INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

During acute hypoxia, a reduction in oxygen consumption (O₂) has been observed in many mammalian species (1). This occurs particularly during exposure to cold and in species of small or medium size, which have high thermogenic requirements. In fact, the primary cause for the reduction in O₂ is the hypoxic inhibition of both shivering and nonshivering thermogenesis, a phenomenon that can be accompanied by a reduction in the set point of thermoregulation (1, 2).

In the newborn as in the adult, O₂ is very sensitive to hypoxia, and a reduction in O₂ has been observed in many newborn species, including the human infant, even at moderate degrees of hypoxia (3). In the newborn, the thermogenic processes are almost exclusively dependent on nonshivering thermogenesis, via mechanisms that involve the uncoupling of oxidative phosphorylation by the protein thermogenin in the mitochondria of the brown fat (6).

Whether or not chronic hypoxia during gestation affects the newborn infant's thermogenic capacity is unknown. Experiments with neonatal rats have indicated that the expression of thermogenin is reduced with prolonged hypoxia either during gestation or in the early postnatal period (7), similarly to what is observed in older rats exposed to prolonged hypoxia (8). In a study of 10 infants with arterial oxygen desaturation because of cyanotic heart disease, the thermogenic response was found to be present (9). However, in this latter study the wide range in the subjects' ages (8 d to 7 mo), the large variability in the degree of hypoxemia, and the presence of heart failure in some patients complicate the interpretation of the results.

In the present study we addressed the question of thermogenesis in the hypoxic newborn by studying changes in O₂ in a population of normal, full-term infants born at high-altitude in the Andes during a moderate reduction in ambient temperature. Their response is compared with that of infants of the same ethnic origin born at sea level.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Measurements were made in Peru on two groups, each comprising 20 1-d-old infants, after approval by the ethics committee of the Universitad Peruana Cayetano Heredia. For each neonate, verbal informed consent was obtained from at least one parent. Both groups of infants included Amerindians and mestizos (of probable partial European ancestry). One group was studied at sea level in Lima (50 m altitude, with an average barometric pressure of 760 mm Hg and an inspired oxygen partial pressure of 159 mm Hg, equivalent to 20.93% O₂). The other group was studied in Cerro de Pasco, at an altitude of 4,330 m (average barometric pressure of 450 mm Hg and an inspired oxygen partial pressure of 94 mm Hg, equivalent to 12.42% O₂ at sea level). All of the infants' mothers had lived and had their pregnancies in the corresponding city or immediate surroundings of equivalent altitude. All infants were born at term and appeared healthy at clinical examination. The characteristics of the infants examined in the study are presented in Table 1.

Measurements

A stainless steel cylinder (I.D. = 270 mm, length = 700 mm) served as a humicrib. A submersible pump (3,000 L/h; Cole-Parmer, Niles, IL) placed in an appropriate water bath maintained the desired temperature by circulating water through coils of polyvinyl chloride (PVC) tubing (I.D. = 15 mm) that surrounded the humicrib. The bath was set to provide a humicib internal wall temperature of either 35° C (warm, at the upper end of thermoneutrality [10, 11]) or 26° C (cool). Wall temperature was monitored with two thermocouples attached midway along the ventral and dorsal surfaces of the humicrib. The temperature difference between the two thermocouples did not exceed 2° C. Relative humidity within the humicrib was occasionally checked (Airguide Instrument, Chicago, IL), and was found to range between 50% and 60% at both of the altitudes at which the study was conducted.

The humicrib was sealed at one end except for a gas outlet that connected directly to a smaller plastic cylinder (I.D. = 150 mm, length = 150 mm) situated within the humicrib. An infant placed within the humicrib was positioned in such a way that its head and upper torso were within the smaller chamber, which in turn was partly sealed with a transparent plastic curtain that formed a loose seal with the infant's body. A regulated flow of air (4,640 ml/min, STPD) was drawn via the gas outlet through the smaller chamber by a pump (Model 506; Reciprotor, Skara, Sweden) situated downstream and regulated by a mass-flow controller (Side-Trak 840; Sierra Instruments, Monterey, CA).

