1 Introduction

It has been known for millennia that muscle and fat wasting leads to poor outcomes including deaths in chronic disease states [1]. This strong and practically undeniable association is also observed in any acute condition that is associated with accelerated wasting such as acute illnesses that progress to unintentional weight loss including in critically ill patients. Circumstances that lead to inadequate nutrition are also associated with increased mortality including starvation due to natural disasters (war, famine, earthquakes, and other natural and man-made disasters), anorexia nervosa, intentional starvation such as in hunger strike, and inadequate nutritional support in patients on life support in critical care units or in terminal cancer or other terminal conditions where nutritional support is chosen to be withdrawn. In geriatrics, the aging-associated sarcopenia is another well-known predictor of death, even though the aging process per se is not an acute or chronic illness. Whereas the term “cachexia” is used to refer to a more severe form of weight loss with a low body mass index (BMI) (e.g., BMI <18.5 or <20.0 kg/m2), any wasting condition whereby muscle or fat is lost—sometimes referred to as “pre-cachexia”—can be a prelude to overt cachexia and may have a bearing on life-span [1].

Despite the age-old strong and consistent association of cachexia with mortality and despite the enormous public health burden of cachexia [2], it is surprising that to date no unifying mechanism has been suggested to explain why wasting conditions are associated with an exceptionally high mortality risk. Indeed, there are debates as to whether this association is causal. This century-long drawback may have played an important role in why the field of cachexia treatment research has not shown major advances, especially since it has frequently been argued that cachexia is a mere epiphenomenon and not a true cause of death. In this short review paper, we examine the causality of the cachexia–death association and review briefly potential pathophysiologic mechanisms whereby muscle and fat wasting results in premature death.

2 Hill’s considerations for causality inferences in the observational studies of cachexia

Given the inherent occurrence of cachexia as a universally unwanted terminal condition, it is highly unlikely that an interventional study to randomize human subjects to any cachectic versus non-cachectic state could be considered ethical. Hence, notwithstanding the potential role of animal models, the main source of evidence and inference pertaining to the cachexia–death association will remain observational or epidemiologic in nature. In epidemiological studies, the natural occurrences of the events are observed in a given population without an external “intervention” or “randomization” to an intervention. The ultimate goal of the epidemiological studies is to disclose the unbiased association between the candidate exposures—also called risk factor—and the outcomes of interest in the study population and then to extrapolate it to a larger population [3]. This association, however, is often confounded by third factors. Given the lack of randomization, the epidemiological associations—including those related to the cachexia–death association—are invariably subject to confounding bias. In other words, there is invariably a third factor or confounder, metastatic cancer for instance, that leads to both cachexia and death independent of each other, hence creating an apparent cachexia–death association that would not exist without the underlying cancer. Nevertheless, cachexia may also directly lead to death. We suggest that there are several biologically plausible scenarios pertaining to the cachexia–death association, as shown in Fig. 1.

Fig. 1
figure 1

Three hypothetical “causal” models of the cachexia–death association

Full size image

Given the inherent limitations of epidemiologic studies, it can be argued that no matter how rigorous the confounders are adjusted for in sophisticated multivariate statistical models there may be a number of unknown residual confounders leading to the observed cachexia–death associations without true biologic plausibility [4]. Figure 1 shows a number of these associations. Whereas in model 1 cachexia is the true cause of death, in models 2 and 3, cachexia is a mere epiphenomenon and is secondary to the wasting disease or its underlying risk factors, respectively, which kill patient independent of cachexia (see Fig. 1). Hence, the inability of observational studies to prove causality is massively limited no matter what kind of statistical techniques are employed [5].

Despite the foregoing view about causal inferences in epidemiological studies of the cachexia–death association, there are some considerations to examine the leap from association to causation, the most well known of which was presented in the 1965 article of Sir Austin Bradford Hill, “The Environment and Disease: Association or Causation” [3], in which several benchmarks—subsequently refined and expanded to nine criteria—that “suggest” causality were listed (Table 1). The most important one is the “temporal relationship,” which indicates that the cause or “exposure,” e.g., weight loss to overt cachexia, should precede the effect or “outcome,” for instance death. An essential problem in studying the causes of death is the fact that death is inherently the final event preceded by presence of any risk factor or condition. Hence, temporality is universally present in this association and cannot help differentiate causality.

