Low body condition scores and reduced eating time budgets emerged as significant indicators of equine mortality in this study. Analogously, weight loss and diminished appetite are well-established predictors of mortality risk in elderly humans, dogs, and cats without terminal illnesses [40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56]. Beyond serving as a prodromal marker of various life-limiting diseases, weight loss contributes to mortality both directly, through its immediate physiological impacts, and indirectly, by accelerating the development and progression of sarcopenia and frailty, conditions that independently increase mortality risk [25, 43, 54, 57,58,59,60,61].

Notably, in humans, weight loss is associated with an increased mortality risk regardless of baseline weight, even in overweight or obese individuals [40,41,42, 45,46,47, 53, 55, 56]. This phenomenon was also evident in this study where horses’ body condition scores ranged from ideal (BCS = 5) to overweight (BCS = 7), with only a few classified as thin. The equine body condition score is positively correlated with body weight, percentage of body fat, and plasma prostaglandin E2 (PGE2), making it a better indicator of adiposity and the associated increased systemic inflammation than body weight [35, 62]. While obesity in humans predisposes to sarcopenia, weight loss results in concurrent losses of not only body fat but also muscle and bone mass and may thus paradoxically exacerbate age-related sarcopenia and frailty [25, 43, 54, 57,58,59,60,61]. In older adults, where age-related body composition changes already lead to increased fat mass and reduced muscle tissue in both sexes [60, 63], such losses further impair physical function, deplete physiological reserves, and heighten vulnerability to stressors [25, 43, 54, 57,58,59,60,61]. Similarly, in geriatric horses, muscle loss can progress to the point where they are unable to rise from a recumbent position without assistance, a well-documented clinical problem. However, the etiopathogenesis of sarcopenia and frailty in horses has not yet been systematically studied, leaving a gap in our understanding of these conditions in the equine population.

The relationship between weight loss and mortality is more pronounced in men, where a > 10% weight loss is associated with a 289% increase in mortality risk compared to a 114% increase in women [40]. This disparity is attributed to body composition differences, as men typically have a higher muscle and bone mass, whereas women have a greater proportion of fat tissue [40]. In this study, no sex-specific differences were observed in horses, likely reflecting the similar testosterone levels in mares and geldings, both of which are significantly lower than those in stallions.

Poor appetite in the elderly, termed anorexia of aging, frequently results in inadequate food intake, leading to nutritional deficiencies and weight loss [40, 43, 50, 52, 54, 64,65,66,67,68,69,70,71]. Appetite regulation is a complex process orchestrated by the central nervous, endocrine, and sensory systems [40, 43, 50, 52, 54, 64,65,66,67,68,69,70,71]. In older adults, appetite reduction has been attributed to lower plasma levels of the orexigenic hormone ghrelin and elevated concentrations of insulin, leptin, cholecystokinin, and proinflammatory cytokines. Additional contributors include slower gastric emptying, altered taste and smell, dental issues, chronic pain, and psychological or social challenges [40, 43, 50, 52, 54, 64,65,66,67,68,69,70,71]. Ageing horses share many of these challenges, particularly chronic orthopaedic pain and susceptibility to dental problems. However, in this study, regular veterinary care, including routine dental examinations, ruled out oral health problems as a primary cause of weight loss.

While horses that died (DAC) or were euthanized (DCC) exhibited higher pain scores than surviving horses (NC), no significant differences in pain scores were observed between the two deceased groups (DAC mean, 9.3; DCC mean, 11.05). Interestingly, weight loss was observed exclusively in DAC, suggesting contributing factors beyond pain. Further studies examining associations between appetite-regulating hormones, inflammatory cytokines, and appetite changes in geriatric horses are needed to identify species-specific and conserved mechanisms of aging-related weight loss.

While the overall pain score did not differ between DAC and DCC horses, horses that died naturally exhibited significantly higher levels of depression and withdrawal compared to all other groups (DCC, NC, Y) reflected by the score’s subitems ‘position in their enclosure’ and ‘demeanour’. In elderly humans, depression is strongly associated with an increased risk of short- and mid-term mortality, with risk levels comparable to or even exceeding those of major health conditions such as myocardial infarction or diabetes [72,73,74,75,76,77]. Horses can exhibit behavioural and postural profiles that share striking similarities with depressive states in humans. These include indifference to environmental stimuli and human interaction, reduced selective attention, lower plasma cortisol concentrations, diminished sucrose intake (an indicator of anhedonia), reduced head and ear movements, and a fixed gaze [78,79,80,81,82]. However, associations between depressive symptoms and mortality risk in horses have not been previously explored or reported.

This study, like any longitudinal research conducted under field conditions, has limitations that should be considered when interpreting the findings. The control group of healthy horses was relatively small (n = 6), reflecting the specific population of the animal sanctuary, which by nature houses primarily animals with chronic health conditions. Despite the limited control sample size, the statistically significant differences observed between groups indicate that the sample was sufficient to detect meaningful effects. Nonetheless, future studies with larger control groups would be valuable to confirm and expand upon these findings. The sanctuary-based setting also introduces an inherent selection bias, as this population may not be representative of horses kept in different husbandry systems, such as privately owned pleasure horses, working horses, or those in performance contexts. Thus, while the sanctuary provides a unique opportunity for long-term, detailed observation of aging equines, further research is needed to assess the generalizability of these results across diverse equine populations and husbandry environments. In addition, while adjustments were made for major known confounders such as age and sex, residual confounding or multicollinearity cannot be entirely excluded and should be considered when interpreting the findings.

Euthanasia is intended to serve as a humane method to end an animal’s suffering by providing a peaceful and painless death, yet the decision to euthanise is often fraught with ethical and emotional dilemmas for animal owners and veterinarians alike [83,84,85,86,87,88,89,90,91,92,93,94,95,96,97]. These end-of-life decisions are rarely based solely on the animal’s quality of life (QoL), best interests, or medical and ethical considerations. Instead, they are frequently shaped by emotional, economic, social, and legal factors, which add layers of complexity, conflicting responsibilities and conflicts of interest to the decision-making process [83,84,85,86,87,88,89,90,91,92,93,94,95,96,97]. In this study, where the horses were owned by a sanctuary, end-of-life decisions could be made exclusively based on the animal’s best interests, highlighting the challenge of defining ‘best interest’ in the context of equine euthanasia.

Animals cannot weigh the future benefits of life-extending treatments against current suffering, nor can they autonomously request euthanasia [83,84,85,86,87,88,89,90,91,92,93,94,95,96,97]. Consequently, the ethical responsibility of making these decisions rests on caregivers and veterinarians. These decisions are guided by observable signs of quality of life, medical diagnoses, available treatment options, and prognoses, but they require complex and subjective predictions about the animal’s experiences and responses to interventions [83,84,85,86,87,88,89,