soldiers, and enlarged jaws and head in ants responsible for leaf-cutting [57]. Among the social insects, the honeybee worker is a well-established model in behavioural, neurobiological and gerontological research [61]. The extensive list of experimental paradigms in honeybee workers includes social learning, decision-making, learning and memory, and questions asked within the context of ageing take advantage of the flexible ageing trajectory of honeybee workers. This trajectory is partially determined by colony age demographics which can be manipulated under semi-wild conditions [62,63].

Figure 1.

Division of labour in insects is associated with ‘cultural’ achievements such as (a) ‘agriculture’ in ants, termites and the ambrosia beetle that grow fungi inside their colonies, (b) ‘architecture’ in termites and ants that build cities with waste disposal systems, ventilation shafts, gardens, nurseries and other amenities [57], and (c) ‘warfare’ of ritualized tournaments, raiding and enslavement between ants [58]. The dance ‘language’ of honeybees (d), moreover, is a unique form of abstract communication between active foragers and new recruits, in which foragers use movement to communicate the value, direction and flight distance to food sources [59].

Although lacking the extensive genetic toolkit of Drosophila, several social insect genomes are either sequenced or in the process of being completed, including the honeybee Apis mellifera, the fire ant Solenopsis invicta, the red harvester ant Pogonomyrmex barbatus, the invasive Argentine ant Linepithema humile and the bumblebees Bombus terrestris and Bombus impatiens (see [60] for a commentary) [64]. The sequencing of social insect genomes has paved the way for researchers to use the CRISPR/Cas9 system to understand social insect physiology and behaviour. This genetic tool has been employed to understand the role of antennal lobe glomeruli in the social behaviour of the clonal raider ant, Ooceraea biroi, and the nutritional regulation of dimorphic size development of reproductive organs in honeybees [65,66]. In addition, RNAi-mediated gene knockdown is successful in eggs, larvae and adult workers honeybees [67–69], in female reproductives of S. invicta and Camponotus floridanus [70,71], and in larvae of the primitively eusocial wasp Polistes metricus [72]. Additionally, the genome biology of A. mellifera includes dynamic CpG DNA methylation, which is a crucial epigenetic system for gene regulation in vertebrates.

3. Lines of investigation

(a). Cognitive function

Cognitive function includes all mental processes that involve symbolic operations such as perception, complex learning and memory, and it encompasses awareness, thinking and capacity for judgement. It is discussed whether insects have cognitive capacities, but at least some species solve complex computational problems that including rule learning, non-elemental learning and delayed-matching-to-sample tasks [73–75]. These abilities are best documented in honeybees [76], which also solve intricate categorization tasks and show numerical processing abilities for numbers up to and including four [77,78]. Some complex brain functions known from honeybees are absent from Drosophila melanogaster and remain poorly explored in other insects [74]. It is not well resolved how these functions compare between insects and mammals, but central computational brain regions (figure 2) appear to be homologous between the animals [79–81].

Figure 2.

Central computational brain regions (orange) can be homologous between the insects and the vertebrates [79–81]: Stylized mushroom body neuropiles of a honeybee brain (a), and the human cortex (b).

(i). Physiological signatures of cognitive decline

Learning, memory and behavioural deficits are hallmarks of cognitive decline and associated with numerous age-related cognitive disorders in humans. Loneliness has been identified as a risk factor for Alzheimer's disease and dementia [82], while perceived social support and both physical and virtual (i.e. social media) social engagement associate with improvements in multiple aspects of cognitive function [83–85]. Similar relationships between ageing, cognitive decline, and the protective effects of social interaction can be found in the worker caste of A. mellifera. Honeybee workers begin to senescence approximately 14 days after their initial foraging flight. The age at which a worker takes its initial foraging flight is independent of chronological age, rather, determined by a combination of genetic, environmental, and social influences [61]. Senescent foragers tend to have decreased learning and memory ability compared to younger foragers and in-hive nurses, however, the progression of senescence itself can be modulated by social factors such as brood presence [86]. Fascinatingly, cognitive decline in honeybee workers can also be reversed by experimentally removing young in-hive bees from the colony (i.e. altering colony age demographics) (figure 3). This manipulation induces senescent honeybees to revert back to performing in-hive tasks such as brood care. This switch to tasks associated with young worker bees is accompanied by a relative increase in learning and memory ability over workers that continue foraging [87]. These unique ageing trajectories of honeybee workers make them a tractable insect model for elucidating mechanistic relationships between social contact and cognitive function, a pressing yet difficult-to-approach question in human studies.

