Stress is defined as a constellation of events consisting of an external (i.e., environmental, psychosocial, mechanical) or internal (i.e., physical, biogenic) stimulus, whether actual or perceived, that precipitates a reaction in the brain followed by a physiological response nonspecific and specific to the stimulus to maintain homeostasis (Selye, 1950; Selye, 1976; Goldstein and Kopin, 2007; Nicolaides et al., 2015; Sher et al., 2020). In turn, homeostasis is characterized as the physiologic stability between interdependent biological systems (Billman, 2020). As such, stress encompasses an integrated definition to create a three-pronged construct: (i) a stressor (stimulus), (ii) a stress perception (detection and interpretation in the brain), and (iii) a stress response involving the activation of physiological fight-or-flight and neuroendocrine systems that serve as an adaptive mechanism to maintain homeostasis (Finnell et al., 2017; Gianaros et al., 2017; McEwen and Gianaros, 2011; Cohen et al., 2016; Schneiderman et al., 2005).

The physiological stress response involves the integration of different brain regions and neuronal circuits responsible for the detection and interpretation of physical, psychological, or environmental stressors and the pro-survival and adaptive mechanisms that follow (McEwen and Gianaros, 2011; Cohen et al., 2016). Though diverse stressors engage distinct brain regions for processing and interpretation (Gianaros et al., 2017; McEwen and Gianaros, 2011), the initiation of the stress response involves the activation of the hypothalamic-pituitary-adrenal (HPA) and sympathetic-adrenal-medullary (SAM) axes for multi-system (i.e., immune, metabolic, cardiovascular) effects (Sher et al., 2020; Miller et al., 2007). Common military training stressors, including energy deficiency or restriction, physical overtraining, cognitive stress, and sleep deprivation/restriction (Figure 3A), have shown to activate the HPA axis and release its neuroendocrine factors for downstream influence on immune and autonomic activity (see Sections 3.2–3.5) by utilizing three primary structures (Figure 3B). The three structures that respond to such stressors include the paraventricular nucleus of the hypothalamus (PVN), the anterior pituitary gland, and the adrenal cortex. The PVN computes and integrates neuronal and humoral inputs to activate a specialized group of cells that control the level of activation of the HPA axis, including the regulation, synthesis, and secretion of corticotropin-releasing hormone (CRH) into the hypophyseal portal vasculature, which serves as a series of veins connecting two venous capillary beds for the transporting and exchanging of hormones between the hypothalamus and the anterior pituitary gland. The release of CRH and its subsequent binding to cognate receptors on corticotropes of the anterior pituitary can trigger the release of adrenocorticotropic hormone (ACTH) into the general circulation to bind to melanocortin-2 receptors on the surface of adrenal zona reticularis and the fasciculata cells of the adrenal gland. This binding triggers the release of glucocorticoids, including cortisol and dehydroepiandrosterone (DHEA), to the circulation (Figure 3B) (Nicolaides et al., 2015; Korte et al., 2005).

Cortisol, the primary glucocorticoid in humans (Oyola and Handa, 2017), binds to ubiquitous mineralocorticoid and glucocorticoid receptors on various body tissues to promote appropriate metabolic responses to environmental perturbations, including glycogenolysis and mobilization of free fatty acids for increased energy availability for expenditure and promote pro- and anti-inflammatory cytokine activation (Herman et al., 2012). DHEA, the glucocorticoid antagonist, serves to prevent excessive systemic inflammation and protects the neurologic machinery (i.e., glucocorticoid or mineralocorticoid receptors) from damaging effects of excess cortisol (Lennartsson et al., 2012; Lennartsson et al., 2022). Glucocorticoid receptor binding of hormones mediates an adaptive, negative feedback response inhibiting further stress hormone production at all levels of the HPA axis (McEwen and Gianaros, 2011; Herman et al., 2012) (Figure 3B).

In cohesion with the neuroendocrine response to stress, the immune system is triggered by the presence of inflammation reflected by circulating pro- and anti-inflammatory cytokine concentrations that mediate HPA axis activity via receptors in tissues associated with the axis (Kargl et al., 2024). Cytokines include mediators of the interleukin family (i.e., interleukin-6), tumor necrosis factor-alpha (TNF-α) (Turnbull and Rivier, 1995), and c-reactive protein (CRP). CRP, a systemic marker of inflammation, is closely linked to HPA activity, meaning that when CRP concentrations rise, the HPA axis also tends to activate, which can lead to increased cortisol and DHEA production (Sproston and Ashworth, 2018). When the immune system is triggered, pro-inflammatory cytokines are released, which can concomitantly stimulate the HPA axis to produce a bi-directional relationship with cortisol and DHEA. Anti-inflammatory cytokines also increase from this response as a negative feedback mechanism (i.e., interleukin-10) to counteract excessive pro-inflammatory effects (i.e., reactive oxygen species) (Kargl et al., 2024).

