Neonatal echocardiograms of macrosomic neonates

OBJECTIVE: To compare echocardiograms of macrosomic and healthy full term neonates whose weight was appropriate for gestational age (AGA). METHODS: Echocardiography was performed on 9 healthy full term AGA neonates and 15 macrosomic neonates. A data base was generated from valid echocardiographic measurements on each infant. RESULTS: Macrosomic infants were heavier than control infants and had a greater body surface area, but their mean cardiac dimensions were similar. The mean LVES volume was smaller than that of the control group when expressed as a fraction of individual LVED values (0.61 +/- 0.04 vs 0.64 +/- 0.02; p = 0.02). When comparing IVS/PW, it was observed that the ratio was up to and including 1.33 in the control group, while the upper limit of the ratio of the nondiabetic macrosomic infants was 1.6. The shortening fraction (SF%) was increased in comparison to infants of normal weight (40.67% +/- 3.34 vs 36.00% +/- 1.89; p = 0.0009). The thickened IVS did not elevate SF% by decreasing LVES; the macrosomic infants had a smaller LVES mean volume. CONCLUSION: IVS/PW ratio of macrosomic infants fell outside of the normal range. Macrosomic neonates were found to have an increased SF% secondary to a proportionally smaller LVES volume, regardless of IVS thickness.     

Serum ferritin levels in preterm infants after multiple blood transfusions

We have examined the effect on iron stores of blood transfusions given to premature neonates during hospitalization in the neonatal intensive care unit as reflected by serum ferritin levels measured for 6 months after discharge. Premature infants who were transfused with more than 100 ml packed cells (group D; n = 11) had higher ferritin levels for a longer period than premature infants who were transfused with smaller volumes (group c; n = 9) or premature and mature infants who were not transfused at all (group B; n = 24 and group A; n = 21, respectively). At 4-5 months the serum ferritin levels in group D (489.8 +/- 132.1 micrograms/L; mean +/- SEM) were significantly higher (P less than 0.001) than those of the other groups. The level of group A term infants (77.5 +/- 12.5 micrograms/L) was higher than those of group B premature infants who did not receive a blood transfusion (33.0 +/- 7.1 micrograms/L) or group C who received less than 100 ml (36.5 +/- 8.8 micrograms/L packed red blood cells. However, these differences were not statistically significant. Our data demonstrate that very-low-birthweight infants who receive a large volume of packed cells during hospitalization may accumulate iron stores sufficient for red cell production during the first 6 months of life. Administration of large amounts of supplemental iron, in such cases, may be curtailed.     

Neutrophil chemotaxis and random migration in preterm and term infants with sepsis

The aim of this study was to evaluate neutrophil chemotaxis and random migration in healthy newborn infants and septic neonates with similar gestational and postnatal age. Possible relationships between chemotactic activity, random migration, causative microorganisms, and clinical course of septic infants were also investigated. The neutrophil chemotaxis and random migration was evaluated in 24 healthy newborn babies and 34 septic neonates and 20 healthy adults by modified Boyden technique. The mean neutrophil chemotaxis of healthy preterm-term infants and adults were similar (66.6 +/- 18.9, 64.4 +/- 19.9, and 74.7 +/- 17 microm, respectively). The mean neutrophil random migration of healthy term infants was not different than that of adults. But the mean neutrophil random migration of healthy preterm infants was lower than that of adults (36.9 +/- 13.7 and 43.5 +/- 1 1.8 microm, respectively) (p = 0.03). The mean neutrophil chemotaxis and random migration septic term infants were not different from the value of healthy term infants (p > 0.05). Although the mean random migration of septic and preterm infants were similar (p > 0.05), the mean neutrophil chemotaxis of septic preterm infants was lower than the value of healthy preterm infants (p = 0.04). Not only mean neutrophil chemotaxis of septic preterm and term infants were significantly lower than that of adults (p = 0.002 and p = 0.006, respectively), but also neutrophil random migration of septic preterm and term infants were significantly lower than that of adults (p = 0.001 and p = 0.005, respectively). There was no relationship between the nature of causative microorganism and neutrophil random migration or chemotactic activity. Polymorphonuclear leukocytes chemotaxis was significantly lower in preterm with sepsis compared with healthy preterm-term infants and adults. These findings may indicate deterioration in neutrophil functions in premature babies under stress but more detailed studies with larger groups are needed.     

