Travel to High Altitude during Pregnancy: Frequently Asked Questions and Recommendations for Clinicians

Under what circumstances do pregnant women travel to HA?

We have to distinguish different circumstances that lead to different risks:

- Ascent by cable car or chair lift for a few hours with or without planned exercise at HA.

What are the physiologic changes experienced by women during pregnancy at SL and how does exposure to HA alter these changes?

Cardio-respiratory responses to pregnancy begin early in gestation reflecting, in some respects, a continuation of those occurring with each menstrual cycle (Chapman et al., 1997), and persist for one or more weeks postpartum. These are due to the combined effects of progesterone and estradiol that serve to increase the respiratory drive and, together with a rise in metabolic rate, raise ventilation up to 50%, due mainly to a 40% increase in tidal volume (Bobrowski, 2010; Moore et al, 1987a). As a result, PaCO₂ decreases and PaO₂ increases, reaching values as high as 106–108 mmHg during the 1st trimester or 100–103 mmHg in the 3rd trimester at SL (Weinberger et al., 1980). This does not change SaO₂ which is already maximal at SL. According to the increased metabolic demands of mother and fetus, maternal oxygen consumption increases 20% during pregnancy, being greatest at term, 10% higher in twin gestations, and proportional to maternal weight gain.

Studies comparing minute ventilation in pregnant women at SL and HA show a similar increase from nonpregnant values in both groups, with the absolute values being higher at HA (McAuliffe et al., 2001; Moore et al., 1982a). It is important to mention again that all the data available have been obtained in HA residents. Pregnant women have higher PaO₂ and lower PaCO₂ than nonpregnant women at HA (Hellegers et al., 1961, McAuliffe et al., 2001). SaO₂, which is not maximal at HA, increases. Combined with the higher hemoglobin levels present in these HA residents (despite hemodilution due to plasma volume increases), CaO₂ is maintained (Moore et al., 1982a). At HA, renal excretion of bicarbonates does not compensate fully for the respiratory alkalosis and blood pH is slightly more alkalotic than at SL during pregnancy (Bobrowski 2010; McAuliffe et al., 2001).

During pregnancy at SL, cardiac output increases about 40% as a result of increased heart rate, stroke volume, and blood volume. The increase in blood volume is mainly due to a 40% expansion in plasma volume. Despite this increase, blood pressure decreases slightly during normal pregnancy, as peripheral vascular resistance is reduced throughout the systemic circulation (Kametas et al., 2004a). Due to a disproportionately lower vascular resistance, uterine artery (UA) blood flow increases early in gestation and continues to rise throughout pregnancy as the result of cytotrophoblast invasion of spiral arterioles and changes in upstream vessels. The net result is a ∼60-fold or greater increase in total blood flow to the uteroplacental circulation that is due to an increase in UA diameter and faster blood flow velocity (Palmer et al., 1992).

Cardiovascular changes at HA are similar to but less marked than those seen at SL; for example, a study in Peru found that cardiac output increased only 17% due to a lesser rise in heart rate, stroke volume, and blood volume than at SL (Kametas et al., 2004a). There is also less rise in UA blood flow during pregnancy at HA (Julian et al., 2009a; Zamudio et al., 1995a). Placental morphology is well preserved at HA and likely to benefit placental exchange, with larger placenta relative to fetal size, membrane thinning, and greater branching of fetal capillaries (Moore et al., 2011). These modifications are more marked in women of Andean vs. European ancestry (Jackson et al., 1987).

Are there harmful effects of pregnancy at HA?

The first recognized effect of HA was the increased incidence of low birth weight (LBW) due to intrauterine growth restriction (IUGR) (Lichty et al., 1957). Since confirmed in multiple studies over more than 50 years, the reduction is due primarily to the effects of chronic hypoxia and not some other attribute associated with high-altitude residence (Giussani et al., 2001; Jensen and Moore, 1997). The incidence of gestational hypertension and preeclampsia are increased ∼ 3-fold at high than lower altitudes in Colorado (Moore et al., 1982b, Palmer et al., 1999) and elsewhere in the world (review in Moore et al., 2011). Preeclampsia increases maternal, fetal, and neonatal morbidity and mortality, whereas LBW is also an important risk factor during the neonatal period as well as later in life.