Figure 3.

Peroxiredoxins (Prxs, red dots) can have protective effects in the ageing brain. It is not clear if Prxs respond to social contact between people (top panel), but one enzyme, Prx-6, is elevated in old bee brains that experience improved function after a social change. This social change involves the removal of young bees that do in-nest duties in the colony, such that old bees must switch from foraging to nursing larvae (mid panels). This task-switch is a behavioural reversion for the bees, because about all of them were nurses before they became foragers. Old reverted nurses that improve in brain function have greater than twofold higher Prx-6 levels than those bees that continue to forage (bottom panel).

Other social insects may serve as excellent models for understanding how social isolation impacts higher order cognitive processes necessary for the maintenance of social structures. Individuals of the paper wasp species, Polistes fuscatus, use their unique yellow and black facial markings as visual signals of individual identity, which they use to establish social dominance hierarchies within a colony [88]. Workers that are isolated from conspecifics during the first 6 days after eclosion do not develop individual recognition, which suggests that social interaction is a crucial component of brain development in young wasps [89]. A similar phenomenon is observed in ant species, where social isolation impairs the ability of ants to recognize individuals via hydrocarbon profile, as well as the hydrocarbon profiles themselves [90–92]. Individuals of the ant species, C. floridanus, reared in isolation have attenuated mushroom bodies, as well as reduced aggression when presented with a dead ant (a stimulus which induces an aggressive response to normally reared individuals) [93]. This suggests that social interaction is necessary for the proper development of the mushroom bodies, the seat of learning and memory in insects. Further studies on the interactions between social environment and brain development in insects may unravel fundamental concepts on how early social experiences can shape an individual's lifespan.

Brain senescence can manifest through a variety of means, including cellular senescence, inflammation, protein aggregation and neuronal damage [94,95]. These senescent processes can be triggered through a variety of genetic, disease or trauma-related events, and often underlie the cognitive impairments observed in both healthy ageing and age-related brain pathology [94]. An outstanding question in the social modulation of ageing is whether social interaction directly alleviates senescent processes (i.e. telomere shortening, tau aggregations), promotes resilience in cognitive function amidst senescent processes, or simply whether social individuals are intrinsically resilient to cognitive impairment. Experiments with the honeybee worker model seem to support the former two possibilities, while the minor workers of another eusocial insect, Pheidole dentata, support the latter. Similar to honeybee workers, P. dentata minors increase their task repertoire with age, eventually performing outside-nest tasks such as foraging [96]. However, aged P. dentata minors experience negligible senescence in brain structures that regulate task performance, and in fact display an increased ability to switch tasks given colony needs. Old P. dentata workers retain a full range of sensory abilities, as well as increased titers of serotonin and dopamine in the brain. These counterintuitive characteristics of older individuals may ultimately be related to the lower neurometabolic costs of individuals in highly integrated societies which can practice ‘collective cognition’ [97]. Although this ultimate hypothesis may not translate to humans, investigating the intrinsic resilience of P. dentata minors to senescence may illuminate fundamental proximate mechanisms that govern variation in our own intrinsic resilience to senescence.

(ii). Molecular signatures of cognitive decline

Protein kinase A (PKA) is a multi-functional family of cyclic AMP-dependent enzymes that can play a role in maturation of amyloid precursor proteins in Alzheimer's disease [98]. Brain ageing in Drosophila is also exacerbated by PKA activity in brain structures called mushroom bodies [99] that are homologous to vertebrate cortex (figure 2) [79]. It is unclear whether PKA can be modulated socially in the human brain, but levels are changed in brains from rats that experience chronic social defeat [100], as well as altered in the frontal cortex of people with regressive autism [101]. Changes in brain PKA gene expression can also regulate gregarization in desert locusts [102] and correlate with variation in social behaviour and division of labour between honeybees [103]. Thus, it can be worthwhile to investigate the social modulability of PKA and use insect models to explore its connections to cognitive ageing.

