Exploratory analysis of high-altitude effects on brain morphology and myelination in native Tibetan infants

Introduction

Background

High altitude introduces environmental stressors distinct from those at sea level. The primary adverse environmental factor at high altitude is the reduction in the partial pressure of oxygen, although it is also associated with low temperature, aridity, and ultraviolet radiation (1). Since the human brain consumes about 20% of the body’s oxygen intake (2), high-altitude environments pose unique challenges to its adaptability. Aside from direct effects, physiological adaptations to long-term high-altitude exposure, such as altered circulatory and respiratory function, can lead to cumulative changes in the brain through afferent feedback (3). There is mounting evidence that long-term high-altitude residents tend to suffer cognitive, memory, mood, and behavior impairments (3-5), suggesting chronic damage to brain structures.

Rationale and knowledge gap

Magnetic resonance imaging (MRI) is a widely-used, non-invasive neuroimaging technique for visualizing and analyzing brain structure (6-12) and myelination (13,14). Studies in healthy populations have shown brain structure alterations following exposure to high-altitude environments in both high-altitude immigrants (9,15-17) and native populations (18-20). Tibetan and Andean populations have inhabited high-altitude regions for more than 25,000 and 12,000 years, respectively, and, through natural selection, appear well adapted to these environments, exhibiting more refined brain modulation. Notably, high-altitude natives show better hematological adaptability, increased gray matter volume (21,22), and increased neural activity (22) compared with high-altitude immigrants. Early-life exposure to high altitudes may affect neonatal brain development; however, the structural and myelination characteristics of neurologically normal Tibetan neonates remain insufficiently understood. The exploration of brain alterations in high-altitude environments could provide insights into the clinical treatment of hypoxic diseases.

Objective

To address these gaps in the literature, we conducted a multicenter, case-control exploratory study of 30 neurologically normal infants from a high-altitude region (3,650 m) and 30 matched infants from a sea-level region (4 m). The participants in the groups were matched 1:1 using propensity score matching (PSM) for gestational age at birth, birth weight, sex, and postnatal age at MRI examination. We aimed to compare volumetric differences in intracranial volume (ICV) and segmented brain tissues (gray matter, white matter, and lateral ventricles) between the two groups, and among altitude subgroups, as well as to assess myelination status using T1-weighted/T2-weighted (T1w/T2w) ratios. Our findings provide insights into the effects of high-altitude exposure on neonatal brain structure and myelination. We present this article in accordance with the STROBE reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-778/rc).

Methods

Study design and population

This was a multicenter, case-control, exploratory study. Infants were recruited from two neonatal intensive care units (NICUs): the NICU at the People’s Hospital of Xizang Autonomous Region in Lhasa, China (high-altitude group; altitude: 3,650 m), and the NICU at the Children’s Hospital of Fudan University in Shanghai, China (sea-level group; altitude: 4 m). The inclusion criteria for the high-altitude group were as follows: (I) admission between June 2020 and April 2022, with a cranial MRI examination performed during hospitalization; (II) both parents of native Tibetan ethnicity; (III) no chromosomal anomalies or major malformations; (IV) no history of neurological disorders or relevant symptoms including seizures, syncope, or abnormal muscle tone; (V) normal findings on cranial MRI reports; and (VI) discharge following complete recovery. For the sea-level group, standard inclusion criteria were applied (i.e., admission during the study period and performance of a cranial MRI examination) without any ethnic restrictions. For both the high-altitude and sea-level groups, infants who required chest compressions or epinephrine in the delivery room, or who were diagnosed with sepsis, bronchopulmonary dysplasia, or severe anemia during hospitalization, were excluded from the study to minimize the impact of additional clinical conditions. Additionally, infants with poor-quality cranial MRI images were also excluded from the study. For the sea-level group (altitude: 4 m), infants were selected from a large clinical database of the NICU at the Children’s Hospital of Fudan University in Shanghai.

Because the original dataset was extensive and some baseline variables required manual supplementation, it was impractical to include all potential cases. Therefore, we identified 90 consecutive infants who underwent brain MRI during the same study period (June 2020–April 2022) in reverse chronological order. After image quality evaluation, 79 infants with adequate MRI quality and complete clinical information were retained for analysis and served as the candidate pool for matching. Propensity scores were calculated based on gestational age at birth, birth weight, sex, and postnatal age at MRI examination. A 1:1 nearest-neighbor matching without replacement, with a caliper of 0.2 standard deviations (SDs) of the logit of the propensity score, was performed to select controls for the high-altitude infants.

