Temporal inversion of the acid-base equilibrium in newborns: an observational study

Background A considerable fraction of newborn infants experience hypoxia-ischaemia and metabolic acidosis at birth. However, little is known regarding the biological response of newborn infants to the pH drift from the physiological equilibrium. The aim of this study was to investigate the relationship between the pH drift at birth and postnatal acid-base regulation in newborn infants. Methods Clinical information of 200 spontaneously breathing newborn infants hospitalised at a neonatal intensive care centre were reviewed. Clinical variables associated with venous blood pH on days 5–7 were assessed. Results The higher blood pH on days 5–7 were explained by lower cord blood pH (−0.131, −0.210 to −0.052; regression coefficient, 95% confidence interval), greater gestational age (0.004, 0.002 to 0.005) and lower partial pressure of carbon dioxide on days 5–7 (−0.005, −0.006 to −0.004) (adjusted for sex, postnatal age and lactate on days 5–7). Conclusion In relatively stable newborn infants, blood pH drift from the physiological equilibrium at birth might trigger a system, which reverts and over-corrects blood pH within the first week of life. Given that the infants within the study cohort was spontaneously breathing, the observed phenomenon might be a common reaction of newborn infants to pH changes at birth.


43
The blood pH equilibrium is persistently controlled within a narrow range via innate buffers 44 and respiratory and renal systems, and is essential in maintaining regular metabolism in human 45 organs (Pocock G et al. 2017). These systems also play important roles when the body 46 experiences an acute derangement of the pH homoeostasis. Adults may encounter severe acute 47 acidosis on limited occasions at critical events (Jung et al. 2011). In contrast, virtually all 48 newborn infants experience some form of hypoxia-ischaemia and acidosis due to disruption of 49 the placental oxygen supply before the establishment of spontaneous breathing (Gleason CA & 50 Juul SE 2018). Severe birth asphyxia causes critical energy depletion and profound 51 acidosis, (Novak et al. 2018) ultimately leading to critical events, such as hypoxic-ischaemic 52 encephalopathy (Douglas-Escobar & Weiss 2015). Relatively less severe perinatal stress may 53 also cause serious neurological consequences. Periventricular leukomalacia is a milder form of 54 cerebral injury in preterm infants, (Stoll et al. 2015) the incidence of which increases with the 55 presence of both spontaneous and induced hypocarbia (Okumura et al. 2001;Stenzel et al. 2020). 56 A study from a large-scale cohort suggested that newborn infants who transiently required 57 respiratory support at birth, but did not require hospitalisation, are at increased risk of developing 58 adverse neurodevelopmental outcomes (Odd et al. 2009). Studies in vivo reported increased 59 fraction of apoptotic neuronal death when cultured neurons were transiently exposed to hypoxic-60 ischaemic environment, and then, to alkalotic environment with sufficient oxygen and energy 61 substrates (Robertson et al. 2013;Vornov et al. 1996). Although these studies highlight the 62 importance of understanding the biological response to the covert disruption of the pH 63 homeostasis at birth, there currently is limited knowledge regarding foetal and neonatal reactions 64 to the pH drift. The elucidation of such reactions and their association with the injury cascade 65 may contribute to the development of novel strategies for the early screening, diagnosis and 66 treatment of transition failure at birth and subsequent neurodevelopmental impairments.

68
We performed a retrospective observational study to test a hypothesis that the pH equilibrium at 69 birth influences acid-base regulation within the first week of life in spontaneously breathing 70 newborn infants.  In this unit, routine blood gas analysis is performed (i) at birth (umbilical venous samples), (ii) 89 on day 0 (venous samples typically obtained shortly after admission), on day 3 (venous/capillary 90 samples) and between days 5 and 7 (venous/capillary samples obtained at the time of the 91 newborn screening test). Blood gas analysis on days 5-7 is not performed for infants who have 92 been discharged home or moved to the step-down unit. The blood pH; partial pressures of carbon 93 dioxide (pCO 2 ) and oxygen; bicarbonate (HCO 3 -); sodium; potassium; calcium; chloride; lactate; 94 glucose; total, carboxyl and foetal haemoglobin; and total bilirubin levels are simultaneously 95 measured (ABL800; Radiometer, Copenhagen, Denmark). However, for cord blood samples, 96 only the pH was available from the electronic record. Blood gas data from all routine and 97 additional blood tests performed within 24 hours of birth and between 5 and 7 days of birth were 98 incorporated.  To highlight the selection bias, clinical backgrounds were compared between the infants within 108 the final study cohort and those, who were excluded, using the Student's t-test, Mann-Whitney's 109 U test or chi-square test. To assess temporal changes in blood pH after birth, relationships 110 between cord blood pH and blood pH on days 5-7 were assessed adjusting for clinical 111 backgrounds and priori covariates; blood samples obtained later than day 7 were not considered 112 because of the selection bias derived from non-routine sampling of these samples. A generalised 113 estimating equation was used to correct for repeated sampling and postnatal age at blood 114 sampling (SPSS version 25; IBM, Armonk, NY, USA). Findings from the univariate analysis 115 were not corrected for multiple comparisons; however, p-values between 0.01 and 0.05 were 116 regarded as "chance level". To evaluate the potential influence of respiratory support on the 117 relationship between cord blood pH and blood pH on days 5-7, the univariate analysis was 118 repeated in (i) a restrictive population of infants, who never experienced invasive respiratory 119 support (n = 157) and (ii) an expanded population of the final study cohort and those, who 120 remained intubated at the time of blood sampling on days 5-7 (n = 302 in total).