Peroxiredoxins (Prxs) are an enzyme family of sulfhydryl-dependent peroxidases. Six isozymes (Prx1–6) have been found in mammals. Prxs appear to have protective roles in neurodegenerative disorders involving oxidative and inflammatory stress, including Alzheimer's (Prx1, Prx2, Prx6), HIV-associated dementia (Prx3) and Huntington's disease (Prx1, Prx3, Prx6, 102, 125), and the enzymes are targets of drug discovery for treatments of dementia [104–107]. Perhaps similarly, elevated Prx6 levels correlate with improved brain function in old A. mellifera. Cognitive deficits develop in aged worker bees that forage for food, but foragers which revert to in-hive duties once (young) in-hive bees are removed from the colony show brain Prx6 levels that are more than twofold elevated (figure 3) [108]. It is not clear if Prx1–6 can respond to changes in peoples' social lives, but the human Prx enzymes show circadian changes [109] and circadian biology can often be modulated by social factors—in insects and people alike [110,111].

Excitatory amino acid transporter-2 (EAAT2) is another protein that correlates with recovery of brain performance after senesced foragers revert to in-nest tasks [108]. In contrast to Prx6 that is elevated, EAAT2 is reduced greater than twofold in reverted individuals that recover from ageing. EAAT2 is an important transporter of the excitatory neurotransmitter glutamate, and EAAT2 dysregulation has been implicated in Alzheimer's and Huntington's disease [112]. The regulation of EAAT2 has not been firmly connected to social factors in mammals, but old rats that experienced maternal separation can show elevated EAAT2 expression in the hippocampus [113]. Maternal separation is a juvenile stress paradigm that increases the risk of Alzheimer's pathology during ageing [114,115]. Perhaps research on insects can give more insight into how social relationships can affect EAAT2.

(b). Physical disability

This blanket term covers dysfunctions of limbs and fine or gross motor ability that leads to physical impairments, activity limitations and participation restrictions, which are common during ageing. Old people are also at higher risk of biomechanical wear, falling and fracturing [116,117]. Insects can experience similar losses of physical abilities, and old individuals often fly or walk less and spend more time stationary or in a supine position [118]. In Mediterranean fruit flies (Ceratitis capitata), the first occurrence of supine behaviour predicts remaining length of life, like the onset of physical disability correlates with time of death in elderly people [119]. Thus, it has been argued that insects can help identify common principles in the disablement process of ageing as well as assist in the discovery of preclinical markers of disablement [118,120].

Socially induced physical activity (SIPA) is motor activity that results from social motivation or contact. Physical activity protects against disability, reduces the risks of dementia and morbidity, and promotes health even in the oldest old humans (greater than or equal to 85 years, see (128) for an example) [121–123]. Emotional support from social networks correlates with better physical performance in elderly [124], and social motivation can impact the quantity and frequency of physical activity at any age [125]. Even so, research on D. melanogaster indicates that SIPA effects may be conditional on several factors (figure 4). Scientists used a mutant fly stock with genetic changes in antioxidant enzyme Cu/Zn superoxide. These flies are less stress resistant and shorter lived than normal flies [126]. Mutant lifespan and climbing ability, which served as a measure of physical performance, was improved by social contact with young, healthy, and normal (wild-type) individuals. Social contact with old wild-type flies, by contrast, did not prolong the mutant's lifespan. Disabling the young companion flies also had negative implications: mutant lifespan was less extended or even shortened, depending on how severely disabled the companions had become [126]. These results (figure 4) might imply that SIPA is less effective, less beneficial, or even harmful in social networks that are dominated by old or disabled individuals.

Figure 4.

Mutant fruit flies (mutant) with genetic changes to antioxidant enzyme Cu/Zn superoxide are less stress resistant and live shorter lives than normal wild-type (wild) flies (top panel). Mutant lifespan is improved by social contact with young wild-type individuals, but not by contact with old wild-type flies (mid panel). Contact with disabled wild-type flies can shorten the lifespan of mutants even more (bottom panel).

Social role valorization (SRV) is a theory about the enablement of valued social roles for people. Social devaluation can lead elderly to devalue their residual capacities even further [127], and poor self-perceived abilities is a risk-factor of depression, immobility and health decline during the ageing process [128,129]. Improvements in physical fitness can directly counteract these symptoms by promoting a sense of self-perceived ability [130]. Interestingly, some insect societies have task systems that enable contributions from aged and disabled animals. In the leaf-cutter ant Atta cephalotes, workers divide labour based on size [57]. Tiny workers tend the fungus garden (figure 1), eggs and larvae, and they patrol the terrain around the colony in search of intruders. Larger workers are foragers that perform tasks of leaf-cutting and transport. Cutting is achieved by chewing, causing progressive teeth-wear that compromises old ants' ability to harvest plant material. Instead, the old ants transport leaves cut by other foragers [131].