Definitions

There is no universal indication for MRI scanning in preterm infants in China. However, most hospitals routinely perform predischarge or full-term MRI scanning in all preterm infants who survive (23). Gestational age at birth was calculated using the best obstetric estimate based on early prenatal ultrasonography, the last menstrual period, obstetric examination, or a combination of all three. Bronchopulmonary dysplasia was defined as the need for supplemental oxygen for at least 28 days, and was classified as moderate or severe depending on the oxygen requirements and ventilator support at 36 weeks postmenstrual age (PMA) (24). Sex was designated at birth, based solely on the visible external anatomy of the newborn.

Exposures

The People’s Hospital of Xizang Autonomous Region is located in Lhasa, China, at an altitude of 3,650 m. However, the altitudes of the maternal permanent residence areas vary, ranging from 3,500 m (Zhag’yab County, Chamdo Prefecture, China) to 4,800 m (Namco Township, Damxung County, China). The high-altitude group was further divided into two subgroups: high-altitude subgroup 1 (3,500 to <4,000 m) and high-altitude subgroup 2 (4,000 to 4,800 m). High-altitude subgroup 1 comprised 14 infants, while high-altitude subgroup 2 comprised 16 infants.

Head circumference measurement by MRI images

Head circumference was measured using MRI images (25). Briefly, two readers (B.Z. and W.C.) independently reviewed the images using ITK-SNAP (version 4.0.2). An ovoid region of interest (ROI) was placed on an axial plane, beginning at the supraorbital bulge and covering the largest supra-auricular head circumference. For asymmetric head shapes, the raters were advised to place the ROI as accurately as possible to best approximate the outer circumference (Figure S1). Interrater reliability was assessed using the intraclass correlation coefficient (ICC) based on a two-way random-effects model (26). Interrater agreement was excellent with an ICC of 0.987 [95% confidence interval (CI): 0.978–0.992] between the two independent raters for the head circumference measurement. Subsequently, the average value of the two measurements was calculated for further analysis.

MRI data collection

Structural T1w and T2w MRI scans were retrospectively acquired at each study site as part of routine clinical care, with sedation using rectal chloral hydrate administered when clinically indicated. As this was a retrospective multicenter study, the MRI acquisition parameters followed site-specific clinical protocols. The MRI data were acquired using a Skyra 3.0T scanner with a 64-channel head coil (Siemens) at the People’s Hospital of Xizang Autonomous Region, and Avanto 1.5T (n=9) or 3.0T (n=21) scanners with a 64-channel head coil (Siemens) at the Children’s Hospital of Fudan University. The T1w images were acquired using standard spin-echo or gradient-echo–based sequences, and the T2w images were obtained using turbo spin-echo–based sequences according to site-specific clinical protocols. Representative acquisition parameters included a repetition time of 250–500 ms and an echo time of 2–12 ms for the T1w images, and a repetition time of 4,000–6,500 ms and an echo time of 90–130 ms for the T2w images. The slice thickness ranged from 3.5–5 mm. MRI image quality was evaluated by visual inspection by an imaging scientist (Y.Z.) and neonatologist (B.Z.). Infants with poor-quality cranial MRI images, characterized by incomplete scan coverage, artifacts, excessive noise, signal dropouts, spatial distortion, and anatomical anomalies impeding brain segmentation, were excluded from the study.

Although both groups used the same brand of MRI device (Siemens), the models differed across sites. To standardize the imaging protocols and ensure comparability, several measures were implemented. First, we used the uAI Research Portal (Shanghai United Imaging Intelligence Co. Ltd., Shanghai, China) (27), an image analysis tool that has been shown to be effective in volumetric analysis across various MRI models (28). For myelination assessment, our analysis was restricted to 3.0T MRI data to minimize inconsistencies in T1w and T2w intensity values that can arise from differences in scanner models (29). Additionally, N5 bias field correction and normalization procedures were applied to reduce inter-device variations. To further assess the potential influence of different MRI device models, a mixed-effects linear model was used, in which the device model was treated as a random effect, and PMA, sex, and birthweight were included as fixed effects. The results showed negligible variance in the random effect, indicating that differences in the MRI device models had minimal effects on volumetric measurements and myelination status.

Volumetric measurement

In this study, T1w images were used for manual volumetric segmentation. Volumetric measurements were preprocessed using the uAI Research Portal (Shanghai United Imaging Intelligence Co., Ltd.), an image analysis tool (27). Briefly, the preprocessing included skull stripping, bias correction, and image resampling to 1-mm isotropic resolution. After preprocessing, the images were manually segmented into white matter, gray matter, and lateral ventricles (by B.Z.) (Figure 1A). To ensure the accuracy of the methodology, an experienced radiologist (J.L., a senior pediatric radiologist with more than 10 years of experience) reviewed the methodology slice by slice. If any discrepancies were found, the segmentation process was repeated and rechecked to ensure accuracy. The volume of each structure was analyzed, along with the ICV, which was defined as the total cerebrum volume after skull stripping, excluding the cerebellum.