The physiological stress response also involves the activation of SAM axis owing to its neuronal synaptic circuitry relying on neurotransmitter communication between higher regions of the brain to peripheral receptors (Sapolsky et al., 2000). SAM activation involves the release of enzymes driven by sympathetic nervous system activity, such as salivary α-amylase (sAA), which is triggered by the release of norepinephrine and epinephrine and bind to β-adrenergic receptors in the salivary glands for its production. sAA is a reliable indicator of SAM activity (Nater and Rohleder, 2009), and has previously been used as a biomarker to assess the SAM response in ambient and military settings (Arhakis et al., 2013; Habersaat et al., 2018; Ishitobi et al., 2010; Klaus et al., 2019). The following sections aim to review empirical research of randomized controlled trial or observational cohort study designs that evaluated the influence of individual stressors commonly experienced during military training, including energy restriction or deficiency, physical overtraining or exertion, cognitive distress, and sleep deprivation or restriction, on the key drivers of allostatic load, including activity of the primary mediators of neuroendocrine, immune and autonomic systems (McEwen BS., 1998). Studies of healthy adults that assessed individual stressors and avoided confounding of additional stressors were included. Each summary included biomarkers commonly used in allostatic load research based on previous reviews (Juster et al., 2010; Carbone et al., 2022) to be consistent in the reporting of mediators across studies.

Energy restriction and/or deficits are commonly experienced during military training courses, with previous research reporting average deficits of 9.7 MJ/day over 8-days of military training (Alemany et al., 2008) and deficits ranging between 1000 and 4000 kcal d^-1 over a 61-d US Army training course (Henning et al., 2014). Energy restriction or deficits during training have been attributed to restricted feeding times (Koltun et al., 2023), food choices, and periods of practiced energy restriction to simulate operations (Friedl et al., 2000). Koltun et al. observed that negative eating behaviors, such as energy restriction and deficits, are associated with worsened military health outcomes, including MSKI risk in both sexes (Koltun et al., 2023). A synopsis of the influence of energy deficit or restriction on activity of primary mediators of allostatic load in healthy, non-obese individuals can be shown in Table 1. From this synopsis, there are two main findings from the literature that may be drawn (Alemany et al., 2008; Henning et al., 2014; Ruffing et al., 2022; Degoutte et al., 2006; Huovinen et al., 2015; Buffenstein et al., 2000; Pasiakos et al., 2011) (Table 1).

First, it is observed that neuroendocrine, autonomic, and immune biomarker concentrations depend on the relative severity (i.e., intensity) and duration of the energy restriction or deficit, such that the magnitude and length of the deficit propagates an inverted “U-shaped” curve where substantial restrictions or deficits in severity and/or prolonged duration promote acute increases in end-product concentrations followed by dampened responses with concentrations falling below baseline (Alemany et al., 2008; Henning et al., 2014; Ruffing et al., 2022; Degoutte et al., 2006; Buffenstein et al., 2000; Pasiakos et al., 2011). Energy deficits exemplified by active individuals consuming 4 MJ d^-1 (<1000 kcal d^-1) less than normal over 1 week (Degoutte et al., 2006) demonstrated significant increases in ACTH (+∼30%, p < 0.05), cortisol (+∼20%, p < 0.05) and DHEA (+5%, p < 0.05) and reductions in DHEA:Cortisol ratio (−20%, p < 0.001) from baseline (Degoutte et al., 2006). Over 2 days of near complete energy restriction where healthy individuals consumed less than 10% estimated calorie requirements, Pasiakos et al. (2011) (Pasiakos et al., 2011) observed a significant decrease in circulating cortisol (−70%, p < 0.001) and an upregulation of DHEA (+68%, p < 0.001) from baseline suggested to be driven by an inverse relationship with lower concentrations of leptin (Pasiakos et al., 2011). Likewise, Pritchard et al. (1999) (Pritchard et al., 1999) reported an 86% increase in dehydroepiandrosterone-sulfate (DHEA-S) following a 3-month energy deficit in healthy adult men (Pritchard et al., 1999). In contrast, however, Alemany et al. (2008) (Alemany et al., 2008) observed a significant reduction in DHEA concentrations following an 9.7 MJ d^-1 energy deficit over an 8-day military training exercise (Alemany et al., 2008). However, as this training included additional stressors, this finding should be interpreted with caution.