On average, newborns weigh 100 g less for each 1000 m increase in altitude. Stratified studies show a threshold effect with the decrease in birth weight beginning above 1500 m in USA and being more pronounced above 2500 m (Yip, 1987). In Peru, the marked effect begins a little higher, ∼2000m (Mortola et al., 2000). The decline in birth weight is due to a slowing of fetal growth that is detectable beginning in the third trimester (Krampl et al., 2000; Unger et al., 1988). Studies comparing populations with differing durations (in generations) of HA residence show that the magnitude of birth-weight fall is about half as great in multigenerational vs. shorter-resident HA populations (Moore, 2001a, Moore et al., 2011). Higher maternal ventilation and consequently higher levels of SaO₂ provide some protective effect in Colorado, and Andean and Tibetan populations (Moore et al., 1982c, Moore et al., 1986, Moore et al., 2001b, Vargas et al., 2007) but the main factor protecting from IUGR in multigenerational- vs. shorter-resident groups in long-term residents is higher UA blood flow (Julian et al., 2009, Julian et al., 2011; Zamudio et al., 2007). New genomic techniques have been used for identifying gene regions that are distinctive in Andean and Tibetan populations (Beall et al., 2010, Bigham et al., 2010, Moore et al., 2011). However the physiological mechanisms by which such genetic variation influences fetal growth or other attributes of HA populations, and the relative roles of lower UA blood flow or decreased oxygen, glucose, or other nutrient delivery due are presently unknown (Zamudio et al., 2010).

HA newborns have higher cord blood hemoglobin and percent fetal hemoglobin, likely reflecting greater fetal erythropoiesis due to hypoxia (Ballew and Haas, 1986), and an increased frequency of neonatal hyperbilirubinemia (Leibson et al., 1989).

While hypertensive complications of pregnancy are more common at HA, such conditions appear to influence birth weight in the same way as at low altitude with these two risk factors acting additively, not synergistically (Jensen and Moore, 1997). However, hypertensive complications of pregnancy appear to have a greater impact on intrauterine or infant mortality at HA, raising the frequency of fetal deaths and placental abruption (Browne et al., 2011; Keyes et al., 2003).

Other complications of HA pregnancies have not been studied as extensively as preeclampsia/gestational hypertension and IUGR. There are some reports from the Andes and Colorado of greater bleeding, placenta praevia, polyhydramnios, premature rupture of membranes, and preterm labor (Keyes et al., 2003; Yancey and Richards, 1994). However, the incidence of preterm deliveries is no greater at high than lower altitudes in Colorado, and the average gestational age is only 0.5 week lower at the highest altitudes (Jensen and Moore, 1997). A few articles report a slight increase in mainly craniofacial congenital anomalies at HA (Alderman et al., 1995, Castilla et al., 1999).

There are no epidemiologic studies on pregnancy complications in HA visitors. The limited data consist of one report on foreign women giving birth to babies at HA (3600 m) in La Paz, Bolivia, and one survey among obstetrical care providers in Colorado. The report from Bolivia concerned the pregnancy outcomes of 12 pregnant women, six of whom had been living at HA since before conception and who delivered babies weighing on average 2917 g (2600–3100 g range), four women who arrived in La Paz during their 1^st trimesters and two who arrived during their 2^nd trimesters. Three of the four of the women arriving in the 1^st trimester delivered prematurely and the fourth had a threatened abortion; their babies weighed on average 2175 g (range 1800–2800 g). One of the two women arriving in the 2^nd trimester delivered preterm and the other at term to babies weighing, on average 2900 g (Falk, 1983). The survey of obstetrical care providers indicated that the most frequent complications were spontaneous abortion in the 1st trimester and intrauterine fetal demise, placental abruption, premature rupture of membranes, and preterm labor after mid-pregnancy, but no actual incidence figures or altitude comparisons were available (Niermeyer, 1999). Dehydration and vigorous exercise were cited as risk factors for such complications.