The air exiting the chamber was subsampled (with a small pump producing a flow  200 ml/min), passed through a drying column (Drierite; Hammond Drierite Co., Xenia, OH), and analyzed for fractional concentrations of oxygen and carbon dioxide with CO₂ and O₂ analyzers arranged in series (OM-11, Beckman Instruments, Anaheim, CA; and CD-3A, Applied Electrochemistry, Pittsburgh, PA, respectively). Baseline fractional values could be checked at any time by subsampling room air and without disturbing the flow through the chamber. Both analyzers had their spans calibrated daily, the O₂ analyzer with ambient air and the CO₂ analyzer with a 10% CO₂ standard gas. Because both analyzers sensed partial pressure, appropriate adjustment was made for differences between the two study localities in barometric pressure. The signal from each analyzer was recorded on chart paper (NGI 102; Servogor, Neudorf, Austria) at a speed of 5 mm/ min and with a fractional resolution of 0.0001 per division.

System integrity (e.g., absence of leaks or mixing problems) was verified by bleeding a known flow (STPD) of CO₂ (determined volumetrically) into the inner chamber and, from the deflection on the O₂ analyzer, determining the actual flow through the system through a modification of a previously reported technique (12). The flow determined in this way was within 2% of the flow delivered by the pump, implying that the system was leak-free and that the gases used were well mixed.

Protocol

At the time of investigation, each infant, breast fed and quiet, and wearing only a disposable napkin, had its body temperature (Tb) measured with a thermocouple inserted 30 mm into the rectum. Two electrodes for heart rate monitoring (Polar PE 4000, with telemetric transmission to a Mini-logger 2000; Mini-mitter, Sunriver, OR) and three skin thermocouples (intrascapular, Tscap; abdomen, Tabd; and calf, Tcalf) were positioned. The infant was then placed in a supine position on a thin mattress (10 mm thick, cloth covered) and transferred to the humicrib, which was preset to 35° C. Recording of O₂ and CO₂ was begun immediately, and heart rate (HR, calculated from a running average of four consecutive interbeat frequencies and read from a display unit) together with the three skin temperatures and two humicrib temperatures was noted every 4 min (digital thermometer with a six-channel multiplexer, Model 871A; Omega, Stamford, CT). After 20 min the temperature of the humicrib was lowered to 26° C by transferring the pump to the cool-water bath. Recording continued for a further 20 min, with the humicrib equlibrating to the new temperature in about 10 min (Figure 1). The infant was continually observed throughout the experiment, and in most cases remained quiet (not necessarily sleeping) and without gross body movements. Mild disturbances, particularly during cooling, were considered part of the normal response; however, if the infant became distressed, as noted by loud crying, measurements were interrupted and the experiment discarded. This was the case for seven infants in Lima and four infants in Cerro de Pasco. The data presented here refer to 20 infants at each altitude on which the complete experiment was conducted. At the end of the experiment the infant was removed from the humicrib and Tb was measured again.

View larger version (14K): [in this window] [in a new window]   Figure 1.   Time profile of ambient temperatures within the humicrib. The epochs used for calculation of values under warm and cool conditions are indicated. Values are means ± SE.

Data Analysis

Over the last 8 min of the warm and cool periods (Figure 1), oxygen consumption (O₂) and carbon dioxide production (CO₂) were determined from the area between the baseline and the O₂ and CO₂ tracings with the help of a graphics tablet connected to a minicomputer. The area divided by the recording time represents the average deflection from baseline that occurred during this period. This value, corrected for calibration, was multiplied by the flow and normalized by the weight of the infant to yield the average O₂/kg (STPD) and CO₂/kg (STPD) during the 8-min epochs in the warm and cool conditions. The values presented for HR, Tscap, Tabd, and Tcalf were obtained by averaging the three readings encompassed by the 8-min epochs (i.e., minutes 12, 16, and 20) for both warm and cool exposures.

Statistical analysis was done through repeated measures analysis of variance (ANOVA), with one grouping factor (the two locations) followed by a priori, post hoc contrasts (between values under warm and cool conditions at the same altitude, and between the corresponding temperature at 50 m and 4,330 m), with Bonferroni's limitation of the modified two-tailed t test. A statistically significant difference was defined as one with a value of p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The two groups differed in their birth weights, with the infants born at high altitude being approximately 500 g lower in weight than those born at sea level, and the metabolic data presented are therefore normalized by body weight. Under warm conditions the infants born at high altitude had the same O₂/kg, CO₂/kg, and HR values as those born at low altitude (Table 2, Figure 2). Body temperature and the various skin temperatures were all lower in the infants at Cerro de Pasco than in their counterparts at low altitude, by an average of about 0.5 to 1.0° C. 