The effect of prenatal administration of dexamethasone and ritodrine on cord blood cortisol and glucose concentrations in premature infants with respiratory distress syndrome.

The influence of maternal dexamethasone and ritodrine administration during pregnancy on cord blood cortisol and capillary serum glucose concentrations and on the incidence of respiratory distress syndrome (RDS) was studied in 30 premature infants (gestational age 27-36 weeks), and compared with a matched control group of 37 premature infants where no such medications were administered. RDS occurred less often in the treated group of infants (13.3%) than in the controls (35.1%, p < 0.01). The healthy treated infants had a significantly lower mean umbilical cord plasma cortisol concentration (5.5 +/- 1.8 ug/dl, mean +/- SD) than that observed in the controls (11.2 +/- 3.9 ug/dl, p < 0.01). Mean cord plasma cortisol concentrations increased with duration of pregnancy. No significant difference in the capillary serum glucose at 30 minutes post-delivery was found between the healthy, RDS, treated and non-treated infants. No adverse effects of steroid and ritodrine therapy were observed.     

Sucking patterns of neonates during bottle feeding: comparison of different nipple units.

The present study was designed to compare the sucking pattern of term and preterm infants during bottle feeding with different types of nipple units (Enfamil single-hole nipple units for term and preterm infants and SMA Nuk nipple units). In addition, the sucking pattern of term neonates during a feeding regimen commonly used in many feeding studies was evaluated (reservoir nipple system). In this system milk flows from a reservoir through a tube and depends on the sucking pressure generated by the infant. Only the Enfamil single-hole nipple units for term and preterm infants were compared in preterm infants. No significant difference in sucking frequency was observed in term neonates with different types of nipple units. Although the mean sucking pressures generated tended to be less among nipple units with higher flow, these differences were not statistically significant. Similarly, no significant difference in total sucking or feeding time was observed among the three nipple units tested. Sucking pressures generated by term infants were significantly less when milk flow was increased markedly utilizing the reservoir system. In preterm infants no differences in sucking frequency, sucking pressure, mean flow, or total feeding time were observed when sucking patterns with term and preterm nipple units were compared. Implications of these findings in feeding neonates are discussed.     

Neonatal jitteriness of unknown origin and circulating catecholamines

Jitteriness is a common problem affecting neonates. Although the cause can sometimes be determined by history or conventional laboratory evaluations, nevertheless, in many instances the cause of the jitteriness is unknown. To determine if either intracranial hemorrhage (ICH) or elevated catecholamines are responsible for jitteriness in neonates, we studied 34 healthy term neonates with normal hematocrits, serum glucose, Ca, Mg, Na, K and P. Thirteen of the infants had jitteriness as their only clinical problem. There were no differences in Apgar scores, birthweight, or gestational age in the two groups. Norepinephrine levels were significantly elevated in the jittery group as compared to the control group: 1276 +/- 574 vs. 914 +/- 338, p less than 0.05. Epinephrine levels were not different in the two groups. Intracranial hemorrhage was not found in any of the patients. Jittery neonates have increased sympathetic activity. What influence the increased levels have on other metabolic and hormonal systems is yet to be determined.     

Does extreme prematurity affect kidney volume at term corrected age?

Objective. Extreme prematurity exposes the neonate to a number of potential renal insults that may result in a reduced number of glomeruli and/or renal size. This may predispose these individuals to cardiovascular disease later in life. The objective was to determine using magnetic resonance imaging (MRI) whether extreme prematurity results in decreased renal volume.

Methods. Neonates <29 weeks' gestation and term infants undergoing MRI of the brain were enrolled in the study. An MRI was performed at term corrected age in the premature neonate and within the first 4 weeks of life in the term neonate.

Results. Seventeen preterm infants and 13 term infants had MRIs performed. There was no significant difference in weight and length at the time of MRI (p = 0.76 and 0.11, respectively). There was no significant difference in total renal volume or total kidney volume to weight ratio between the preterm and term neonates (p = 0.83 and 0.6, respectively).

Conclusions. At term corrected age, extremely premature neonates have the same renal volume as term infants. It is unclear whether renal volume is a good indicator of glomerular number.     