Figure 1 Schematic overview of regional segmentation (A) and myelination assessment (B). ICBM, International Consortium for Brain Mapping; T1w, T1-weighted; T2w, T2-weighted.

Myelination assessment

There is growing evidence that the T1w/T2w ratio can be used to investigate spatial variation in myelin formation in the neonatal brain (13,14,30,31). This analysis was also performed using the uAI Research Portal (Shanghai United Imaging Intelligence Co. Ltd.) (27). In practice, the N5 bias field correction algorithm is often used in conjunction with normalization processing to maximize the reduction of device-to-device differences. Initially, the N5 algorithm was applied to remove the bias field, achieving a more homogeneous intensity distribution. Subsequently, normalization processing was conducted to scale the intensity values to a standardized range, ensuring that the images obtained from different devices or scanning conditions were comparable.

These preprocessing methods can significantly enhance the accuracy and reliability of medical image analysis, particularly in multicenter studies or when comparing images acquired on different devices. The process of generating the T1w/T2w ratio map involved several key steps, including the co-registration of T1w and T2w images, bias field correction, and intensity normalization. The intensity normalization was performed using a linear scaling procedure as described previously (29). Fifty ROIs were defined using the International Consortium for Brain Mapping (ICBM) Deep Nuclei probabilistic atlas (http://www.loni.usc.edu/ICBM). Among which, the following eight ROIs exhibiting myelination at 40 weeks’ PMA (32-34) were selected for comparison: the middle cerebellar peduncle (MCP), the pontine crossing tract (PCT, a part of the MCP), the left and right inferior cerebellar peduncles (ICP-L and ICP-R), the left and right superior cerebellar peduncles (SCP-L and SCP-R), and the left and right posterior limbs of the internal capsule (PLIC-L and PLIC-R). The individual ROIs of ICBM were acquired using the template T2w images and co-registered to each individual’s T2w images. Based on these translated ROIs of ICBM, the T1w/T2w signal ratio was calculated for brain myelination assessment (14) (Figure 1B).

Outcomes

The primary outcomes were differences in the manually segmented brain region volumes and ICV between the high-altitude and sea-level groups. The secondary outcomes were differences in the manually segmented brain region volumes among the high-altitude subgroups and the sea-level group. The T1w/T2w mapping technique was applied for further assessment of brain myelination.

Statistical analysis

Descriptive analyses were performed to examine the demographic characteristics of the total study population. Differences between the control and high-altitude groups were assessed using the chi-squared test for the categorical variables, and either the student t-test or Wilcoxon rank-sum test for the continuous variables as appropriate.

Linear models were used for further volumetric or myelination comparisons. The confounders adjusted in the multivariable model included gestational age at birth, infant sex, birth weight, age at MRI scan, and PMA at MRI scan. Due to the collinearity between PMA, gestational age at birth and postnatal age at MRI scan, two linear models were employed. In linear regression model 1, gestational age at birth, infant sex, birth weight, postnatal age at MRI scan, and altitude grouping were included as the independent variables, while in linear regression model 2, infant sex, birth weight, PMA at MRI scan, and altitude grouping were included as the independent variables. We further optimized two models based on distinct criteria: adjusted R-squared and the Mallows Cp statistic. This optimization process aimed to enhance the models’ predictive performance and reduce complexity. Following this optimization, we conducted an analysis of variance to identify the optimized models. All of the regression models were checked for linearity, homoscedasticity, and absence of multicollinearity, and the residuals approximated a normal distribution. β coefficients, 95% CIs, and P values were reported. The analyses were performed using R4.2.3 software. G*Power (version 3.1.9.6) was used to assess the statistical power of our study given the sample size and effect size. A post-hoc power analysis was conducted for the independent samples t-test with the type I error set at 0.1 and a two-tailed test. P values <0.05 were considered statistically significant. Bonferroni correction was used for multiple comparisons.

Sensitivity analysis

We conducted sensitivity analyses on the primary results using two distinct subsets selected from the overall population: infants who were preterm and underwent MRI examination within 56 days after birth, and infants who underwent MRI examination within 41 weeks of PMA. These analyses aimed to minimize the effects of gestational age and postnatal interventions.

Ethics

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study procedures were reviewed and approved by both The People’s Hospital of Xizang Autonomous Region Institutional Review Board (No. ME-TBHP-24-KJ-035) and The Children’s Hospital of Fudan University Institutional Review Board (No. 2021-6-25). The requirement for informed parental consent was waived due to the retrospective nature of the study.