Similar dose-response relationships are revealed concerning autonomic and immune biomarkers in response to energy restriction or deprivation (Durguerian et al., 2018; Hennigar et al., 2021). Among competitive weightlifters, Durgeurian et al. observed a significant increase in resting sAA (+364.60%, p < 0.05, Cohen’s d = 1.72) following a 6-day energy restricted diet of 8.4 ± 3.6 MJ d^-1 when compared to a control group consuming 15.4 ± 5.0 MJ d^-1 (Durguerian et al., 2018). However, after a 4-week energy restriction protocol with greater absolute severity (<25% of calories from baseline; 1650 ± 911 kcal d^-1), Waldman et al. found no significant difference in resting sAA concentrations in firefighters (Waldman et al., 2020). Although Durgeurian et al. showed a conservative energy deficit (∼2007.65 kcal d^-1) than Waldman et al. (∼1650 kcal d^-1), the relative change in Durgeurian et al. (−1438 kcal d^-1) was more severe than Waldman et al. (−400 kcal d^-1), which may have contributed to the results (Table 1). Further, an acute 2-d energy restriction protocol has shown to evoke autonomic nervous system balance toward sympathetic dominance (Solianik and Sujeta, 2018), with more chronic energy deficits demonstrating a blunting in sympathetic activity measured by heart rate variability (Jenkins et al., 2022; Mazurak et al., 2011). During a 72-h sustained combat and training operations (SUSOPS) in active-duty personnel, Hennigar et al. observed a −43% energy deficit (−2047 kcal d^-1) in a restricted diet group (2515 ± 171 kcal d^-1) compared to an +18% energy deficit in a balanced diet group (5437 kcal d^-1 ± 377 kcal d^-1; p < 0.001). Further analysis^, observed significant increases following SUSOPS in CRP and IL-6 in both groups, with the restricted group showing a 59% greater increase in CRP than the balanced group (+2.6 ± 5.3 mg L^-1, p < 0.001) (Hennigar et al., 2021). However, as the SUSOPS also included heavy physical training and sleep restriction (<4 h∙night^-1), which may confound immune responses (Kargl et al., 2024), these results should be interpreted with caution. Over a 2-year energy restriction protocol, Trevizol et al. observed a significant reduction in IL-6 to indicate improvement in inflammation (Trevizol et al., 2019). Likewise, a 2020 meta-analysis demonstrated that energy-restricted diets reduce CRP concentrations compared to baseline through a considerable length of intervention (≥2 months) (Wang et al., 2020). In contrast, a 2022 systematic review observed an increase in circulating inflammatory cytokines in military personnel following field training exercises; however, this result may be confounded by additional stressors (E Silva et al., 2022).

Second, there is a general consensus that a reduction in body mass, as often observed following military training (Friedl et al., 2000; Nindl et al., 1997; Nindl et al., 2012), may serve as an indicator of increased basal HPA activity (Ruffing et al., 2022; Schorr and Miller, 2017; Villanueva et al., 1986). Ruffing et al. (2022) observed a significant correlation between reduced body mass and increased 24-h area-under-the-curve (AUC) cortisol concentrations (r = −0.473, p = 0.030), suggesting that cortisol could respond to reduced chronic energy stores (Ruffing et al., 2022). Additional studies observed increased basal cortisol concentrations among females with anorexia nervosa (Schorr and Miller, 2017; Misra and Klibanski, 2014) and in exercising females who were in a chronic energy deficit (Villanueva et al., 1986). Third, the hunger signal ghrelin may play a modulatory role in HPA axis activity during energy deficits (Misra et al., 2005). Previous studies observed a positive correlation between ghrelin and cortisol concentrations in females with anorexia nervosa (r = 0.480, p = 0.002) (Misra et al., 2005), with military-simulated energy restriction observing a negative association between satiety and cortisol (r = −0.550, p < 0.05) and DHEA-S concentrations (r = −0.620, p < 0.05) (Pasiakos et al., 2011). Hence, the augmented HPA activity during energy deficit or restriction appears to produce a dose-response relationship associated with deleterious changes in body composition and subjective satiety changes. Taken together, energy restriction or deprivation may serve as a potent driver of the primary mediators of allostatic load (Feigel et al., 2025a).