The same Atta is known for its system of hazardous waste management. Waste materials from the colony are potentially infectious, and specific workers handle the waste following a convey belt model, with increasingly dangerous tasks being performed by older individuals [132]. This role allocation is functional: compared to young ants, the total lifespan of old individuals is minimally affected by the danger, because the remaining life-expectancy of old ants is short regardless of what they do. A similar rationale was put forward by the Skilled Veterans Corps after the 2011 nuclear crisis at the Fukushima power station in Japan. The Corps of retired engineers and professionals volunteered to stabilize the plant, arguing that they would be less impacted by long-term health risks than younger people [133]. Here, it was age that modulated task allocation, providing an important social role to a group traditionally at risk of social devaluation. In this context, SRV theory does not propose solutions, but provides a framework for satisfying the needs and motivations of old individuals in the society they belong to. Perhaps studies of groups at risk of social devaluation and their functional roles in insect societies can stimulate ideas of role-valorization in our own communities.

(c). Inflammatory responses

Inflammation attempts to eliminate harmful agents, irritants or damaged tissue, and initiate healing after cell injury. The process is part of the innate immune response that has conserved molecular mediators in humans and insects [134,135]. Many age-related disorders in people occur together with abnormally persistent inflammation, but elevated inflammatory activity (inflammaging) is also present in normal ageing, perhaps promoting neurodegeneration, depression and loss of muscle mass [25,136]. Pro-inflammatory markers that increase in elderly (TNF-α, IL-1β, reactive oxygen species, and others) can activate the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) transcription factor that controls a battery of genes in innate immunity, inflammation and ageing. An NF-κB inhibitor was recently tested in D. melanogaster where it extended lifespan [137]. Overall, insects provide many tools for studying links between ageing and the immune system [138].

Sleep loss elevates NF-κB in both humans and D. melanogaster [139–141]. This increase could translate sleep disturbances into systemic inflammation and increase the risk of diseases. Ageing often comes with disruption to the daily sleep-wake cycle, with as many as 50% of elderly reporting difficulties in sleeping [142,143]. At the same time, stable relationship histories such as marriage or having a partner might promote human health span by positive effects on sleep [144,145]. Nuances exist in the relationship between marriage and health, as the act of staying married correlates with lower levels of inflammatory marker activity in older men, while women tend to have higher levels of inflammation if the quality of their marriage is low [146,147]. Social network support among elderly correlates with better sleep [148], while social activities combined with exercise can improve sleep patterns in nursing home and assisted living residences [149]. In Drosophila, cockroach and many social insects, circadian behaviours like sleep can be manipulated in social experiments [150–154]. Such methods could probe for causal connections, for instance between NF-κB regulation and clock-related proteins with roles in immunity and inflammation—such as the Prxs enzymes [155,156].

Social stress correlates with higher inflammatory marker activity in young, middle-aged and old adults [26,100]. This pattern is seen after stressful episodes and during chronic social stress [157]. Social relationships with conflicts, rejection, mistrust or instability could, thereby, contribute to low-grade systemic inflammation and poor health. It is, however, challenging to test this connection because many additional factors correlate with relationship qualities, including socio-economic status, environmental risks of aggression or depression, and several health behaviours including sleep [158]. Complex connections between social factors and innate immunity are also found in insects such as the ant Harpegnathos saltator, where individuals that achieve upward social mobility become highly resistant to infection and live longer lives while those that experience social stress have increased susceptibility to infection [159,160]. Possibly, links between social status and mobility, relationship qualities, stress, innate immunity and inflammation, can lead to a ‘spiraling’ (a positive or negative reinforcement) of individual health outcomes. In the former example of H. saltator, the outcome is positive, in the latter, negative. In another example from a different social insect, the outcome is also negative: honeybee workers with lipopolysaccharide (LPS)-induced systemic inflammation experience more frequent rejection by healthy social partners [161]. Social rejection causes physical injury, innate immune activation, and mortality in honeybees. LPS is also an inflammatory inducer in mammals [162], and the mammalian LPS-response can probably be modulated by social partners [163]. Stress associated with opposite-sex contact also exists in Drosophila, where both sexes suffer reduced lifespan upon interacting with the opposite sex [41,42]. The negative impact of social interaction on lifespan is greater in males, and interestingly, this effect is also seen in genetically masculinized female flies. [164]. Perhaps these connections can be better understood by further research on insect immunity and social interactions.