Results

Demographic characteristics

The entire cohort comprised 60 infants: 30 Tibetan infants in the high-altitude group (Figure S2) and 30 matched infants in the sea-level group. The descriptive characteristics of both groups before matching are summarized in Table S1, while the descriptive characteristics of the included infants after matching are presented in Table 1. Briefly, the mean gestational age at birth of the neonates was 34.69 (SD 3.02) weeks. Because full-term infants generally undergo brain MRI only when neurological disorders are present, while preterm infants routinely undergo brain MRI at a PMA of 37 weeks, 76.7% of the study population were preterm infants. The mean PMA at MRI examination was 37.86 (SD 2.58) weeks, and the mean postnatal day at MRI examination was 22.20 (SD 16.55) days. There were no significant differences in the demographic characteristics between the two groups (Table 1).

Comparison of manually segmented brain regions

Across all three models, ICV was significantly larger in the high-altitude group compared to the sea-level group [model 1: β coefficient (95% CI): 27,851.35 (12,923.06–42,779.64), P value <0.001; model 2: β coefficient (95% CI): 27,595.67 (12,123.51–43,067.84), P value <0.001 and adjusted model: β coefficient (95% CI): 23,728.00 (8,000.96–39,455.04), P value =0.004, respectively]. Near-significant differences in lateral ventricle and gray matter volumes were observed. However, in the white matter volume comparison, no significant difference was detected between the different altitude groups (Figure 2). The power analysis for the comparison between the two groups showed a value of 0.600 for ICV, 0.631 for lateral ventricle volume, 0.301 for grey matter volume, and 0.221 for white matter volume.

Figure 2 Comparison of manually segmented brain volumes between high-altitude and sea-level groups. Model 1 was adjusted for gestational age at birth, infant sex, birth weight, and postnatal age at MRI scan. Model 2 was adjusted for infant sex, birth weight, and PMA at MRI scan. The adjusted model denotes the optimized model selected based on adjusted R² and Mallows’ Cp criteria. CI, confidence interval; HA, high-altitude group; MRI, magnetic resonance imaging; PMA, postmenstrual age; SD, standard deviation; SL, sea-level group.

Comparison of manually segmented brain volume among the sea-level group and high-altitude subgroups

The high-altitude group was subdivided into high-altitude subgroup 1 (3,500 to <4,000 m) and high-altitude subgroup 2 (4,000 to 4,800 m). Subgroup 1 comprised 14 infants, while Subgroup 2 comprised 16 infants. No significant differences were observed in the demographic characteristics among the three groups (Table S2). The ICV in altitude subgroup 2 (4,000 to 4,800 m) was significantly larger than that in the control group [model 1: β coefficient (95% CI): 35,969.00 (18,428.96–53,509.04), P value <0.001; model 2: β coefficient (95% CI): 36,027.00 (17,842.12–54,211.88), P value <0.001 and adjusted model: β coefficient (95% CI): 31,806.00 (13,172.28–50,439.72), P value =0.001, respectively], and the differences in white matter volume and grey matter volume were near significant (Figure 3). However, there were no significant differences detected in ICV, lateral ventricle volume, white matter volume, or grey matter volume between altitude subgroup 1 (3,500 to <4,000 m) and the control group (Figure 3).

Figure 3 Comparison of manually segmented brain volumes among altitude subgroups. Model 1 was adjusted for gestational age at birth, infant sex, birth weight, and postnatal age at MRI scan. Model 2 was adjusted for infant sex, birth weight, and PMA at MRI scan. The adjusted model denotes the optimized model selected based on adjusted R² and Mallows’ Cp criteria. CI, confidence interval; MRI, magnetic resonance imaging; PMA, postmenstrual age; SL, sea-level group.

Sensitivity analysis

Given that 76.7% of the included infants were preterm, with a wide range of postnatal day at MRI examination from day 1 to day 86 after birth, sensitivity analyses were performed using two subsets: infants who were preterm and underwent MRI examination within 56 days after birth, and infants who underwent MRI examination within 41 weeks of PMA. The comparison of manually segmented brain regions and ICV in these subsets revealed consistent findings with those of all the included infants. Specifically, the ICV and lateral ventricle volumes were significantly larger (Table S3).