Physical overtraining–physical training conducted beyond one’s finite ability to recover adequately between training sessions and supported by reduced physical performance (Pope et al., 2018) - is commonly reported during military training (Nindl et al., 2017; Booth et al., 2006; O’Leary et al., 2018; Chicharro et al., 1998). Previous research reports that external training loads, characterized as the total work performed that contributes to the internal training load (Foster et al., 2017; Impellizzeri et al., 2019), which is defined as the relative physiologic indicator reflecting the psycho-physiological response to external loads (Foster et al., 2017; Impellizzeri et al., 2019), is often operationalized in military training by the total distance covered (O’Leary et al., 2018; Jurvelin et al., 2020; Whittle, 2022). Distances range, on average, between eight and twelve miles per day during initial training courses (Feigel et al., 2024a; Givens et al., 2023a; O’Leary et al., 2018) (Drain et al., 2015; Pihlainen et al., 2023). As a synopsis of the impact of physical overtraining, which may be experienced in military courses (Booth et al., 2006; O’Leary et al., 2018; Chicharro et al., 1998), on the primary mediators of allostatic load, four findings may be drawn (Henning et al., 2014; Roberts et al., 1993; Slivka et al., 2010; Fry et al., 1993; Volek et al., 2004; Cadegiani and Kater, 2017) (Table 2).

First, there is a general consensus that physical overtraining perturbs neuroendocrine activity such that it increases cortisol and DHEA concentrations as short as 1 week of training up to 74 days of a heavy physical training program (Roberts et al., 1993; Slivka et al., 2010; Volek et al., 2004; Cadegiani and Kater, 2017). However, this is not always observed in elite athlete populations owing to differences in physical fitness level and training experience (Slivka et al., 2010; Fernández-Garcia et al., 2002). Second, there are contradictions in results reported by individual variations in the stress response (Fry et al., 1993; Fernández-Garcia et al., 2002). For instance, the occurrence of hypocortisolism during heavy physical training (Fry et al., 1993; Fernández-Garcia et al., 2002) may be indicative of overreaching in some individuals wherein cortisol is blunted and positive adaptations cease (Armstrong et al., 2021). Third, the increase in HPA activity appears to depend on either heightened volume or intensity alone, but increased HPA activity may also be observed during high-volume, low-intensity training alone (Roberts et al., 1993; Hooper et al., 2017; O’Connor et al., 1989) or periodized high-volume, low-intensity training to low-volume, high-intensity training (Volek et al., 2004). Fourth, these observations appear independent of sex (Szivak et al., 2023a; Szivak et al., 2023b). Further, similar to the role of energy restriction (Table 1), heavy physical training demonstrates a dose-response relationship with increased relative severity or duration inducing a rise in neuroendocrine markers followed by a blunting effect (Fry et al., 1993; Fernández-Garcia et al., 2002; Bouget et al., 2006). However, as mentioned previously, differences in results may be attributed to physical fitness level (Slivka et al., 2010; Fernández-Garcia et al., 2002)

Fifth, physical overtraining has also been shown to alter inflammatory cytokine and autonomic biomarker concentrations at rest (Table 2). Fry et al. observed significant increases in markers of inflammation, including IL-2 (+183%, p < 0.001), following a 10-day, twice-daily, high-intensity interval running protocol in military personnel (Fry et al., 1993). Similarly, Tibana et al. showed significant increases in IL-6 (+99–197%), IL-10 (+14.4–21%) and a reduction in IL10:IL-6 ratio (−7.1%–8.9%) after a 2-day high-intensity functional interval training protocol (Tibana et al., 2018), and Tuan et al. observed significant increases in serum TNF-α (+∼100%, p < 0.05) following a 3-day intervention of 30-min running sessions at 85% VO2_max (Tuan et al., 2008). Previous research demonstrated that cytokines, including CRP, increases post-exercise (Tsao et al., 2009), with a peak around 24 h post-exercise (Reichel et al., 2020; Ispirlidis et al., 2008), but does not appear intensity-dependent (Tsao et al., 2009). Excessive resistance training can augment CRP concentrations in which may remain elevated up to 3 weeks (Fatouros et al., 2006). Concerning autonomic markers, Collins et al. (2019) (Collins et al., 2019) observed significant increases in sAA immediately after and 24-h post high-intensity exercise intervention in male athletes (Collins et al., 2019) and Chiodo et al. (2011) (Chiodo et al., 2011) observed significant increases following combat fighting competitions (Chiodo et al., 2011). Only one study to our knowledge investigated the chronic effects of a 10-month heavy physical training on sAA and observed a significant reduction in athletes indicating parasympathetic dominance and potential fatigue (−22%, p < 0.05) (Yasuda, 2025). Although sAA following chronic physical training remains largely uninvestigated, previous studies may support this finding by observing initial increases in sympathetic activity from heart rate variability (Plews et al., 2012) with a time-dependent reduction in sympathetic dominance toward parasympathetic dominance (Plews et al., 2012; Hynynen et al., 2006; Uusitalo et al., 1998; Flatt et al., 2017). Together, these data suggest physical overtraining may serve as a driver of the mediators of allostatic load.