Nitrogen utilization and bone mineralization in very low birth weight infants fed partially hydrolyzed preterm formula

Partialy hydrolyzed formulas have been proposed for term and preterm infants, but decreased nitrogen and calcium intestinal absorption rates, together with plasma amino acid imbalances have been reported in preterm infants. We evaluated a new formula with modified nitrogen and calcium sources (glycerophosphate). During their second week of life, 16 preterm infants were randomly assigned to 1 of 2 groups: 9 were fed the new partially hydrolyzed formula and 7 a conventional formula. A nutrient balance was performed at the end of the first month of life. Amino acid concentrations and whole-body mineralization were measured at the end of study period (theoretical term). Birth weight and gestational age (mean +/- SD) were similar in the 2 groups (28.9 +/- 7.0 wks and 1,183 +/- 242 g v 27.7 +/- 1.0 wks and 1,139 +/- 162 g). Median nitrogen absorption rate (85% v 89%; P = .03) was lower in infants fed the new formula than in those fed the conventional one. After correction for difference in nitrogen intake, there was no significant difference in N retained between the 2 groups (P = .11). Plasma amino acid concentrations were also similar in the 2 groups. At theoretical term, median bone mineral content was not significantly different between the 2 groups (50 g/kg v 55 g/kg; P = .17) and it was close to the reference values obtained in term neonates just after birth. As long as nitrogen content is 10% higher in protein hydrolyzed formula than in entire protein formula, appropriate nitrogen retention, plasma amino acid profile can be achieved with the new partially hydrolyzed formula. In both groups, bone mineralization at theoretical term was close to the mineralization observed term neonates just after birth.     

Relationship between cortisol levels in umbilical cord plasma and development of the respiratory distress syndrome in premature newborn infants.

The aim of this study was to determine whether there is a decrease in fetal cortisol levels associated with the respiratory distress syndrome (RDS). Eighteen newborn infants of less than 37 weeks' gestation who developed moderate to severe forms of RDS did have a significantly lower (P less than 0.02) mean cord plasma cortisol concentration at birth than that observed in 67 unaffected infants of similar gestational age; mean values +/- standard errors were 3.36 +/- 0.42 and 5.58 +/- 0.43 mug per 100 ml, respectively. However, whether or not RDS developed in neonates appeared to depend more upon the degree of prematurity (with a 71.5% incidence in gestations of less than 32 weeks compared to 17.1% in those of 32 to less than 37 weeks) than upon cortisol levels at delivery. Bood cortisol levels in the first days of life of four infants with RDS were considerably increased in comparison to those at birth. Mean cord plasma cortisol concentrations increased with duration of pregnancy, with the previously observed value for term infants (of 37 or more weeks) being approximately twice that for infants of less than 32 weeks' gestation. These findings appear to justify carefully controlled studied with antepartum glucocorticoid administration with the aim of reducing the incidence of RDS in premature newborn infants.     

Renal function and renal failure in the newborn

Renal function in the newborn infant varies with conceptual age and should be evaluated in this context. Very preterm infants less than 34 weeks' conceptual age have reduced GFR and tubular immaturity in the handling of filtered solutes when compared to term infants. Premature infants between 34 and 37 weeks' conceptual age undergo rapid maturation of renal function similar to term infants, with establishment of glomerulotubular balance early in the postnatal period. ARF in neonates differs from that seen in older children and adults in that ischemic (e.g., hypoxic) insults and congenital malformations constitute the major pathophysiologic mechanisms responsible for clinically observed oliguria and azotemia. Principles of conservative management are similar to those used in older children except for the greatly increased insensible water loss requirements of the very preterm and premature infant. Technical advances have added peritoneal dialysis and CAVH to the therapeutic regimen for persistent ARF or life-threatening complications of reduced renal function.     

Preliminary observations of urinary calcium and osteopontin excretion in premature infants, term infants and adults

Osteopontin is an acidic glycoprotein which may prevent nephrocalcinosis and nephrolithiasis by inhibiting the growth and retention of calcium oxalate crystals within the tubular lumen. The purpose of this study was to obtain preliminary data regarding urinary osteopontin in premature infants at risk for nephrocalcinosis. We examined urinary osteopontin concentration in premature infants, term infants and adults, and examined the relationship between urinary calcium and osteopontin concentration in these groups. The urinary osteopontin concentration of 17 premature infants of 3.7 +/- 1.2 microg/ml was not significantly different from the urinary osteopontin concentration of 12 term infants of 6 +/- 6 microg/ml, while the urinary osteopontin concentration in 23 urine specimens from adults of 27 +/- 15 microg/ml was significantly higher than premature infants and term infants (p < 0.05). Urinary osteopontin concentration did not correlate with urinary calcium concentration in premature infants, while there was a correlation between the osteopontin/creatinine ratio and calcium/creatinine ratios in premature infants. Diminished urinary concentration of osteopontin may enhance the risk for nephrocalcinosis in premature infants.     