Myelination assessment

Myelination was assessed in 51 infants with 3.0T MRI data. Of these, 41 infants (23 in the high-altitude group and 18 in the sea-level group) successfully completed the co-registration process, allowing for the calculation of the T1w/T2w ratio for each ROI. The gestational age at birth in weeks was 36.09 (SD 2.99) for the sea-level group and 34.85 (SD 2.78) for the high-altitude group. The PMA at examination was 39.00 (SD 3.43) weeks for the sea-level group and 38.13 (SD 2.13) weeks for the high-altitude group. No significant differences were observed in the demographic characteristics between the two groups. The T1w/T2w ratio in the MCP was significantly lower in the high-altitude group in two of the three models (model 1: β coefficient (95% CI): –0.08 (–0.15, –0.01), P value =0.03; model 2: β coefficient (95% CI): –0.08 (–0.15, 0), P value = 0.04 and adjusted model: β coefficient (95% CI): –0.08 (–0.15, 0), P value =0.054, respectively), suggesting that myelination in the MCP was delayed compared to the sea-level group. The power analysis indicated a value of 0.839. The T1w/T2w ratio in other ROIs tended to be lower the in high-altitude group; however, the differences did not reach statistical significance (Table S4 and Figure 4).

Figure 4 Comparison of T1w/T2w signal intensity ratio between the high-altitude and sea-level groups. Model 1 was adjusted for gestational age at birth, infant sex, birth weight, and postnatal age at MRI scan. Model 2 was adjusted for infant sex, birth weight, and PMA at MRI scan. The adjusted model denotes the optimized model selected based on adjusted R² and Mallows’ Cp criteria. Colored overlays indicate the anatomical locations of the predefined white matter tracts and are shown for visualization purposes only: MCP, PCT (a part of the MCP), ICP-L/R, SCP-L/R, and PLIC-L/R. CI, confidence interval; ICP-L, inferior cerebellar peduncle left; ICP-R, inferior cerebellar peduncle right; MCP, middle cerebellar peduncle; MRI, magnetic resonance imaging; PMA, postmenstrual age; PCT, pontine crossing tract; PLIC-L, posterior limb of internal capsule left; PLIC-R, posterior limb of internal capsule right; SCP-L, superior cerebellar peduncle left; SCP-R, superior cerebellar peduncle right; T1w, T1-weighted; T2w, T2-weighted.

Discussion

To the best of our knowledge, this was the first study to examine brain structure and myelination in Tibetan neonates. We found that across all three models, ICV was significantly increased in the high-altitude group, with near-significant increases in the gray matter and lateral ventricle volumes. In the subgroup analysis, ICV was significantly increased in altitude subgroup 2 (altitudes ranging from 4,000 to 4,800 m) but not in altitude subgroup 1 (altitudes ranging from 3,500 to 4,000 m). While there was no difference in white matter volume, the T1w/T2w ratio in the MCP was significantly lower in the high-altitude group.

It is widely accepted that prolonged exposure to hypoxic environments adversely affects brain development (35), with the extent of the effect depending on the severity, duration, and timing of the exposure. In adults, studies have reported heterogeneous effects of long-term high-altitude exposure on brain structure (36), likely due to the combined analysis of immigrants and native populations (37), or the inclusion of a wide range of altitude levels (19). Because dynamic neurodevelopmental processes, including the consolidation of thalamocortical connections and the continuous development of axons (38,39), occur during the early neonatal period, drawing conclusions in the neonatal population is particularly challenging, and is further compounded by the increased susceptibility of neonates to environmental influences.

In the current study, the high-altitude group, particularly altitude subgroup 2 (altitudes ranging from 4,000 to 4,800 m), exhibited a significantly increased ICV, and a trend toward increased grey matter and lateral ventricle volumes. Previous studies (18-22) have also found increased grey matter volume in high-altitude native adult populations, with some researchers suggesting compensatory processes following exposure to high altitudes. The significant increase in ICV could be attributed to the cumulative differences in gray matter and lateral ventricle volumes. The increased gray matter volume may be associated with neurogenesis, gliogenesis, and vascular changes (40). Blood vessels account for approximately 5% of the volume of gray matter (41). The blood vessels in the developing brain exhibit increased permeability compared to those in the adult brain, and exposure to high-altitude environments might further increase capillary permeability (42). This condition may worsen at altitudes above 4,000 m, which may serve as a threshold of hypoxia severity. Above this threshold, cerebrovascular autoregulation may be impaired (43), even in those well adapted populations like Andean and Tibetan natives (12,44). Given that cerebral blood flow alterations mainly occur in the gray matter (18), and white matter has the lowest density of perfused capillaries (45,46), it appears that vascular alterations may contribute to the regional volume changes. It should also be noted that the genetic and epigenetic backgrounds of high-altitude and sea-level groups may differ (35,47), potentially contributing to the observed volumetric changes independent of environmental hypoxia.