Cognitive stress is the heightened perception of stress that can occur during military training (Conkright et al., 2022; Eddy et al., 2015) owing to physical training (Eddy et al., 2015), negative energy balance (Beckner et al., 2023), sleep deprivation (Passi et al., 2022), environmental conditions, and decision-making tasks (Conkright et al., 2022; Ben-Avraham et al., 2022). Newly recruited soldiers to mandatory military service face challenging psychological demands on a daily basis (Schei, 1994; Mažeikienė et al., 2021), including separation from family and friends, unpredictable and uncontrollable demands on their time, intense routines, and operating in a space laden with laws and hierarchies to contribute to cognitive stress (Boermans et al., 2013). Together, these factors can negatively affect the mental health of soldiers and increase the risk for attrition (Tait et al., 2022; Pope et al., 1999). Reduced cognitive performance may be reflected by cognitive fatigue (Eddy et al., 2015; Main et al., 2023), which can impact the ability to maintain an alert and attentive state and risk poor operational performance (Passi et al., 2022). Hockey (1997) (Hockey, 1997) observed that fatigue, in low controllability and high environmental demand situations, is related to HPA and SAM activation owing to direct engagement with an acute stressor (i.e., active problem-focused coping) (Hockey, 1997). Suarez and Perez found increased HPA activity in urban combat training when physical activity remained at a low level (Suárez and Pérez, 2013). Further analysis revealed cognitive stress caused by uncertainty regarding the location of threats, which led to fatigue and impaired post-combat information processing ability (Suárez and Pérez, 2013).

Table 3 summarizes a representative set of investigations (Henning et al., 2014; Lennartsson et al., 2012; Lennartsson et al., 2022; Budde et al., 2010; Schoofs and Wolf, 2011; Kudielka et al., 2004; Salvador et al., 2003) that assessed the influence of cognitive stress driven by the performance of cognitive tasks under time or duty constraints in occupational, classroom, or laboratory settings. Results of these studies observed significant increases (Lennartsson et al., 2012; Schoofs and Wolf, 2011; Kudielka et al., 2004; Salvador et al., 2003) or no change (Henning et al., 2014; Budde et al., 2010) in cortisol, DHEA, DHEA-S, or ACTH in response to a battery of cognitive assessments, including military training-related tasks (Conkright et al., 2022). Acute cognitive stress increases HPA activity among both sexes (Lennartsson et al., 2012; Schoofs and Wolf, 2011; Kudielka et al., 2004), whereas chronic cognitive stress can lead to blunted HPA activity (Miller et al., 2007). One meta-analysis observed a time-dependent HPA activity response with increased time since the onset leading to blunted morning and daily cortisol concentrations (Miller et al., 2007). However, it should be mentioned that Miller et al. observed that cognitive stress responses can be modulated by the nature (physical, social, traumatic), presence (morning, afternoon), emotional involvement (shame vs. loss), and controllability of the stressor (uncontrollable, controllable), which may affect neuroendocrine biomarker activity (Miller et al., 2007).

Considering the influence of cognitive stress on immune and autonomic system function (Table 3), there is evidence of elevated circulating inflammatory markers owing to recurrent daily stressors (Gouin et al., 2012; Vineetha et al., 2014). Among a group of 53 chronic caregivers (56 months ±44 months) for dementia (≥5 h wk^-1), Gouin et al. observed significantly higher concentrations of CRP and IL-6 compared to non-caregiving controls (Gouin et al., 2012). A systematic review on the influence of chronic occupational stress (i.e., employment, burnout and exhaustion, caregiver stress) identified elevated CRP concentrations than control groups (Johnson et al., 2013). However, previous research shows varied sAA responses to chronic stress (Habersaat et al., 2018; Giessing et al., 2020; Wingenfeld et al., 2010; Unno et al., 2013; Wood et al., 2021; Juster et al., 2011) despite robust increases during acute stress (Man et al., 2023; Knauft et al., 2021; Chacko et al., 2022; Teixeira et al., 2015). Individuals with chronic stress disorders demonstrate increases (Tanaka et al., 2012; Tanaka et al., 2013) or decreases in sAA (Teixeira et al., 2015; Wolf et al., 2008) during acute stressor tasks, but dampened basal sAA concentrations (Klaus et al., 2019; Wolf et al., 2008). Together, these data suggest cognitive stress may also perturb the mediators of allostatic load.

