Adaptation Due to Repeated Cold Exposure

Various occupations, sports and adventures are associated with seasonal and repeated cold exposures over long time periods (months to years). Examples are pearl divers in Korea, winter swimmers, participants in polar and high-altitude expeditions, skiing instructors, mountain guides, and downhill and nordic ski racers. It is difficult to distinguish whether adaptive effects arose from stimuli like cold, altitude, or high-intensity exercise. Winter swimmers (Vybíral et al., 2000) and female pearl divers from Korea (Tipton and Bradford, 2014), who are primarily exposed to cold water, may exhibit two or all three forms of cold adaptation (insulative, hypothermic, and metabolic) and, additionally, local cold adaptation. In winter swimmers shivering occurred considerably later during cooling when compared to not cold-adapted subjects, which was explained by reduced heat loss, more pronounced non-shivering thermogenesis, but less total heat production (Vybíral et al., 2000). Insulative adaptation was demonstrated in the cold-adapted during swimming, and hypothermic adaptation occurred when only sitting in cold water (Tipton and Bradford, 2014). Improved cold tolerance has also been shown in pearl divers daily immersed in cold water for several hours (Hong, 1973). Besides insulative, hypothermic and metabolic forms of adaptation those women also exhibited reduced heat flux in the limbs indicating local cold adaptation. Interestingly, however, pearl divers lose their cold adaptation within 3–5 years when wearing wet suits (Park et al., 1983).

Acclimation and Habituation to Cold Environments

In contrast to the excellent capabilities of humans for heat acclimatization, potential acclimatization to cold remains a matter of discussion (Brazaitis et al., 2014a). Acclimation studies are most appropriate to investigate type and time course of cold responses under standardized conditions and in individuals/populations of interest. Most studies have been performed by cold water immersion due to the high conductivity and thermal capacity of water compared to air. Cold water immersion elicits typically a cold shock response characterized by sympathetic activation, hyperventilation and tachycardia (Castellani and Tipton, 2015). However, repeated immersions in cold water (habituation) considerably decreases these responses, and this depression is lasting for 7–14 months (Tipton et al., 2000) (Table ​Table11). Biological mechanisms involved in habituation of the cold shock response are still unclear. Mental processes seem just as involved as cold stimuli per se and the frontal cortex area seems to play a pivotal role (Glaser and Griffin, 1962; Castellani and Tipton, 2015). Even a small number of repeated short cold exposures (e.g., 3 min × 60 min) resulted in less discomfort, delayed, and reduced intensity of shivering associated with lower skin temperature, which is indicative for insulative adjustment (Makinen, 2010). Longer durations (90 min – 3 h) of cold water immersion (10–18°C) resulted in insulative or combined metabolic insulative acclimation (Young et al., 1986). It has been suggested that skin cooling is sufficient to provoke the vasoconstrictor response, whereas a decrease (0.8°C) in core temperature seems to be necessary to induce sympathetic activation (Makinen, 2010).

Types of acclimatization/acclimation to cold air depend mainly on the duration of exposure. Whereas exposures for up to about 1 h cause habituation with a later onset of shivering and maintenance of skin and core temperatures (Hesslink et al., 1992; Makinen, 2010), longer exposures (3 h to days) result in hypothermic habituation (Mathew et al., 1981). A few studies suggest that chronic cold exposure (weeks), e.g., during camping in remote areas without sufficient cold protection, triggers metabolic acclimatization (Scholander et al., 1958a).

Studies have typically been performed under resting conditions. It is remarkable that exercise during the cold exposure seems to blunt some of the acclimation effects, e.g., the hypothermic acclimation observed during rest (Launay et al., 2002). Another study demonstrated increased susceptibility to hypothermia after multiple days of severe exertion in cold environment (Castellani et al., 2001). This topic has not been investigated extensively and urgently needs further attention.