A higher incidence of early spontaneous abortion has been suspected in HA residents, but the most recent studies do not confirm this (Vitzthum and Wiley, 2003). There are no data in HA visitors or short-term residents. An indirect approach for assessing the potential risk for short-term visitors are reports concerning pregnant flight attendants, who are frequently exposed to pressurized altitudes of ∼2500 m for many hours. A recent review suggested a greater risk of spontaneous abortion or intrauterine fetal demise in such flight attendants; however if spontaneous abortion was considered alone, the risk was not increased. The authors cautioned that the methodology of some studies was questionable and obviously other factors than hypoxia may be contributing (Magann et al., 2010). Another small study reported a high incidence of miscarriage in air medical flight crews exposed to higher altitudes than in commercial flights, but the associated risk factors do not allow establishing a causal link with hypoxia (Van Dyke, 2010).

Are there differences between HA visitors and HA residents (excluding genetic factors)?

Like all HA visitors, pregnant women need to acclimatize and may develop acute mountain sickness (AMS) and other altitude-related complications during the first days at HA. There are no specific studies of AMS in pregnancy. In the Colorado survey, the frequency seemed similar or slightly lower in pregnant visitors than in the general population (Niermeyer, 1999), which could be related to pregnancy-induced hyperventilation.

For pregnant women going from LA to HA, hypoxia would be expected to further increase pregnancy-induced hyperventilation (Niermeyer, 1999), but there are no specific studies to affirm this. Two studies dealing with the effects of exercise at HA (as detailed below) saw no further increase in the preexisting hyperventilation of pregnancy following acute exposure to 1830 m or 2238 m (Artal et al., 1995, Huch, 1996). A study on 10 women on commercial flights also concluded that there was no further increase in pregnancy-associated hyperventilation at a pressurized cabin altitude of 2400 m (Huch et al., 1986). In all cases, women were in their 3rd trimester. Needed are studies earlier in pregnancy and for longer duration.

LA women do not have the benefit of higher pre-pregnant hemoglobin concentrations, although some degree of hemoconcentration occurs quite rapidly following ascent due to diuresis. Hence CaO₂ might be expected to be lower at HA, at least before the erythropoietin-stimulated rise in red blood cell production became apparent after several weeks.

There are no studies on changes in UA blood flow in HA visitors. Such studies are greatly needed and should involve assessing the effects of minor perturbations in UA blood flow caused by postural changes (eg, shifting from the left lateral to recumbent positions) or exercise (Jeffreys et al., 2006).

Are there studies concerning the effects of exercise during pregnancy?

Current recommendations state that healthy pregnant women benefit from regular physical activity and can adopt the same recommendations as the general population: “30 minutes or more of moderate exercise on most, if not all, days of the week”. A wide range of activities are considered safe. Sports carrying a high risk of injury or strenuous activities are not recommended. Proper caloric and fluid intakes are essential (ACOG, 2002, Melzer et al., 2010).

Oxygen consumption during submaximal exercise is higher during pregnancy. Absolute Vo₂ max (l/min) does not change in pregnant women, however if expressed relative to body weight (ml/kg.min) it may be decreased in late pregnancy due to maternal weight gain (Lotgering et al., 1991, Pivarnik et al., 2003). Despite this apparent maintenance of aerobic fitness during pregnancy, it is noticeable that the heart rate (HR) reserve is lower, due to higher resting HR and slightly lower maximal HR (Wolfe, 1993). As a result, it has been recommended to monitor exercise intensity using modified HR target zones and ratings of perceived exertion or using the easy “talk test” to avoid overexertion (Wolfe and Davies, 2003). A common concern is the potential for competition for blood supply between skeletal muscles and the uteroplacental circulation, compromising oxygen and glucose availability to the fetus either acutely or chronically, depending on the intensity, duration, and frequency of exercise. Studies on fetal heart rate show that there is typically an increase in HR with submaximal maternal exercise. There are some reports of transient fetal bradycardia in response to maximal exercise, without adverse birth outcome (Wolfe et al., 1994). There are conflicting results on the effect of exercise on birth weight, probably due to differences in amount and intensity of exercise and in nutrition. However, women engaging in frequent exercise appear to have babies whose birth weights are in the lower part of the normal range and who have less body fat (Pivarnik et al., 2003). Doppler studies of the uteroplacental circulation during or after exercise give conflicting results: some have described an increase in the vascular resistance during strenuous exercise whereas others have not (Jeffreys et al., 2006, Veille, 1996). On the fetal side, exercise does not seem to influence the umbilical artery resistivity index (Veille, 1996). Direct measurements of UA blood flow during exercise in women and sheep suggested that uterine blood flow decreases with both the level and the duration of exercise (Jeffreys et al., 2006, Lotgering et al., 1983).