Majority of military training studies observe that soldiers achieve less than the nightly recommended sleep duration of 7–8 h per night (Givens et al., 2023a; Kargl et al., 2024; Hansen et al., 2021), with most studies observing 4–6 h of sleep per night on average (Edgar et al., 2021; Givens et al., 2023a; Kargl et al., 2024; Taylor et al., 2020). Among a representative set of investigations assessing acute (1-2 nights) partial (4 h∙night^-1) and total (0 h∙night^-1) sleep deprivation on neuroendocrine function (Table 4), there is a general consensus of increased HPA activity across studies (Guyon et al., 2014; Balbo et al., 2010). Additional studies assessing the influence of semi-chronic (4 nights) or chronic (5+ nights) find blunted HPA activity owing to “psychological deactivation” or fatigue (Åkerstedt et al., 1982) and reduced HPA sensitivity (van Dalfsen and Markus, 2018). However, chronic short sleepers (<5 h∙night^-1) have been shown to have elevated cortisol concentrations compared to normal sleepers, suggesting that the downregulation of the HPA axis may fail to occur in some individuals (Balbo et al., 2010).

Investigations on the influence of sleep disturbance, defined as interruptions during periods of sleep, and deprivation has shown to alter immune and autonomic responses (Kargl et al., 2024; Irwin et al., 2016). A 2016 systematic review and meta-analysis (N = 72 studies) on the association between sleep disturbance, sleep duration, and inflammation in adults observed that sleep disturbance was associated with higher concentrations of CRP (ES 0.12; 95% CI = 0.05–0.19) and IL-6 (ES 0.20; 95% CI = 0.08–0.31), with shorter sleep duration, but not the extremity of short sleep, was associated with higher CRP (ES 0.09; 95% CI = 0.01–0.17) but not IL-6 (ES 0.03; 95% CI: −0.09–0.14). However, neither sleep disturbances nor sleep duration was associated with TNF-α (Irwin et al., 2016). Among salivary markers, recent literature suggests that sAA may be a cross-species marker of sleep deprivation in tactical (i.e., first responder) and military populations (Lindsey et al., 2025) and individuals with sleep disorders (Thieux et al., 2024). Pajcin et al. observed the influence of 2-d TSD (50-h) on sAA and observed a significant diurnal profile wherein concentrations increased throughout the morning and afternoon (p < 0.001) and steadily declined in the evening and early-morning (p < 0.001). These results suggested that sAA may be sensitive to the diurnal rhythm for arousal and tracking the diurnal drive for alertness (Pajcin et al., 2017) (Table 4). However, there does not appear to be a consensus as to the direction of sAA with subjective feelings of sleep disturbance and repercussions of sleep debt (i.e., sleepiness and cognitive performance) (Thieux et al., 2024).

Although military training-related stressors do not occur in isolation during training (Nindl et al., 2018; Friedl et al., 2000), and exhibit bi-directional effects, such as sleep deprivation affecting subjective feelings of psychological stress (Schwarz et al., 2018), the above findings may reveal a pattern of the influence of stressors on primary mediators of allostatic load by demonstrating acute increases and chronic reductions in biomarker concentrations during stress (Tables 1–4). Chronic stress can lead to one or more forms of HPA axis dysfunction (Figures 3C,D) and alter immune and autonomic system function, which may serve as a mechanism of allostatic load (Selye, 1950; van Dalfsen and Markus, 2018; Karin et al., 2020) as biomarker concentrations may fall within ‘at-risk’ bounds to support the computation of the ALI (Juster et al., 2010).

HPA axis dysfunction is characterized as a dynamic compensatory response to chronic stress that begins with initial hypercortisolism followed by hypocortisolism (Fries et al., 2005) or diurnal dysrhythmia (McEwen, 2007). The mechanism of HPA dysregulation derives from an evoked HPA axis that engenders an over-responsive system (e.g., hypercortisolism) toward an under-responsive or non-responsive system (e.g., hypocortisolism) (Selye, 1950). Progression from an over to under-responsive system is reflected by one or more forms of HPA axis dysregulation leading to allostatic load (McEwen, 2007; Juster et al., 2011) (Figures 3C,D). Model 1 is one form that illustrates the dynamics of neuroendocrine biomarker responses broken down into three stages when under the influence of chronic stress. These three stages illustrate the trajectory of biomarker concentrations and emulate Han’s Selye General Adaptation Syndrome (GAS) theory (Selye, 1950) (Figure 3C). The first stage of biomarker activity is the “Alarm Stage”, described as an acute, adaptive response to a stressor. This can also be illustrated by the allostasis model (Figure 1), where a response appropriately meets a demand observed by increased cortisol, DHEA, and ACTH concentrations (Korte et al., 2005). If the stress continues, however, the ‘Resistance Stage’ occurs, which is characterized by chronic activation of the physiological stress response (“chronic allostasis”) as observed by heightened biomarker concentrations. Notably, this stage risks degradation of protective negative feedback mechanisms (i.e., deterioration of glucocorticoid receptor sensitivity) or primary outcomes, as shown in the allostatic load model (Figure 2A). Finally, if the stress continues, the “Exhaustion Stage” occurs, which is characterized by a reduction in circulating biomarker concentrations that can signal allostatic load (Sher et al., 2020). Taken together, Model 1 reports that under chronic stress, ACTH will drive cortisol production and deplete DHEA production that leads to reduced cortisol production toward hypocortisolism in the Exhaustion Stage (Stephens and Wand, 2012). Hence, this model benefits the prediction of neuroendocrine biomarker trajectories and downstream immune and autonomic biomarker trajectories to complement the aforementioned results of empirical research on common military training-related stressors (Tables 1–4). Hence, Model 1 may serve as one mechanism leading to allostatic load (McEwen BS., 1998; Sher et al., 2020).