Physiological and Pathophysiological Responses to Acute Cold Exposure

Acute cold exposure causes powerful autonomic homoeostatic responses in order to prevent heat loss and to maintain core body temperature. Cutaneous vasoconstriction and thermogenesis (Table ​Table11), primarily by skeletal muscle contractions, are most important and are particularly powerful when cooling involves both superficial and deep thermoreceptors (Stocks et al., 2004). Thermogenesis may be achieved by voluntary behavioral responses, e.g., physical activity, but occurs involuntarily by shivering. The magnitude of the effector responses to cold are varying depending on the severity and duration of exposure, physical activity and clothing, and individual characteristics such as age, gender, and body composition (Stocks et al., 2004). When these responses are not sufficient to maintain core temperature, the risk of hypothermia will occur if not protected by appropriate clothing (Burtscher et al., 2012). Moreover, peripheral vasoconstriction may also enhance the risk of cold injury and impair manual dexterity. The occurrence of freezing and non-freezing cold injuries depends on the severity and duration of cold exposure (Vale et al., 2017) and may predispose for future cold injury (Gorjanc et al., 2018). Systemic blood pressure (systolic and diastolic) typically increases, and heart rate may increase, e.g., when only legs are cold exposed, or decrease with whole body exposure, again depending on the severity of cold (Korhonen, 2006). The augmentation of the double product (systolic blood pressure × heart rate) indicate elevated myocardial oxygen consumption. This fact may in part explain the cold-induced provocation of angina attacks in patients suffering from coronary artery disease, especially in those with an abnormal baroreceptor function (Marchant et al., 1994). In addition, elevated blood pressure and centralization of blood might cause an increase of diuresis (cold-induced pressure diuresis). The primary effect of cold air on the respiratory system is to decrease minute ventilation and chemosensitivity which is not really efficient in humans. More importantly, however, bronchoconstriction, airway congestion, secretion and reduced mucociliary clearance may compromise pulmonary mechanics in subjects with cold- or exercise induced asthma (Giesbrecht, 1995). In addition, acute cold stress was shown to impair vigilance, overall mood, motor, and cognitive performance (Brazaitis et al., 2014b). However, most of these responses to acute cold exposure diminish or even disappear with habituation or acclimation.

Cellular Responses to Cold Exposure

Prevention of cooling can be achieved by increasing heat production by excessive breakdown of stored nutrients such as fat, which is called non-shivering thermogenesis. Prolonged cold exposure results in an increase in PGC1α protein expression, which in turn stimulates mitochondrial biogenesis, e.g., in skeletal muscle (Chung et al., 2017). In white and in brown adipose tissue the cold-sensing receptor TRPM8 seems to initiate increased mitochondrial thermogenesis (Frontini and Cinti, 2010; Rossato et al., 2014) by up-regulating mitochondrial UCPs (e.g., UCP1). In brown adipose tissue UCP1 uncouples mitochondrial oxygen consumption from oxidative phosphorylation resulting in heat production, which is an important means to regulate body temperature in cold ambient temperatures, in particular in small mammals and human infants (Cannon and Nedergaard, 2004). Such mechanism also occurs in beige adipose tissue (Shabalina et al., 2013), whose formation is stimulated in the cold (Fisher et al., 2012) or by caloric restriction (Fabbiano et al., 2016). Notably, acute exposures (3 h) of healthy humans to cold temperatures (7°C) seems to not elicit changes in skeletal muscle gene expression (Zak et al., 2017). Therefore, it has been discussed that exercise might be a necessary accompanying condition to induce expression of respective genes when acutely exposed to extreme environmental temperatures.

Ventilation and Alveolar Gas Exchange

All HA natives hyperventilate relative to lowlanders at low altitude (Table ​Table22). However, the degree of hyperventilation to a defined hypoxic stimulus (hypoxic ventilatory response, HVR) varies considerably among different ethnic groups. Tibetans seem to have an elevated HVR (Curran et al., 1995), whereas it seems blunted in Andeans (Beall et al., 1997) and in Sherpas (Santolaya et al., 1989) implying differences in the sensitivity of the peripheral chemoreceptor (Moore, 2000).