Fluconazole therapy in neonatal candidemia

We reviewed 62 episodes (from 59 infants) of neonatal candidemia that occurred between January 1994 and June 1999. Except 5 term babies, all infants were premature (median gestational age [GA], 30 weeks) and birth weight was less than 2,500 g (median, 1,300 g). Most infants had reported risk factors and other neonatal problems. The age at onset of candidemia ranged from 15 to 173 days with a median of 34 days. In addition to catheter removal, all but one infants received antifungal agents and candidemia was eradicated subsequently in 46 episodes (75%). Eighteen infants with 19 episodes ever received fluconazole therapy. Fluconazole was administered as the first line agent in 6 episodes and successfully cleared candidemia in 5 episodes. Fluconazole was used as an alternative agent in an additional 13 episodes after amphotericin B (am B) +/- flucytosine were given for a period without a satisfactory result and eradication of candidemia was achieved in 8 episodes subsequently. All 18 infants tolerated fluconazole well and no withdrawal was required on account of its adverse effect. In contrast, am B alone was administered as the first line agent in 55 episodes and successfully cleared candidemia in 32 episodes (58%). This retrospective analysis suggests that fluconazole appears to be safe in neonates and can be used as an alternative agent in treating neonatal candidemia. A large-scaled prospective study may be needed.     

Newborn jaundice technologies: unbound bilirubin and bilirubin binding capacity in neonates

Neonatal jaundice (hyperbilirubinemia), which is extremely common in neonates, can be associated with neurotoxicity. A safe level of bilirubin has not been defined in either premature or term infants. Emerging evidence suggest that the level of unbound (or "free") bilirubin has a better sensitivity and specificity than total serum bilirubin for bilirubin-induced neurotoxicity. Although recent studies suggest the usefulness of free bilirubin measurements in managing high-risk neonates, including premature infants, no widely available method exists to assay the serum free bilirubin concentration. To keep pace with the growing demand, in addition to reevaluation of old methods, several promising new methods are being developed for sensitive, accurate, and rapid measurement of free bilirubin and bilirubin binding capacity. These innovative methods need to be validated before adopting for clinical use. We provide an overview of some promising methods for free bilirubin and binding capacity measurements with the goal to enhance research in this area of active interest and apparent need.     

Cue-based oral feeding clinical pathway results in earlier attainment of full oral feeding in premature infants.

OBJECTIVE: To study whether a cue-based clinical pathway for oral feeding initiation and advancement of premature infants would result in earlier achievement of full oral feeding. STUDY DESIGN: Age of achievement of full oral intake was compared for two groups of preterm infants; a prospective study group vs historic cohort controls. Study infants had oral feedings managed by nurses using a clinical pathway that relied on infant behavioral readiness signs to initiate and advance oral feedings. Controls had oral feedings managed by physician orders. RESULT: Fifty-one infants (n=28 study and n=23 control) were studied. Gender distribution, gestational age, birth weight and ventilator days were not different between groups. Study infants reached full oral feedings 6 days earlier than controls (36+/-1 3/7 weeks of postmenstrual age (PMA) vs 36 6/7+/-1 4/7 weeks of PMA, P=0.02). CONCLUSION: The cue-based clinical pathway for oral feeding initiation and advancement of premature infants resulted in earlier achievement of full oral feeding.     