Our analysis revealed no difference in head circumference between the two groups, suggesting potential changes in the extracerebral space. However, due to the thick-slice nature of the MRI data in this study, further segmentation and analysis of the extracerebral space or cerebrospinal fluid were not performed. For ICV and ventricular volumes, the study achieved moderate statistical power (~0.6). While this fell below the conventional threshold of 0.8, it still indicates a moderate chance of detecting differences. However, the low statistical power for gray and white matter volumes suggests that the sample size may have been insufficient to reliably detect differences. Given the exploratory nature of this study, the results should be interpreted with caution, and future studies with larger sample sizes are recommended to improve power.

Since it is generally accepted that chronic hypoxia impairs myelination and synaptogenesis in developing brains (48,49), we further investigated myelination by analyzing the T1w/T2w ratio (13,14,30,31). A significant difference was observed in the MCP, which is one of the areas in which myelination can first be observed (32,33). Our results indicated delayed myelination in high-altitude neonates, consistent with the previously reported lower motor maturity of native high-altitude (4,300 m) neonates (11). However, longitudinal follow-up research needs to be conducted to determine whether this delay is temporary.

To minimize the effect of gestational age and postnatal interventions, we controlled for these confounders using several approaches, and the results remained robust. First, we restricted the population to neurologically normal and non-critically ill infants. Second, the control group was selected using PSM to balance gestational age at birth, sex, birth weight, and age at assessment. Additionally, three linear models were applied for adjustment during comparison. Third, sensitivity analyses of the primary outcomes were conducted in two subsets of data.

However, this study had some limitations. First, the sample size was relatively small. Second, despite the relative healthiness of the newborns included in the study, potential biases related to underlying illnesses and differences in clinical procedures between centers cannot be excluded. Third, the long-term longitudinal data were unavailable, leaving it unclear whether the alterations observed in high-altitude infants may impair long-term neurological function.

Conclusions

In this study, ICV was significantly increased in the high-altitude group compared with the sea-level group, especially in regions above 4,000 m. Near-significant differences were observed in the lateral ventricle and gray matter volumes, while no significant differences were detected in the white matter volume. The T1w/T2w ratio analysis indicated delayed myelination in the MCP of high-altitude infants. These results highlight the need for further research into potential long-term neurological function to determine whether the structural findings represent natural anatomical adaptations or impaired neurodevelopment.

Acknowledgments

The authors would like to thank Dr. Xinran Dong for her assistance with MRI data analysis and Prof. Peng Shi for his valuable support in statistical analysis.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-778/rc

Data Sharing Statement: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-778/dss

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-778/prf

Funding: This research was supported by the National Key Research and Development Program of China (No. 2024YFC2707700); the Regional Science Fund Program of the Natural Science Foundation of China (www.nsfc.gov.cn) (No. 82260316); and the Natural Science Foundation of Xizang Autonomous Region Group Medical Assistance to Xizang Project (sti.xizang.gov.cn) [No. XZ2022ZR-ZY09(Z)].