Because in HA natives hyperventilation begins with their first breath and is maintained throughout their life, it is likely the main mechanism causing extended growth of the lungs resulting in greater forced vital capacity and increased lung volumes and diffusion capacity in HA natives compared with Han Chinese (East Asian ethnic group) living in the Tibet and with sojourners to the Andes, who moved to HA as adults and lived there for several years (Jones et al., 1992; Brutsaert et al., 2004). Similarly, larger lung volumes and increased diffusion capacity has been found in rodents and dogs raised in hypoxia. Animals raised in normoxia that moved to hypoxia for several years did not acquire larger lungs (Brutsaert, 2016). It has to be noted, however, that also lifestyle favors the development of larger lungs in HA natives, because many study subjects were physically much more active than their low-land-counterparts (Kashiwazaki et al., 1995).

Hypoxia-induced vasoconstriction of small pulmonary arteries (Euler-Liljestrand reflex) seems to be blunted in Tibetans (Groves et al., 1993), which has been discussed to protect from augmented fluid filtration into the alveoli and thus from interstitial and alveolar edema, and to favor alveolar oxygen diffusion in HA natives relative to lowlanders with high pulmonary capillary pressure at HA. In fact, a decreased alveolar-to-arterial oxygen difference has been found in HA natives (Lundby et al., 2004), but results are conflicting.

Together, these results indicate phenotypic adaptation with improved alveolar oxygen diffusion and higher arterial oxygen saturation (SaO₂) at rest and during exercise (Wagner et al., 2002; Lundby et al., 2004; Brutsaert, 2016).

Oxygen Transport by Hemoglobin

An increased oxygen affinity of hemoglobin (Hb-O₂-affinity) in hypoxia is favorable, because it results in higher SaO₂ at a low PO₂ than low oxygen affinity (Mairbäurl and Weber, 2012). It is unclear, however, whether adjustments at this level occur in HA natives. No differences in Hb-O₂-affinity have been found between sojourners and Andeans (Lundby et al., 2006), and between sojourners and Sherpas (Samaja et al., 1979). However, increased Hb-O₂-affinity in HA natives has also been reported (Balaban et al., 2013; Simonson et al., 2014). This is in line with increased Hb-O₂-affinity in animals native to HA relative to their lowland counterparts (Petschow et al., 1977).

Amount of Hemoglobin

An increased hemoglobin concentration ([Hb]) is of advantage in hypoxia because it increases the oxygen content of arterial blood to compensate for decreased SaO₂. Most HA natives show an increased [Hb] in blood (Table ​Table22). The degree of elevation directly relates to the absolute altitude. However, [Hb] varies considerably among different ethnical groups living at comparable altitude. Most data were obtained in the Andes and on Han-Chinese living on the Tibetan plateau, where a pronounced increase in [Hb] has been found (e.g., Hurtado et al., 1945; Cosio and Yataco, 1968; Winslow and Monge, 1987; Wu et al., 2014). However, similar to the animal kingdom, where yaks and alpacas have lower [Hb] than their lowland-relatives, cows and llamas, respectively, when exposed to HA (Bouverot, 1985), also some human populations living at HA have only mildly elevated [Hb]. Those are Tibetans (Simonson et al., 2010) and Ethiopians (Beall et al., 2002), whose [Hb] is a few g/dl lower than that of Andeans and Han Chinese (Beall et al., 1998), and their [Hb] actually is close to normal sea-level-values despite living at altitudes >3500 m. Differences exist even among Andean populations, where Quechuas have lower [Hb] than Aymaras (Arnaud et al., 1981). In any of the aforementioned populations the altitude-related increase in [Hb] exist both in men and in women (Wu et al., 2005; Gassmann et al., unpublished data).

Though most studies have been performed on adult individuals, a few studies indicate that [Hb] is already elevated during childhood in Han-Chinese (Wu et al., 2005) and in Quechuas (Garruto et al., 2003) living at altitudes between 4200 and 4500 m, whereas [Hb] of Tibetan children living at HA is only slightly higher (Wu et al., 2005) than that of lowland-children (Fulwood et al., 1982) of ages between 6 months and 15 years, both in girls and in boys. Data on newborns are sparse (Moore et al., 1982; Niermeyer et al., 1995), but [Hb] appears to be within the range of lowland-newborns, with a tendency of lower [Hb] in Tibetan than in Han Chinese newborns in Lhasa (Niermeyer et al., 1995).