View larger version (32K): [in this window] [in a new window]   Figure 2.   O2 and HR under warm and cool conditions for infants born at 50 m and 4,330 m. *Significant difference from other values. Dashed line is the average of the combined warm values for each altitude. Values are means ± SE.

Under exposure to a cool temperature, the O₂/kg and CO₂/kg of infants in Lima increased by 34% and 42%, respectively, whereas the infants born at high altitude showed no increase in metabolic rate. The HR response followed a similar pattern (i.e., HR increased with cool temperature only in the infants born at low altitude) (Table 2, Figure 2).

Exposure to cool temperature did not alter Tb or the various skin temperatures in either subpopulation of infants. Under both the cold and the warm conditions, these values were lower in infants at Cerro de Pasco than in those at Lima (Table 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Values of O₂ and CO₂ under warm conditions (~ 4.4 ml/ min/kg) did not differ in infants born at low altitude and high altitude, and were similar to those previously reported in term infants during the first day of life at 34° C ambient temperature (10). The similarity in O₂ in infants born at low altitude and high altitude confirms what was previously observed in a study of infants in Santa Cruz (400 m) and La Paz (3,600 m), Bolivia (13), but in that study, ambient temperature was not controlled and the metabolic values were higher than those found in the present study. Also, because pulmonary ventilation was similar in the two groups in the Bolivian study (13), the similarity in gas metabolism indicates that high-altitude infants compensate for the lower inspired O₂ concentration through increased O₂ extraction from the inspired air.

Once exposed to the cool ambient temperature of 26° C, the O₂ of infants in Lima increased by about 34% (i.e., they had a thermogenic response characteristic of all homeotherms, and which is known to be present at birth [10, 14, 15]). On the other hand, the increase in the high-altitude group was small and not significant. In both groups, the HR response was qualitatively similar to that of metabolism.

The metabolic results in the present study can be schematically summarized by use of the thermal conductance model (Figure 3). When O₂ = 0, ambient temperature equals Tb. The slope of the dashed line in Figure 3 indicates the increase in O₂ required to maintain Tb per unit decrease in ambient temperature, and is therefore proportional to the infant's thermal conductance. The intersection of this curve with the thermoneutral O₂ line (Figure 3, dotted line) indicates the lower critical ambient temperature (Tlc) (i.e., the temperature below which O₂ must increase in order to maintain Tb). With this representation, it is apparent that in the infants born at high-altitude, Tlc was about 3° C lower than in those born at low altitude, in part because of a slightly lower Tb, and more importantly because of a lower thermal conductance. The lower body weight of the high-altitude group was not a contributing factor, because the larger surface-to-volume ratio should eventually have increased thermal conductance.

View larger version (16K): [in this window] [in a new window]   Figure 3.   Schematic representation of changes in O2 as a function of ambient temperature. Above the lower critical temperature (Tlc), O2 remained constant as temperature increased. Below Tlc, O2 increased linearly with decreasing ambient temperature. The straight line describing thermogenesis extropolates to the abscissa to intersect at the temperature of the body (Tb). A decrease in conductance reduces Tlc and decreases the slope of thermogenesis. Values are mean ± SE.

Acute hypoxia is known to depress thermogenesis in all its forms (1), and this is particularly apparent in the newborn mammal (4, 10, 16). However, the first postnatal hours are accompanied by a rapid increase in blood oxygenation, from about 27 mm Hg in utero to about 74 mm Hg at 5 h after birth (17). This sudden and rapid rise in oxygenation, or relative hyperoxia, is likely to be experienced also by the infant born at high altitude. Hyperoxia increases the thermogenic response of newborn animals and increases O₂ in infants (1, 18, 19), both at sea level and at high altitude (13). Hence, it seems very dubious that the differences in thermogenic response observed in our study can be attributed to differences in postnatal inspired oxygen pressure. On the other hand, the alternative and much more likely possibility that the results reflect a difference in oxygenation during gestation needs to be weighed against the evidence for a greater prenatal hypoxemia at high altitude.