At HA, in nonpregnant subjects, SaO₂ falls during exercise. Despite a slight increase in hemoglobin due to hemoconcentration, CaO₂ falls as well. HR reserve is decreased due to higher HR at rest and lower maximal HR after acclimatization (Grover et al., 1986); these changes are already present in pregnant women at SL and hence HR reserve could be still lower at HA. There are very few studies on exercise and pregnancy at HA and they involve very few subjects (7–12), usually sedentary and unfit, for a short duration of exercise (3–12 min), following acute exposure to moderate altitudes (1830–2228 m). Moreover, all studies involved women in their 3rd trimester. Artal studied 8 women exercising at SL and after acute exposure at 1830 m. One subject could not complete the test at altitude because of premature labor which required 72 h observation but was without other consequences. Fetal bradycardia for 2 min was observed in another subject after exercise at altitude. In the seven women who completed the study, maximal oxygen consumption and work levels were lower at altitude but submaximal exercise efficiency unaffected. The ventilatory variables were not significantly influenced by altitude, whereas both cardiac output and stroke volume were elevated at altitude at rest, but not during exercise, suggesting a lower cardiac reserve at altitude. Maximum duration of the test was 12 min (Artal et al., 1995). Two studies in Switzerland involved 12 and 6 sedentary pregnant women respectively, one at 2228 m after cable car ascent and the other at simulated altitude of 2200 m (Huch, 1996). There was neither greater hyperventilation nor higher HR at rest or during the 3 min bicycle exercise at altitude compared with pre-ascent data. Six subjects experienced regular mild contractions during the cable car ascent, which did not increase during or after exercise. All fetal heart rate traces were normal except for one subject in each study (both smokers), who experienced moderate and transient bradycardia after exercise and then was followed by reduced variability before returning to normal after descent in one subject. In all studies, women were reported to deliver normally at term but no other details on the newborn were provided.

What evidence exists concerning the effects of travel during pregnancy?

General recommendations for travel by air, to remote areas or exotic countries, or for long duration during pregnancy must be followed (ACOG, 2009, Hezelgrave et al., 2011). Travel during the 2nd trimester is considered less risky for travel in general. Ectopic pregnancy and miscarriage are the main risks in the 1st trimester, preterm delivery in the 3rd.

The pregnant woman must be aware of the risk of being far from medical or obstetrical assistance if needed. Even in developed countries, emergency services in HA resorts are more oriented to traumatic injuries, and practitioners are not always as well trained for obstetrical care. For exotic locations, some vaccinations or drugs useful for prophylaxis or treatment of infections are contra-indicated during pregnancy, and some infectious diseases may be more severe during pregnancy (diarrhea, malaria, dengue, hepatitis E) (Barry and Bia, 1989, Jean et al., 2005). In all cases, it is important to have appropriate health insurance for mother and baby for travels after mid-pregnancy.

Theoretically, pregnancy increases the risk of venous thromboembolism, but there are no reports of higher incidence in pregnant women during air flight or at HA. However, standard means for prevention are at least as important for pregnant women as in the general population (Brenner, 2009, Kametas et al., 2004b).

Is it safe to take drugs to alleviate HA-related illnesses while pregnant?

The most widely used drugs at HA are acetazolamide (used to prevent or treat AMS), nifedipine and dexamethasone (used to treat HA pulmonary edema and HA cerebral edema, respectively). Nifedipine (except sublingual) and dexamethasone are not contra-indicated during pregnancy, but obviously the best recommendation is to prevent or avoid these two complications, which can be life-threatening for both the mother and the fetus. The frequency of AMS can be greatly lowered by gradual ascent and limiting exercise intensity, and such approaches rather than taking drugs are recommended during pregnancy. However, is it possible to use such drugs if AMS or other symptoms develop? What answer should be given to a pregnant woman who took acetazolamide or other drugs before knowing that she was pregnant?