The mechanisms explaining the different patterns in [Hb] in HA natives of different ethnicity are not fully understood. The lower [Hb] in Tibetans than in Han Chinese is associated with putatively advantageous polymorphisms of EPAS1 (HIF-2α) (Beall et al., 2010), the prolyl hydroxylase EGLN1 and PPARα (Simonson et al., 2010) in Tibetans, as well as of the thyroid hormone receptor THRB and, marginally, also of EPAS1 in Ethiopians (Scheinfeldt et al., 2012), which are all associated with low [Hb] thus indicating genetic adaptation.

There is a clear increase in total hemoglobin mass (tHb) with decreasing SaO₂ in North Americans residing at different altitudes (Weil et al., 1968). However, a significant increase in tHb was only found when PaO₂ dropped below ~70 mmHg. This is consistent with data showing that Bolivians living at 2600 m did not have an elevated tHb (Schmidt et al., 2002; Böning et al., 2004) indicating that the stimulus of chronic hypoxia at moderate altitudes was not strong enough to induce polycythemia. In contrast, tHb was elevated by ~15% in La Paz (3600 m) (Wachsmuth et al., 2013), and by up to 80% in residents of Cerro de Pasco (4330 m) (Sanchez et al., 1970). However, excessive high [Hb] is considered maladaptive and represents a key factor of chronic mountain sickness (CMS) first described by Carlos Monge in the Peruvian Andes (Vargas and Spielvogel, 2006).

Alveolar Ventilation and Oxygen Diffusion

Decreased oxygen content in arterial blood is detected by carotid bodies, which generate signals stimulating alveolar ventilation. Hyperventilation at HA slightly raises the alveolar partial pressure of oxygen (PAO₂) at the expense of decreased arterial partial pressure of carbon dioxide (PCO₂) and respiratory alkalosis, and slightly elevates SaO₂ as long as alveolar oxygen diffusion is not impaired. However, the degree of hyperventilation to a defined hypoxic stimulus varies considerably among individuals, where those with the lowest HVR are the least hypoxia-tolerant ones indicated by increased susceptibility to HA pulmonary edema (HAPE) (Hohenhaus et al., 1995). Sojourners to HA also show pronounced hypoxic pulmonary vasoconstriction (Sylvester et al., 2012) (Table ​Table22), which favors alveolar filtration and subclinical edema, and which, if exaggerated vasoconstriction occurs, might result in HAPE with massively impaired trans-alveolar oxygen diffusion (Swenson et al., 2002). Even after long-term sojourns lowlanders did not acquire the improved gas exchange observed in HA-natives (Dempsey et al., 1971).

Cardiovascular Responses

Due to the decreased arterial oxygen content cardiac output increases in order to maintain oxygen delivery to the periphery. At rest, the increase is mainly due to an increased heart rate, while the stroke volume is decreased, likely due to a decreased venous return because of peripheral vasodilation (Alexander and Grover, 1983). However, with prolonged stay at HA, cardiac output decreases and might even fall below low altitude values. Although cardiac output for the same work load is higher at HA than at low altitude, stroke volume remains decreased (Alexander et al., 1967). Maximal heart rate is decreased because of down-regulation of adrenergic β-receptors (Kacimi et al., 1992). In general, it has been assumed that cardiac function is preserved in sojourners to HA (Reeves et al., 1987).

Oxygen Affinity of Hemoglobin

Hyperventilation decreases arterial CO₂ and results in respiratory alkalosis. Together, both effects increase the Hb-O₂-affinity which improves oxygen binding by hemoglobin even at low arterial PO₂ (Mairbäurl and Weber, 2012). Furthermore, elevated levels of organic phosphates, such as 2,3-DPG, increase the Bohr effect on Hb-O₂-binding which favors the release of oxygen from its Hb-bond in peripheral tissues (Mairbäurl and Weber, 2012).