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-778/coif). W.Z. serves as an Editor-in-Chief of Translational Pediatrics. Q.Z., Z.C., R.H. and F.S. are employees of Shanghai United Imaging Intelligence Co., Ltd. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study procedures were reviewed and approved by both The People’s Hospital of Xizang Autonomous Region Institutional Review Board (No. ME-TBHP-24-KJ-035) and The Children’s Hospital of Fudan University Institutional Review Board (No. 2021-6-25). The requirement for informed parental consent was waived due to the retrospective nature of the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Savioli G, Ceresa IF, Gori G, et al. Pathophysiology and Therapy of High-Altitude Sickness: Practical Approach in Emergency and Critical Care. J Clin Med 2022;11:3937. [Crossref] [PubMed]
2. Raichle ME. Two views of brain function. Trends Cogn Sci 2010;14:180-90. [Crossref] [PubMed]
3. Hoiland RL, Howe CA, Coombs GB, et al. Ventilatory and cerebrovascular regulation and integration at high-altitude. Clin Auton Res 2018;28:423-35. [Crossref] [PubMed]
4. Davis JE, Wagner DR, Garvin N, et al. Cognitive and psychomotor responses to high-altitude exposure in sea level and high-altitude residents of Ecuador. J Physiol Anthropol 2015;34:2. [Crossref] [PubMed]
5. Yan X. Cognitive impairments at high altitudes and adaptation. High Alt Med Biol 2014;15:141-5. [Crossref] [PubMed]
6. Alex AM, Aguate F, Botteron K, et al. A global multicohort study to map subcortical brain development and cognition in infancy and early childhood. Nat Neurosci 2024;27:176-86. [Crossref] [PubMed]
7. Park S, Kim HG, Yang H, et al. A regional brain volume-based age prediction model for neonates and the derived brain maturation index. Eur Radiol 2024;34:3892-902. [Crossref] [PubMed]
8. Triplett RL, Lean RE, Parikh A, et al. Association of Prenatal Exposure to Early-Life Adversity With Neonatal Brain Volumes at Birth. JAMA Netw Open 2022;5:e227045. [Crossref] [PubMed]
9. Chen X, Li H, Zhang Q, et al. Combined fractional anisotropy and subcortical volumetric abnormalities in healthy immigrants to high altitude: A longitudinal study. Hum Brain Mapp 2019;40:4202-12. [Crossref] [PubMed]
10. Rimoldi SF, Rexhaj E, Duplain H, et al. Acute and Chronic Altitude-Induced Cognitive Dysfunction in Children and Adolescents. J Pediatr 2016;169:238-43. [Crossref] [PubMed]
11. Saco-Pollitt C. Birth in the Peruvian Andes: physical and behavioral consequences in the neonate. Child Dev 1981;52:839-46.
12. Virués-Ortega J, Bucks R, Kirkham FJ, et al. Changing patterns of neuropsychological functioning in children living at high altitude above and below 4000 m: a report from the Bolivian Children Living at Altitude (BoCLA) study. Dev Sci 2011;14:1185-93. [Crossref] [PubMed]
13. Filimonova E, Amelina E, Sazonova A, et al. Assessment of normal myelination in infants and young children using the T1w/T2w mapping technique. Front Neurosci 2023;17:1102691. [Crossref] [PubMed]
14. Lee J, Kim Y, Lee J, et al. Altered cortical myelination based on gray-to-white matter signal intensity contrast in shift workers. Brain Struct Funct 2025;230:179. [Crossref] [PubMed]
15. Yan X, Zhang J, Shi J, et al. Cerebral and functional adaptation with chronic hypoxia exposure: a multi-modal MRI study. Brain Res 2010;1348:21-9. [Crossref] [PubMed]
16. Zhang J, Zhang H, Li J, et al. Adaptive modulation of adult brain gray and white matter to high altitude: structural MRI studies. PLoS One 2013;8:e68621. [Crossref] [PubMed]
17. Zhang J, Yan X, Shi J, et al. Structural modifications of the brain in acclimatization to high-altitude. PLoS One 2010;5:e11449. [Crossref] [PubMed]
18. Wang X, Wei W, Yuan F, et al. Regional cerebral blood flow in natives at high altitude: An arterial spin labeled MRI study. J Magn Reson Imaging 2018; Epub ahead of print. [Crossref]
19. Wei W, Wang X, Gong Q, et al. Cortical Thickness of Native Tibetans in the Qinghai-Tibetan Plateau. AJNR Am J Neuroradiol 2017;38:553-60. [Crossref] [PubMed]
20. Zhong M, Zeng H, Wang D, et al. Structure and activity alteration in adult highland residents’ cerebrum: Voxel-based morphometry and amplitude of low-frequency fluctuation study. Front Neurosci 2022;16:1035308. [Crossref] [PubMed]
21. Zhang X, Xie W, Du W, et al. Consistent differences in brain structure and functional connectivity in high-altitude native Tibetans and immigrants. Brain Imaging Behav 2023;17:271-81. [Crossref] [PubMed]
22. Zhang X, Xie W, Liu Y, et al. Brain Structural and Functional Alterations in Native Tibetans Living at High Altitude. Neuroscience 2023;520:134-43. [Crossref] [PubMed]
23. Bai R, Jiang S, Guo J, et al. Variation of Neonatal Outcomes and Care Practices for Preterm Infants <34 Weeks’ Gestation in Different Regions of China: A Cohort Study. Front Pediatr 2021;9:760646. [Crossref] [PubMed]
24. Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med 2001;163:1723-9. [Crossref] [PubMed]
25. Rau A, Demerath T, Kremers N, et al. Measuring the Head Circumference on MRI in Children: an Interrater Study. Clin Neuroradiol 2021;31:1021-7. [Crossref] [PubMed]
26. Wirtz M, Caspar F. Beurteilerübereinstimmung und Beurteilerreliabilität: Methoden zur Bestimmung und Verbesserung der Zuverlässigkeit von Einschätzungen mittels Kategoriensystemen und Ratingskalen. Göttingen: Hogrefe Verlag für Psychologie; 2002. 287 p.
27. Wu J, Xia Y, Wang X, et al. uRP: An integrated research platform for one-stop analysis of medical images. Front Radiol 2023;3:1153784. [Crossref] [PubMed]
28. Teng L, He Y, Cao Z, et al. Unified Model for Children's Brain Image Segmentation With Co-Registration Framework Guided by Longitudinal MRI. IEEE J Biomed Health Inform 2025;29:4623-32. [Crossref] [PubMed]
29. Ganzetti M, Wenderoth N, Mantini D. Whole brain myelin mapping using T1- and T2-weighted MR imaging data. Front Hum Neurosci 2014;8:671. [Crossref] [PubMed]
30. Nazeri A, Krsnik Ž, Kostović I, et al. Neurodevelopmental patterns of early postnatal white matter maturation represent distinct underlying microstructure and histology. Neuron 2022;110:4015-4030.e4. [Crossref] [PubMed]
31. Soun JE, Liu MZ, Cauley KA, et al. Evaluation of neonatal brain myelination using the T1- and T2-weighted MRI ratio. J Magn Reson Imaging 2017;46:690-6. [Crossref] [PubMed]
32. Dubois J, Dehaene-Lambertz G, Kulikova S, et al. The early development of brain white matter: a review of imaging studies in fetuses, newborns and infants. Neuroscience 2014;276:48-71. [Crossref] [PubMed]
33. Kinney HC, Brody BA, Kloman AS, et al. Sequence of central nervous system myelination in human infancy. II. Patterns of myelination in autopsied infants. J Neuropathol Exp Neurol 1988;47:217-34. [Crossref] [PubMed]
34. Paus T, Collins DL, Evans AC, et al. Maturation of white matter in the human brain: a review of magnetic resonance studies. Brain Res Bull 2001;54:255-66. [Crossref] [PubMed]
35. Ducsay CA, Goyal R, Pearce WJ, et al. Gestational Hypoxia and Developmental Plasticity. Physiol Rev 2018;98:1241-334. [Crossref] [PubMed]
36. Luo Q, Zhang JX, Huang S, et al. Effects of long-term exposure to high altitude on brain structure in healthy people: an MRI-based systematic review and meta-analysis. Front Psychiatry 2023;14:1196113. [Crossref] [PubMed]
37. Zhang YQ, Zhang WJ, Liu JH, et al. Effects of Chronic Hypoxic Environment on Cognitive Function and Neuroimaging Measures in a High-Altitude Population. Front Aging Neurosci 2022;14:788322. [Crossref] [PubMed]
38. Bethlehem RAI, Seidlitz J, White SR, et al. Brain charts for the human lifespan. Nature 2022;604:525-33. [Crossref] [PubMed]
39. Kostović I, Jovanov-Milosević N. The development of cerebral connections during the first 20-45 weeks’ gestation. Semin Fetal Neonatal Med 2006;11:415-22. [Crossref] [PubMed]
40. Zhang X, Zhang J. The human brain in a high altitude natural environment: A review. Front Hum Neurosci 2022;16:915995. [Crossref] [PubMed]
41. Barbier EL, Lamalle L, Décorps M. Methodology of brain perfusion imaging. J Magn Reson Imaging 2001;13:496-520. [Crossref] [PubMed]
42. Lafuente JV, Bermudez G, Camargo-Arce L, et al. Blood-Brain Barrier Changes in High Altitude. CNS Neurol Disord Drug Targets 2016;15:1188-97. [Crossref] [PubMed]
43. Jansen GF, Krins A, Basnyat B, et al. Role of the altitude level on cerebral autoregulation in residents at high altitude. J Appl Physiol (1985) 2007;103:518-23. [Crossref] [PubMed]
44. Aboouf MA, Thiersch M, Soliz J, et al. The Brain at High Altitude: From Molecular Signaling to Cognitive Performance. Int J Mol Sci 2023;24:10179. [Crossref] [PubMed]
45. Klein B, Kuschinsky W, Schröck H, et al. Interdependency of local capillary density, blood flow, and metabolism in rat brains. Am J Physiol 1986;251:H1333-40. [Crossref] [PubMed]
46. Callewaert B, Gsell W, Lox M, et al. Intravoxel incoherent motion as a surrogate marker of perfused vascular density in rat brain. NMR Biomed 2024;37:e5148. [Crossref] [PubMed]
47. West JB. High life: a history of high-altitude physiology and medicine. New York: Springer; 1998.
48. Parikh NA, Sharma P, He L, et al. Perinatal Risk and Protective Factors in the Development of Diffuse White Matter Abnormality on Term-Equivalent Age Magnetic Resonance Imaging in Infants Born Very Preterm. J Pediatr 2021;233:58-65.e3. [Crossref] [PubMed]
49. Wang F, Yang YJ, Yang N, et al. Enhancing Oligodendrocyte Myelination Rescues Synaptic Loss and Improves Functional Recovery after Chronic Hypoxia. Neuron 2018;99:689-701.e5. [Crossref] [PubMed]

(English Language Editor: L. Huleatt)