A second mechanism of allostatic load is shown in Model 2 (Figure 3D). Model 2 purports that HPA dysfunction and the downstream effects of immune and autonomic function arise from changes in the total functional masses of the HPA hormone-secreting glands including the adrenal cortex and anterior pituitary (Karin et al., 2020). Karin et al. reported that this mechanism occurs from an initial hypercortisolism followed by reduced DHEA production and glucocorticoid receptor and mineralocorticoid receptor sensitivity of the cells in the hypothalamus and anterior pituitary that impair the negative feedback loop of the HPA axis (Karin et al., 2020). Reduced cell receptor sensitivity with stress hormone production may render the hormones of the HPA as growth factors for the glands in the axis where excessive secretion of CRH can drive pituitary corticotroph cell growth and ACTH can drive adrenal gland hypertrophy/hyperplasia to result in larger functional masses capable of greater binding affinity to maintain the stress response over weeks. Consequently, this form of HPA axis dysfunction can lead to a similar over-to-under responsive neuroendocrine, immune, and autonomic nervous system activity (Henning et al., 2014; Kargl et al., 2024; Karin et al., 2020) as observed in response to military training-related stressors (Tables 1–4). Hence, Model 2 may serve as a second mechanism of allostatic load. Owing to the role of military training-related stressors on allostatic load, the next section summarizes the impact of allostatic load quantified by ALI on physical performance and psychological and musculoskeletal health, in non-obese, healthy adults and, where available, military personnel.

Previous evidence observes that ALI is negatively associated with physical performance outcomes assessed by a battery of maximal strength and balance assessments (Germano et al., 2023; Hansen et al., 2016). Among 1101 healthy volunteers (65–74 years), Germano et al. observed ALI was inversely associated with score on the Short Physical Performance Battery (SPPB) score (β = −0.234, p < 0.001), with indirect effects evidenced between age and socioeconomic status (Germano et al., 2023). Similarly, among 5467 healthy volunteers (48–62 years), Hansen et al. found that the ALI mediated the association between education and physical performance (chair rise ability, postural balance, sagittal flexibility) and muscle strength (jump height, trunk extension, and flexion, handgrip strength) and accounted for 2%–30% of the total effect among women (Hansen et al., 2016). As a secondary analysis of data from the MacArthur Studies of Successful Aging study, Seeman et al. observed step-wise reductions in similar physical performance outcomes with every 1-unit increase in ALI (Seeman et al., 1997) and during a follow-up of 7 years (Seeman et al., 2001). Seeman et al. also found the ALI outperformed predicting physical dysfunction to a greater degree than its individual sub-components, suggesting its benefit of determining risk of physical performance decline (Seeman et al., 2001). However, these investigations were conducted among older (mid to late-life) adult populations (Germano et al., 2023; Hansen et al., 2016). Nevertheless, recent research has linked high ALI with worsened physical performance in younger (<40 years) populations and in military personnel (Feigel et al., 2025a; Hastings et al., 2022). Feigel et al. observed a significant negative association between change (Δ) in ALI from baseline and change in physical performance in men from elements of the USMC Physical Fitness Test (PFT) (ΔPullups: β = −0.88, R² = 0.60, 95% CI: −1.55, −0.21; ΔPush-Pull PFT Score: β = −2.87, R² = 0.60, 95% CI: 4.99, −0.75; Δ Total PFT Score: β = −3.48, R² = 0.58, 95% CI: −5.76, −1.19) to suggest a potential role of chronic stress on military physical performance (Feigel et al., 2025a). Data from the National Survey of Midlife Development in the United States (N = 2055, 26–86 years) demonstrate a negative association with ALI and physical function (grip strength: β = −0.11, 95% CI: −0.15, −0.07; gait speed: β = −0.20, 95% CI: −0.24, −0.16) (Hastings et al., 2022). Although military training studies lack use of the ALI, previous research demonstrates an association between altered neuroendocrine and autonomic hormone profiles following training and worsened physical fitness characteristics, such as muscular power and strength (Szivak et al., 2018), which are important attributes for occupational task performance (Feigel et al., 2024b). Further evidence suggests that increased inflammatory cytokine concentrations can hinder muscle protein synthesis of lean muscle mass (Miller et al., 2022) and negatively influence upper and lower-body muscular strength (Sharma Ghimire et al., 2023). Hence, further research of ALI on physical performance in in-training personnel is warranted.