Acetazolamide is teratogenic in different species of rodents but not in monkeys (Nakatsuka et al., 1992; Scott et al., 1990). Safety in human pregnancy has not been established, so it is usually not recommended, especially during 1st trimester, except in case of imperative indication without other alternatives such as would be the case if the potential benefit to the mother justifies the potential risk to the fetus. Its use has been reported in a small series of patients with neurologic or ophthalmologic indications, given at doses 750 to 1000 mg daily (much higher than usual doses of 250 to 500 mg daily for preventing or treating AMS) and throughout pregnancy, including 1st trimester, without teratogenic effect (Lee et al., 2005). There is only one report of an association with sacrococcygeal teratoma (which is a tumoral process, not a malformation) and therefore a causal link cannot be affirmed nor excluded (Worsham et al., 1978). There are a few reports of metabolic disorders in the neonate, mainly metabolic acidosis, after treatment throughout pregnancy or only in the few days preceding delivery (Merlob et al., 1990, Ozawa et al., 2001). All cases were transient and not severe. Metabolic acidosis may increase the risk of severity of neonatal jaundice for a given level of bilirubin, but again no severe case has been reported (Sawitsky et al., 1968). Even if there is little evidence for significant adverse effects, it is advisable to avoid taking acetazolamide for AMS prevention and to use, instead, staged ascent to achieve natural acclimatization. Moreover, in animal studies, decreased uterine blood flow potentiates acetazolamide teratogenicity (Ugen and Scott, 1987) and therefore may be an additional theoretical reason to avoid it in HA pregnancies.

Conditions for which travel to HA is contra-indicated

There are clear contra-indications for going to HA when some condition(s) may already compromise the pregnancy at LA, as such a condition would be expected to be aggravated by added hypoxia (Jean et al., 2005). Most of these contra-indications concern pregnancies after 20 wk, when the risk of preeclampsia begins and/or the major increase in fetal size occurs:

• maternal heart or lung disease (throughout pregnancy depending on severity)

Pregnant women who are anemic and who plan to stay for weeks to months at HA should be evaluated for iron stores (ferritin level) to cope with the stimulated erythropoiesis, and iron supplementation may be advisable. Gonzales et al. (2009) recently found that the optimal hemoglobin level or that with the lowest risk of adverse outcome for pregnancies at HA was between 11 and 13 g/dl; hence higher values are not advisable as they could decrease UA blood flow due to increased viscosity.

Smoking, apart from its impact on the mother, may have greater effects on the fetus given the higher affinity of fetal than adult hemoglobin for binding carbon monoxide (Di Cera et al., 1989). Consistent with this, Moore observed a 2- to 3-fold greater depressant effect of maternal smoking on birth weight at 3100 m compared to lower altitudes (Moore et al., 1982c), and Huch observed anomalies in fetal heart rate during HA acute exposure only in smokers (Huch, 1996).

Considerations relating to the period during pregnancy when traveling to HA

Considerations pertaining to exercise at HA

The few studies available show that acute exposure up to 2500 m with moderate or brief duration exercise poses very little risk to mother and fetus during the 3rd trimester in nonsmoking women experiencing a normal pregnancy. Given that fetal demand is greater during the 3rd trimester, safety is likely throughout pregnancy. There are no data for longer exercise or at higher elevations.

As pregnancy, altitude, and exercise cause hyperventilation, it is important to maintain adequate hydration for all women, especially given the low humidity levels present at altitude and the potential risk of dehydration for pregnancy complications (Niermeyer, 1999). The pregnant woman must also be aware that weight gain, change in the center of gravity and joint laxity may increase the risk of falls, injuries and strains, and so it is recommended to avoid activities with high risk of fall, such as skiing, after mid-pregnancy.

Risks of giving birth at HA

If possible, it is always preferable for the mother and the fetus to deliver in normoxia. But the answer is dependent on the level of health-care facilities at the HA location, as the baby may need supplemental oxygen at birth and during the cardio-pulmonary transition undergone during the first days after birth. This point and other questions about potential later effects have been discussed in other studies (Bakr and Abbas, 2006, Jayet et al., 2010, Niermeyer et al., 1993, Niermeyer et al., 1995, Niermeyer et al., 2009, Yancey et al., 1992).