Serum tobramycin levels in low- and very low-birthweight infants

This study was designed to evaluate the serum concentration of tobramycin sulfate following a 2.5-mg/kg intravenous infusion in 43 premature infants on days 1, 3, and 5 of age (therapy). Twenty premature infants weighing 1500 gm or less at birth and 23 others whose birthweights ranged from 1501 to 2500 gm made up the study population. Serum tobramycin levels were measured by an enzymatic immunoassay (EMIT) at one, four to six, and 12 hours after injection. Peak serum levels increased from day 1 (means, 5.2 +/- 2.2 mcg/ml) to day 3 (means, 6.1 +/- 2.6 mg/ml) and then remained unchanged at day 5 (means, 6.1 +/- 2.4 mg/ml). Approximately 40% of the study population presented trough levels above 2 mcg/ml on day 1 and over 70% on days 3 and 5. No evidence of renal toxicity or auditory dysfunction was observed. In light of the high trough levels observed during the first week of life in premature infants, it may be judicious to monitor serum tobramycin concentration and to decrease the dosage or to prolong the dose interval in order to maintain trough concentrations below 2 mcg/ml.     

Serum netilmicin levels in premature AGA infants

Safe use of aminoglycosides requires close monitoring of serum concentrations. Limited information coupled with marked changes in fluid compartments and renal function during the first week of life in premature neonates makes interpretation of peak and trough levels very difficult. This study was designed to measure serum netilmicin levels following a 2.5 mg/kg IV push infusion. Blood samples were taken on the 5th day of therapy 1 hour before and 1, 6, and 11 hours after a dose. Fifteen premature infants weighing 1000-1500 gm at birth and 20 others whose weight ranged from 1501-2750 gm comprised the study population. All premature infants were appropriate for gestational age (AGA) and of them, only two were severely asphyxiated. At the time of the study, 10 neonates were still on respirators. Serum and urine sodium and creatinine, BUN, and urinalysis were obtained in 28 of these infants. No evidence of renal dysfunction was found. All infants received 100 mg/kg IV ampicillin every 12 hours, but none were being treated with diuretics. Serum netilmicin levels were measured by an enzymatic immunoassay, peak and trough were calculated by extrapolating the first order decay curve. Peak levels ranged from 3.4 to 14 micrograms/ml (means 6.1 +/- 2.5 micrograms/ml SD) and 90% of them were above 4 micrograms/ml. Half of the small premature infants (1000-1500 gm birthweight) presented trough values above 3 micrograms/ml. Pharmacokinetic analysis of our data predicts that a 2.5 mg/kg loading dose followed by 2 mg/kg given every 12 hours will decrease by one-half the number of small prematures exceeding the considered "safe" trough level (greater than 3 micrograms/ml).     

THE AUSTRALIAN & NEW ZEALAND JOURNAL OF OBSTETRICS & GYNAECOLOGY

In order to assess the possible influence of fetal polypeptide hormones on birth size, somatomedin-like receptor activity (SmLRA) (n = 281), prolactin (n = 158), growth hormone (n = 133) and insulin (n = 108) concentrations were measured in the cord blood of 281 singleton infants born after different complications of pregnancy. Infant sex did not significantly affect the concentration of any hormone. SmLRA concentrations appeared to rise from 25 to 38 weeks of gestation, but thereafter fell. Concentrations of prolactin, growth hormone and insulin correlated negatively with gestational age. Insulin emerged as the hormone most clearly related to fetal growth, since concentrations were high (mean +/- SD = 20.6 +/- 11.5 uU/ml) in serum from infants of diabetic mothers (IDM) and other large-for-dates infants (10.1 +/- 4.8 uU/ml), but low (5.3 +/- 0.5 uU/ml) in infants who were small-for-dates (SFD). In contrast, SmLRA concentrations were high in serum from SFD infants (0.63 +/- 0.29 U/ml) and low in IDM (0.43 +/- 0.16 U/ml). Prolactin concentrations were higher in serum from SFD infants (212 +/- 101 ng/ml) and from IDM (237 +/- 182 ng/ml) than from normal infants born at term (139 +/- 68 ng/ml). Administration of intramuscular betamethasone to pregnant women in premature labour resulted in significant elevations in the concentrations of prolactin and insulin in cord blood.     

Urinary uric acid/creatinine ratio as an additional marker of perinatal asphyxia.