Previous epidemiological studies demonstrate a positive association between ALI score and symptoms of reduced psychological wellbeing in healthy individuals (Guidi et al., 2020), including anxiety (Gou et al., 2025; D’Alessio et al., 2020), depression (Gou et al., 2025; D’Alessio et al., 2020), and perceived stress (Guidi et al., 2020; Beckie et al., 2016; Juster et al., 2010; Geronimus et al., 2006) and a negative association with resilience (Felix et al., 2023), all of which are reported during military training (Forse et al., 2024; Bulmer et al., 2022b; Taylor et al., 2009; Guo et al., 2021). However, cognitive reappraisal was indirectly associated with lower ALI, whereas the tendency to use emotion suppression was indirectly associated with greater ALI (Ellis et al., 2019). Sleep health has been tied to allostatic load from a systematic review demonstrating a positive relationship between chronic sleep difficulty level and ALI (Christensen et al., 2022). Guidi et al. observed poorer objective and subjective sleep quality was associated with ALI in four separate investigations (Guidi et al., 2020). A 2022 systematic review and meta-analysis by Christensen et al. observed a significant negative association between sleep health, characterized by sleep duration and sleep quality (i.e., greater time in restorative sleep stages as compensation) on ALI among epidemiological studies. However, it was also reported that sleep may be bi-directional where poorer sleep quality may contribute to ALI if sleep health is not improved with intervention (Christensen et al., 2022). Interestingly, recent work from our group observed a significant negative association between Δ in sleeping difficulty level and Δ ALI by the end of a 10-week military training course in both sexes (ΔSD: β = −1.25 to −0.56, R² = 0.35 to 0.82, p < 0.001–0.046) (Feigel et al., 2025a). Further research into the influence of ALI on psychological wellbeing outcomes in military training populations is also warranted.

Gallagher and Barbe (2022) examined the role of allostatic load on musculoskeletal disorders, characterized as injury, dysfunction, or impairment to the muscle, ligament, tendon, or bone, during occupational settings (Gallagher and Barbe, 2022). Their findings suggested that musculoskeletal disorders, including overuse MSKIs, may result from impaired tissue repair mechanisms driven by mechanical stress. These impairments, in turn, are influenced by underlying inflammatory, autonomic, and neuroendocrine dysfunctions associated with allostatic load, which is precipitated by chronic psychological and physical stress. Overuse MSKIs are comprised of inflammatory and degenerative conditions in musculoskeletal tissues involving muscles, tendons, ligaments, and peripheral nerves (Edwards, 2018). The authors found that physical work risk factors for overuse MSKIs included high force demands, repetitive work, adoption of non-neutral postures, and repeated heavy lifting, with a combination of high psychosocial work demands during work settings (da Costa and Vieira, 2010), which are common attributes experienced during military training courses (Vaara et al., 2022). Additional psychological risk factors included perceived stress at work, psychological job demands, and low job control (Bongers et al., 1993; Bongers et al., 1993; Deeney and O’Sullivan, 2009). Together, the authors purported that the presence of psychological and physical stress may lead to allostatic load and overuse MSKI owing to a slower-than-normal healing response in the tissue. This may result in faster damage development in musculoskeletal tissues and higher overuse MSKI risk (Gallagher and Barbe, 2022).

Epidemiological studies demonstrate relationships between ALI and musculoskeletal disorders (Guidi et al., 2020; Mori et al., 2014). Among 703 healthy men and women (median age: 56), mixed-effects linear regression controlling for clustering within families and adjusted for age, gender, race/ethnicity, body mass index, menopausal transition stage, childhood socioeconomic status, adult finances, education level, and study center, each standard deviation increment in ALI was associated with between 0.10 and 0.11 standard deviation decrements in lumbar spine bone mineral density (all p < 0.05) (Mori et al., 2014). Symptom frequency and intensity were associated with higher ALI among chronic fatigue syndrome patients compared to controls (Maloney et al., 2006; Goertzel et al., 2006). Further research on the influence of ALI on musculoskeletal health on training personnel is warranted (Feigel et al., 2024a).