Table 1 Hill’s considerations for causality inference in observational associations. Each causality benchmark is examined for cachexia–death association
Full size table

It is important to note that even though Hill’s consideration can advance steps in the direction of causal inferences, epidemiology can virtually never prove causality. Table 1 shows pros and cons pertaining to each of the nine Hill’s considerations. Given the inconsistency of these benchmarks in the cachexia–death association, the current state of knowledge makes it virtually impossible to allow naming weight loss or even full-blown cachexia as the main cause of death in chronic and progressive disease states. Nevertheless, the consistency of the associative data, the remarkable strength of the cachexia–death association, the early occurrence of death following progressive weight loss, and emerging evidence from basic science and animal models suggest that the causality element may be present and can soon be discovered.

3 Putative pathophysiology of the cachexia–death link

Several hypotheses have been advanced to explain a biologically plausible association between cachexia and death, as shown in Table 2. We favor the platelet pathway hypothesis, according to which relative thrombocytosis (increased platelet count), potentially further enhanced by platelet function activation or other platelet aspect augmentation such as increase in platelet volume, leads to a predisposition to thromboembolic events including cardiovascular or cerebrovascular events and sudden death. In a recent cohort study of 40,000 people with end-stage kidney disease undergoing chronic hemodialysis treatment, Molnar et al. [6] have shown that relative thrombocytosis, i.e., platelet count >300,000 × 103/μL, was incrementally associated with higher death risk. Interestingly, this association was mitigated after adjusting for surrogates of malnutrition–inflammation complex including protein–energy wasting (PEW), which suggests that the platelet link may be in the causal pathway between PEW and mortality.

Table 2 Putative mechanisms of action underlying the higher mortality associated with protein–energy wasting in chronic kidney disease
Full size table

There are several other reports about the causes and consequences of relative thrombocytosis and its relationship with thromboembolic events and death including in iron deficiency [7, 8]. Similarly, hypothyroidism, which can occur under wasting conditions, may also lead to increased platelet numbers as well as platelet activation [9]. Of note, a recent study in dialysis patients found that plasma level of gelsolin, a marker of healthy muscle mass, is decreased in dialysis patients, while circulating actin, a marker of myocyte damage, is increased [10]. Low gelsolin and high actin levels may activate platelet pathways, hence explaining why sarcopenia and muscle mass loss over time are strongly associated with death risk [10, 11].

Another plausible hypothesis invokes the potentially causal role of arrhythmias in cachectic conditions [12]. This pathway may explain why beta-blockers can improve survival in patients with chronic heart failure, where cardiac cachexia is quite common. It is still not clear what exact mechanisms, mediators, or pathways may lead to higher occurrence rate of arrhythmia-related death in cachectic conditions (if they indeed are present) [13].

Endothelial dysfunction may be another important causal pathway in the cachexia–death association. Progressive wasting leads to fat loss, leading to lower level of anti-inflammatory cytokines such as interleukin 10 and adiponectin, along with the decline in circulating level of lipoproteins. According to the endotoxin–lipoprotein hypothesis by Rauchhaus et al. [14], cachectic patients may be subject to enhanced endotoxin bioactivity in the circulation. Lipoproteins are known to have the ability to block endotocin bioactivity in vivo [14]. The lack of neutralizing circulating lipoproteins may explain why patients in chronic illness often show better survival when lipoprotein levels are high [15, 16]. In hypocholesterolemic patients, the lack of an adequate lipoprotein pool leads to free circulating endotoxins, which per se results in the generation of a pro-inflammatory and oxidative cascade with subsequent endothelial dysfunction and cardiovascular events [17]. Some traditional hypotheses pertaining to the pathophysiologic link between wasting and cachexia are based on the suppressive effect of wasting on the immune system, leading to a propensity for infectious diseases and deaths caused by infectious events. Some hypotheses pertaining to the cause of death in protein–energy wasting of CKD patients are listed in Table 2.

4 Conclusions

Cachexia is a major public health issue [18]. Millions of people with chronic disease states are more likely to die when they lose weight and develop progressive cachexia. There are opposing views as to whether cachexia is a biologically plausible cause of increased death risk in patients with wasting disorders or whether this is a mere epiphenomenon. The pathophysiologic pathways whereby cachexia leads to death are not clear. Increases in platelet numbers and activation pathways, and arrhythmias appear to be causes of sudden cardiovascular events. The field of applied cachexia will remain in its millennium-old infancy as long as there is no major advancement in disclosing the true etiology of death in cachexia.