(d). Nutrition

Many elderly loose appetite and develop an anorectic syndrome [165,166], but an increasing number of obese individuals become elderly and these individuals may stay overweight for the rest of their lives [167]. At the same time, the body weight that confers maximal survival is not constant over the life-course: it is higher in old people than in middle-aged adults, and this ‘obesity-mortality paradox’ causes debate about criteria for recommending weight-management strategies to elderly [168]. Also, several social insects show significant and stable weight loss among old individuals and offer many opportunities for studying social aspects of diet including food consumption and food choice [169–171]. Drosophila moreover, can provide genetic models for a number of weight-related pathologies because the pathways involved in metabolic biology are highly conserved between the species [172].

Food consumption is often reduced after age 70–75 years and leads to loss of body fat and lean muscle. People with ‘anorexia of ageing’ are at risk of pathological weight loss, malnutrition, morbidity and mortality. This syndrome is not completely understood, but dysregulation of orexigenic (appetite increasing) and anorexigenic (appetite reducing) signal peptides from brain, pancreas, adipose tissue and gastrointestinal tract is likely to play a role [165,166]. Social relationships also affect elder nutrition and eating behaviour (figure 5). These patterns are not well explained mechanistically, but appetite-controlling molecules like the anorexigenic peptide hormone leptin can be influenced by social situations [175–177]. In honeybees, abdominal adiposity, carbohydrate metabolism and gustation is linked to signalling by juvenile hormone [69,169], which is similar in structure, function and mode of action to vertebrate thyroid hormone [178]. Juvenile hormone is sensitive to social interactions, and young workers kept in social isolation develop levels identical to those of old bees. Social contact with only 2–3 other workers is sufficient to normalize the hormone level [179]. Drosophila larvae also benefit from group feeding in natural conditions, as feeding aggregates are able to dig ‘wells’ into deeper, nutrient-rich layers of decaying plant matter. Groups tend to be larger when more closely related larvae are in the local environments, and more local kin is associated with higher survival rates and greater female weights (a robust marker of lifetime fecundity) [180]. Thus, the ability for a fly to recognize and cooperate with kin within a social environment may be beneficial for overall fitness (at least in females) and may serve as a useful model for the effects of group feeding on individual physiology.

Figure 5.

From left to right: old people who live alone or report feeling lonely consume fewer daily meals and total calories, tend to eat less protein, fruits and vegetables, and are at increased risk of being underweight and malnourished compared to those who have a partner or live in traditional family surroundings [173]. Hospitalized elderly, similarly, increase their intake of protein and total calories when a personal volunteer, rather than a professional nurse, socialize with them and assist with feeding at mealtimes [174].

Food choice is motivated by many factors including sensory appeal. Sensory capacities like taste and smell are often diminished during ageing [166,181]. Sensory changes can make familiar meals seem undesirable [182], and push elderly individuals to prefer meals with more intense flavours (i.e. salty, sweet), but which lack proper nutritional value [183]. Interestingly, research on humans and many animals show that social modelling (figure 6) can increase food acceptance for meals that are novel, different or varied [31,184]. Drosophila can use social modelling to make decisions about food [40,185], but there is surprisingly limited evidence for whether or not (having) choices is favoured by animals like insects [186].

Figure 6.

Young children accept novel foods more readily after observing their mother eating them (left panel), while adults modulate their food intake to match that of social companions (right panel). Perhaps care strategies that incorporate aspects of this social modelling can help maintain a varied and calorically sufficient diet in the elderly.

Grasshoppers can optimize development and growth if they mix their diet, but increased feeding time in mixtures suggest inefficient decision making when many options are available to animals [186,187]. Perhaps in contrast, social insects overcome the difficulty of multiple choices through collective decision making. These decisions are based on communication between animals with limited individual information. Each animal knows about one or perhaps some options, but together, the group has assessed many options and arrives at decisions together. In the ant Temnothorax rugatulus, for instance, groups handle dual versus multiple choices (eight options) with equal efficiency, while single ants are less able to make good decisions about eight options [188]. Honeybees, moreover, use collective decision making during foraging [189] but food-choice behaviour also has genetic components in the bee [190]. Genes involved in ovarian and insulin-associated pathways affect the bees' (sweet) taste and food foraging bias [191,192]. Similar relationships have been reported in humans [193,194], but it is unclear how these mechanisms respond to social contact.