121
A multivariate model to explain the blood pH on days 5-7 was then developed for the final 122 study cohort. First, the relationship between cord blood pH and blood pH on days 5-7 was 123 assessed adjusting for sex and postnatal age. The influence of gestational age at birth and pCO 2 124 and lactate levels on days 5-7 was assessed using the forward stepwise selection algorithm.   (Table 1 and  157   158 The first multivariate model (Model 1) was adjusted for the postnatal age and sex, which 159 showed a negative relationship between the cord blood pH and blood pH on days 5-7 (p < 0.001, 160 Table 3). Model 2 included the variables in Model 1 and gestational age, whereas Model 3 161 included the variables in Model 2 plus pCO 2 on days 5-7, both of which supported the 162 relationship between the cord blood pH and blood pH on days 5-7 (p = 0.005 and p < 0.001, 163 respectively; Table 3). The final model (Model 4) was adjusted for the same variables as Model 164 3 plus the lactate level on days 5-7, where lower cord blood pH, greater gestational age and 165 lower pCO 2 levels on days 5-7 (regression coefficient: -0.131, 0.004 and -0.005; 95% confidence 166 interval : -0.210 to -0.052, 0.002 to 0.005 and -0.006 to -0.004; and p = 0.001, p < 0.001 and p < 167 0.001, respectively) explained the higher venous blood pH on days 5-7 (Table 3). To identify the 168 intermediate variable explaining the relationship between the cord blood pH and blood pH on 169 days 5-7, Model 5 was additionally adjusted for the covariates in Model 4 plus HCO 3on days 5-170 7, where the role of cord blood pH as an independent variable for blood pH on days 5-7 was 171 replaced by HCO 3on days 5-7 (p < 0.001, Table 3) 172 173

174
In spontaneously breathing newborn infants, a higher blood pH on days 5-7 was paradoxically 175 related to a lower cord blood pH. Considering that the pH equilibrium of these infants was likely 176 to be determined by their own spontaneous regulation, this temporal inversion of the blood pH 177 homeostasis within the first week of life might represent a physiological response in newborn 178 infants to the drift in the pH equilibrium at birth.

180
In our current study, higher blood pH levels on days 5-7 were best explained by the cord blood 181 pH, as well as gestational age, pCO 2 and lactate levels of the same blood sample obtained on 182 days 5-7. Considering that the role of the cord blood pH as an independent variable for the blood 183 pH on days 5-7 was superseded by HCO 3levels on days 5-7, it would be relevant to speculate 184 that the pH drift in the cord blood towards an acidic (alkalotic) equilibrium triggered the active 185 accumulation (elimination) of blood HCO 3to temporally invert the acid-base homeostasis 186 within the first week of life. Although our study cohort did not include those with severe birth 187 asphyxia or excessive immaturity and all of the infants who were included had been weaned 188 from mechanical ventilation by the time of blood gas analysis on days 5-7, even transient 189 resuscitation and mechanical ventilation may well influence the acid-base regulation thereafter.
190 However, the relationship between cord blood pH and blood pH on days 5-7 was consistently 191 observed in both restrictive (those who never experienced invasive respiratory support) and 192 expanded (the original cohort plus those who remained intubated at the time of blood sampling 193 on days 5-7) study populations, suggesting that the influence of the respiratory support on the 194 temporal change of the blood pH may be limited. Provided that the blood pH equilibrium of 195 these infants was determined as a consequence of spontaneous acid-base control, the negative 196 relationship between the cord blood pH and the blood pH on days 5-7 might be a physiological 197 response in newborn infants.

199
In foetuses, acidosis is compensated for by innate buffers, such as bicarbonate and 200 haemoglobin, which help eliminate carbon dioxide via the placenta (Blechner 1993). After birth, 201 the role of the placenta in maintaining the acid-base homeostasis is replaced by the lung and 202 proximal tubule of the kidney, which, together with the innate buffering system, contribute to 1