Direct indication of a more marked fetal hypoxia in humans during gestation at high altitude is limited. In one study, scalp Pa_O2 of the human fetus during delivery was almost the same at 4,200 m (average: 19 mm Hg) as at sea level (average: 21.5 mm Hg) (20). The degree of fetal hypoxia is probably correlated with the degree of maternal adaptation to hypoxia (21), which can be influenced by genetic factors (22), and with the time of residence at high altitude (23). Indeed, it is known that heritable characteristics selected through long-term residence at high altitude improve uteroplacental circulation (24) and obviate the need for a major increase in the oxygen carrying capacity of the fetal blood. In fact, hematocrit and hemoglobin of the high-altitude newborn infant would seem not to be or only slightly increased (21, 25). It should be noted that even a small decrease in Pa_O2, possibly within the instrumental error and insufficient for major hematopoietic responses (which are Pa_O2-dependent), could have physiologic implications because of the steepness of the fetal hemoglobin curve and the significant reduction in arterial O₂ content. Body weight has been consistently found to be reduced in infants born at high altitude, and this was also the case in a survey of several million birth files in the United States, for infants born at altitudes of up to 3,100 m, in which socioeconomic factors were taken into account (28). The birth weights in Cerro de Pasco in the present study are in agreement with this and extend the general trend (Figure 4), which for altitudes above 1,300 to 1,500 m suggests a decrease in birth weight of about 100 g for every 500 m in altitude. Additionally, reports of a positive correlation between the degree of maternal hyperventilation and infant birth weight (21, 31) further indicate that in the fetus at high altitude, oxygen supply is less than at low altitude.

View larger version (13K): [in this window] [in a new window]   Figure 4.   Body weight at birth in Lima (50 m) and Cerro de Pasco (4,330 m), combined with previously published data from a large survey of infants from many localities in the United States (28) and from La Paz, Bolivia (3,600 m) (29). Barometric pressure (Pb) and the inspired oxygen pressures (PO2) at the various altitudes (H, km) (y-axes, right side) were calculated from the simplified version (30) of the formula provided by the International Civil Aviation Organization, where Pb (mm Hg) = 760 · [288.15/(288.15  6.5  H)]5.256 and PO2 = 0.2094 · Pb.

Prenatal hypoxia increases catecholamine release, and this can promote peripheral vasoconstriction (32) and a reduction of Tlc (33). In such a case, the reduced thermal conductance of the infant born at high altitude (Figure 3) could be interpreted as the result of more efficient mechanisms of heat conservation via peripheral vasoconstriction, as reflected by lower skin temperatures (Table 2). Alternatively, these results for the high-altitude infant are compatible with the possibility of a lower thermogenic capacity. Adult highlanders have a reduced metabolic response to cold despite their normal shivering, and the response is normalized upon descent to sea level (34). This has been interpreted as indicting that high altitude primarily inhibits the nonshivering component of thermogenesis. In 2-d-old rats hypoxic during gestation but studied under normoxiac conditions, thermogenesis was as in control rats (7), presumably because, as mentioned earlier, the neonatal condition of relative hyperoxia may have increased their O₂. Nevertheless, the gestationally hypoxic rats' brown fat adipose tissue was hypoplastic, with a decrease in its mitochondrial thermogenin content, which by uncoupling oxidation and phosphorylation provides nonshivering heat production (7). Hence, it is possible that in the high-altitude infant, reduced thermal conductance reflects reduced thermogenesis rather than, or in addition to, greater efficiency in the control of heat loss. In this case, Tb in the present study might not have remained at the warm value had the cooling stimulus been maintained for a longer period. In other words, in the high-altitude group of infants, the maintenance of Tb during cooling could have occurred only because of an experimental exposure that for obvious ethical reasons had to be confined to a short period. In this respect, the general trend toward lower Tb and skin temperature values in the infants in Cerro de Pasco may be interpreted as the effect of a lower thermogenic capacity and of a lower set-point for thermoregulation. These alterations need not necessarily be interpreted as disadvantageous at high altitude, an environmental condition often characterized not only by hypoxia but also by lower ambient temperatures. For example, a slight reduction in Tb may carry an evolutionary advantage by increasing the affinity of hemoglobin for oxygen and reducing the gradient for heat loss. A lower value and decreased set point of Tb also imply that for any given thermal conductance, homeothermy can be achieved with less metabolic effort.

In conclusion, we found that 1-d-old infants at high altitude showed a smaller increase in O₂ in response to cooling temperatures than did infants born at sea level. Several considerations among a number of possibilities suggest that this may reflect a decreased thermogenic capacity because of the greater hypoxia during gestation at high altitude.