Metabolic syndrome describes a group of disorders and conditions that increases the risk of cardiovascular disease, diabetes and mortality. The syndrome becomes progressively more common and poses many challenges to human quality of life, health and economy [195]. Metabolic syndrome is more prevalent in obese people [196], but even though obesity predicts higher mortality in 75–84 year-olds, it has protective effects from age 85 and onwards [168,197,198]. These results suggest that the oldest old should not be encouraged to lose weight or avoid obesity, although the obesity-mortality paradox of ageing remains an active area of study [197,199].

As core metabolic pathways regulating energy homeostasis are highly conserved, insect models have proven useful for unravelling aspects of this paradox. Drosophila can be made obese by feeding a high-sugar diet, high-fat diet, calorie restricted diet, or by genetic manipulation. Each of these manipulations acts on a different set of tissues and pathways linked to lifespan and can be used interchangeably to tease apart the effects of diet on lifespan [200]. Much of this research centres on the effects of diet on the highly conserved insulin/insulin-like growth factor (IIS). IIS reduction is associated with high-sugar and high-fat diets, similar to the diabetic phenotype in humans. However, whereas reduced insulin signalling in humans has negative health implications, its reduction in flies is associated with increased lifespan [172,200]. Here, the honeybee model may be better suited to unravel the connections between insulin signalling and lifespan regulation. Ihle et al. recently discovered that reduction in insulin signalling is associated with an accelerated behavioural transition to foraging, resulting in a reduction in total lifespan [201]. In a natural context, the worker transition to foraging is partially regulated by the needs of the colony, suggesting an interplay between social environment, metabolic pathways, and lifespan in honeybees.

4. Ecological processes

Ecology is the study of environmental systems that focus on relationships between living organisms and the natural world. Humans and many other advanced social species have unique strategies for how they interface with nature, perhaps particularly with regard to the scale by which they explore, shape, manage, culture and mine the environment for the extraction of resources (figure 1) [57,202]. From this extraction emerges economies based on the production, distribution and consumption of the resources, as well as derived goods and associated services. Economies create influential environmental systems that affect individual activities, resource availability, social status and health [203]. Ecology is often called the economy of nature, but it does not typically extend its scope to environments (like economies) that emerge from social organization. Nonetheless, we believe a focus on social systems is essential in a discussion of social impacts of/on the ecology of ageing. This is an established concept in elder care where ‘ecology of ageing’ often relates to environments that are highly organized, including neighbourhoods and other living-arrangements for ageing independently (see [188] for an overview) [204].

Never before in the history of the planet has the world contained so many elderly, or such a large fraction of them compared to young people [205]. This situation raises many questions in societies, including how economies are impacted by the elderly and how economic forces impact their lives, or ecology. It is commonly claimed that current elder care systems are unsustainable, and this concern has inspired research on potential solutions. One is ‘smart house’ (robotic) technologies for independent living that cut the costs of having actual (human) carers [206]. A different realization is that elderly represent a substantial and predictably growing consumer group [207]. National and international innovation, development and trade in healthcare services may thus become a major economic indicator in the future. If any of these solutions and predictions come through, it may profoundly change the environment of elderly.

Research on ageing in model animals is largely focused on the immediate biological and environmental causes of senescence. Such data do not provide a good basis for discussing how the economics of societies can impact the ecology of ageing. Economies of resource gathering and allocation, however, are studied in several social insect species [57,202]. Social insects use a variety of strategies for harvesting and processing natural resources that fuel colony growth, reproduction and survival. Materials and foods that are brought to the nest are often processed in one or several steps involving different worker groups, before the resulting goods are distributed to members or areas of the colony [60]. Old individuals typically have specific roles in these processing and distribution chains, defining their ecology of ageing.