PURPOSE: To study the validity of urinary uric acid (UA) as a marker of perinatal asphyxia in term and premature infants. METHODS: The urinary ratio of UA to creatinine (Cr) was obtained within 24 hours after birth in four groups of infants: 17 term infants and 18 premature infants with perinatal asphyxia, and 22 healthy term infants and 20 premature infants without perinatal asphyxia. Perinatal asphyxia was defined as an Apgar score of 3 or less at 1 minute or 5 or less at 5 minutes, and/or a first blood gas pH of less than 7.25 and a base deficit of at least 12 mmol/L. RESULTS: The urinary ratio of UA to Cr was significantly higher in term infants with perinatal asphyxia than in term infants without asphyxia (1.53 +/- 0.71 vs 0.73 +/- 0.45; p < 0.005). The same result was found between premature infants with and without perinatal asphyxia (3.89 +/- 1.84 vs 2.45 +/- 0.88; p < 0.01). The urinary ratio of UA to Cr in premature infants was significantly higher than in term infants. When the urinary ratio of UA to Cr was greater than 0.95, perinatal asphyxia was identified with a sensitivity of 80% and a specificity of 71% in term infants. In premature infants, a cut-off value of UA/Cr for perinatal asphyxia of 2.9 had a sensitivity of 71% and a specificity of 70%. CONCLUSIONS: The results of this study indicate that the urinary ratio of UA to Cr may be used as an additional marker of perinatal asphyxia in term and premature infants. In comparison with other markers such as xanthine, hypoxanthine, and ascorbic acid, it is a simple, quick, and inexpensive way to detect hypoxic episodes in a neonatal intensive care unit within 24 hours after birth.     

Increased peripheral chemoreceptor activity may be critical in destabilizing breathing in neonates

Periodic breathing and apnea are common in neonates, yet the physiological mechanisms involved are not clear. A low arterial PO2 might magnify peripheral chemoreceptor contribution to breathing, with its baseline variability inducing major changes in ventilation, leading to instability of the respiratory control system. We hypothesized that neonates: (1) would depend much more on the peripheral chemoreceptor contribution to breathing than adult subjects and (2) their baseline arterial PO2 would sit on the steep portion of the ventilation/arterial PO2 relationship on the adult nomogram, making breathing prone to oscillate. We analyzed data from previous polygraphic recordings in four groups of subjects: small preterm infants [SPI; postconceptional age (PCA) 33+/-2 weeks; n = 40], large preterm infants (LPI; PCA 36+/-2 weeks; n = 34), term infants (TI; PCA 42+/-1 week; n = 24), and adult subjects (AS; weight 63+/-2 kg; age 29+/-3 years, n = 16). Peripheral chemoreceptor activity was measured by: (1) the immediate decrease in ventilation and (2) apnea time during brief inhalation of 100% O2 (about 1 minute). We found that: (1) the immediate decrease in ventilation with 100% O2 was more pronounced in infants than in adult subjects (38+/-2 versus 6+/-5%), and in infants breathing periodically versus those breathing continuously; (2) the apnea time during 100% O2 was also significantly longer in periodic breathing infants; and (3) the TcPO2 was much lower in infants than in adult subjects (65+/-1 versus 93+/-1 Torr), and also lower in periodic versus continuously breathing infants. It was located significantly to the left of values for the adult subject, on the ventilation/arterial PO2 diagram. The data suggest that: (1) a substantial portion of baseline breathing activity early in life is maintained by increased peripheral chemoreceptor activity; and (2) neonates breathe irregularly with apneas due to the position of their arterial PO2 values on the ventilation/arterial PO2 diagram, in which a change in PO2 produces a more significant change in ventilation than that observed later in life.     

Hypoxic-ischemic encephalopathy: cerebrovascular carbon dioxide reactivity in neonates

Using noninvasive measurement of cranial blood flow, we previously demonstrated that full-term asphyxiated neonates have decreased cerebral perfusion that can persist up to 5 days of age. In an attempt to test their postischemic cerebrovascular CO2 reactivity, we measured cranial blood flow in ten asphyxiated term (39 +/- 0.8 weeks and 3078 K 400 gm) infants with and without inhaled carbon dioxide (3 percent). The end tidal CO2 (PaCO2) increased significantly, from 28.8 +/- 1.0 mm Hg to 32.3 +/- 2.0 mm Hg after CO2 inhalation (p less than 0.01), whereas the cranial blood flow showed no significant change (38.5 +/- 5.0 ml/min/100 gm brain weight to 37.6 +/- 6.0 ml/min/100 gm brain weight). We conclude that term infants with hypoxic-ischemic encephalopathy have low cranial blood flow at 3 days of age. Their cerebrovascular response to inhaled CO2 is variable and suggests some impairment.