Old individuals usually do risky tasks in insect societies (figure 7) [132,208–210]. This work includes foraging for food, construction materials, resins and water, as well as defensive strategies, waste handling and removal. These tasks expose the animals to wear, predators, and pathogens that are not encountered by other workers [211]. Old individuals may also be treated differently—and at least partly because of their exposures: old worker honeybees receive less social feeding than younger adults (reviewed in [194]) [210], whereas old worker ants contaminated with waste and similarly immune-compromised worker bees can be rejected by other colony members [212,213]. Such environments of risk and rejection are detrimental to lifespan [211]. However, the allocation of old, undernourished and infected individuals to this ecology is good colony economy because the animals' remaining life-expectancy is short—regardless. Social insect ‘economic policy’ is then, in short, to minimize physical costs in the young population and to balance resource use with remaining life-expectancy in the elderly.

Figure 7.

In honeybee societies, younger individuals (left panel) work in the protected nest while older individuals (right panel) take care of risky tasks like foraging. This age-associated division of labour is good colony economy. The total lifespan of old bees is less affected by the dangers of foraging because the remaining life-expectancies of the individuals are short (red arrow) regardless of the tasks they perform for the colony.

Statements like ‘protecting the young’ and ‘balance resource use’ can resonate with people. However, the ecology of ageing in social insect colonies is clearly not a useful model for the future of elder care in human societies. Most animal models come with advantages and disadvantages, and each system can only address specific questions [214]. Insect models can contribute to deciphering specific genetic, molecular and physiological mechanisms that connect social contact to ageing, while being more broadly useful in questions of how resource acquisition and allocation (economy) can impact the ecology of ageing.

5. Evolutionary processes

Species age at different time-scales [215]. Some animals and plants age rapidly, like the nematode worm Caenorhabditis elegans and the thale cress Arabidopsis thaliana that have lifespans of about 30 days. Others age slowly or even negligibly, like the rockfish Sebastes aleutianus and the bristlecone pine Pinus longaeva with lifespans of about 200 and 4500 years, respectively [215,216]. Evolutionary theories of ageing seek to explain such differences and why mortality increases with age. They predict that there is no selection for longevity beyond the period of reproductive capability (reviewed by [194]) [210]. Social species, however, can provide an exception. This is because post-reproductive individuals in social groups can gain (inclusive) fitness by helping their offspring reproduce. Lee developed a theory of ageing that accounts for helping in the form of ‘intergenerational resource transfers’ [217]. The theory linked an economic model for resource use in caring (such as feeding, warming, fanning and guarding) to predictions of age-specific mortality. Lee's formalism presents the force of selection on longevity as a weighted average of remaining fertility and remaining resource transfers (help) to be given to others, and, it shows that the remaining capacity to help at any age is what shapes senescence in social species with continuing care for offspring. This conclusion is in line with the ‘grandmother hypothesis’, which posits that grandmothers have enhanced their children's’ fitness and the survival of their grandchildren through helping [218].

Evolutionary implications of Lee's theory and the grandmother hypothesis are supported by results from pre-modern human populations, where women with prolonged post-reproductive lifespans had more grandchildren, and hence greater inclusive fitness [218]. It is, on the other hand, not demonstrated that the inclusive fitness benefits of grandmothering are large, and recently, it was suggested that the benefits of helping are not large enough to explain why women outlive their reproductive period by several decades [219]. Inclusive fitness theory has also been used to understand helping in social insects, and it is questioned whether inclusive fitness benefits are sufficient to explain that workers (sterile helpers) evolved [202]. These debates are still ongoing, and perhaps defines an area where scientists can come together to ask broader questions about the role of helping in the evolution of social life histories.

6. Concluding remarks

Geriatrics is becoming an increasingly important specialty, and research on ageing is an engaging, proactive science [220]. Although many ‘smart’ technologies are accessible to elder care, there is resistance against substituting ‘warm’ human hands with the ‘cold’ technology [221]. Our review indicates that some of this resistance can be appropriate. A number of studies reveal significant correlations between social factors, health and ageing in human populations, but it is difficult to extract specific and causal relationships from the clinical surveys. Senescence results from a cumulative imbalance between damage and repair, and insects may help reveal whether and how social processes can reduce damage and increase repair. This understanding may lead to new elder care strategies that can improve the health and quality of life for many people.

Data accessibility

This article has no additional data.

Authors' contributions

T.P.Q. and G.V.A. contributed equally to this manuscript.

Competing Interests

We declare we have no competing interests.

Funding

The work was supported by the Norwegian Research Council (180504, 191699), The PEW Charitable Trust and the National Institute on Aging (PO1AG22500).