A coordinated shift toward a macrocytic and hypochromic erythrocyte phenotype in adults with neurofibromatosis type 1 Authors: Yoshimasa Nobeyama, Satoko Imai, Akihiko Asahina Published: 2026 ABSTRACT AbstractNeurofibromatosis type 1 (NF1) is a multisystem disorder, but its hematological features remain incompletely characterized. We investigated whether erythrocyte-related indices show coordinated alterations in adults with NF1. In this retrospective single-institute study, 225 adults with NF1 were analyzed. Z-scores for erythrocyte-related parameters were calculated relative to a reference population. One-sample tests, correlation analyses, and comparisons across tumor burden (D-class) were performed. A composite score, defined as the mean of parameter-specific Z-scores, was used to assess coordinated changes. Z-scores for red blood cell count, hemoglobin, hematocrit, and mean corpuscular volume were modest but significantly increased, whereas mean corpuscular hemoglobin concentration shifted in the opposite direction. Individual effect sizes were small; however, the composite score showed a moderate effect size (Cohen’s d = 0.43), indicating a coherent directional shift. Mean corpuscular volume was inversely correlated with red blood cell count (r = −0.390, P < .001), and a higher tumor burden was associated with lower mean corpuscular hemoglobin concentration. Adult patients with NF1 exhibit subtle but coordinated erythrocyte alterations, suggesting a macrocytic and relatively hypochromic profile, with erythrocyte hemoglobin concentration tending to decrease with increasing tumor burden. These findings may represent a systemic feature of NF1 and warrant further investigation.Keywords: Hemoglobin, MCHC, MCV, NF1, Red blood cell Introduction Neurofibromatosis type 1 (NF1) is an autosomal dominant disorder caused by germline mutations in the NF1 gene (Gutmann et al, 2017). The disease manifests with a wide spectrum of clinical features that affect multiple organ systems, including pigmented macules, neurofibromas, malignant peripheral nerve sheath tumors, optic pathway gliomas, skeletal deformities, cognitive impairment, and vasculopathy. NF1-associated vasculopathies include renal artery stenosis (Minhas et al, 2025), moyamoya disease (Brosius et al, 2022), cardiovascular disease (Cutruzzolà et al, 2020), and retinal vascular occlusion (Özdemir Zeydanlı and Özdek, 2023). While these occlusive vascular disorders have been primarily attributed to structural abnormalities, such as vascular stenosis, emerging evidence suggests that erythrocyte-related factors may also play a role in microcirculatory disturbances, particularly in small vessels, such as those in the retina. Indeed, beyond the context of NF1, recent studies have indicated that erythrocyte indices, such as mean corpuscular volume (MCV), are associated with cardiovascular events (Hsieh et al, 2017) and deep vein thrombosis (Xu et al, 2024).Furthermore, Gamaldo et al (2013) reported that elevated MCV levels are associated with poorer cognitive function. Regarding NF1, certain genotype–phenotype correlations have been documented (Bettegowda et al, 2021; Koczkowska et al, 2018; Scala et al, 2021). However, interpreting hematologic data in NF1 is complicated by intrinsic sex-related differences in peripheral blood parameters in the general population. Consequently, neither sex-pooled analyses nor direct comparisons of raw data can distinguish NF1-specific alterations from baseline physiological differences. To address this, we employed Z-score transformation based on age- and sex-matched controls. This approach standardizes the data and allows for combined analysis across sexes, while enabling direct comparison with reference values. By minimizing the influence of physiological variation related to age and sex, Z-score normalization facilitates a more precise characterization of the integrated hematologic phenotype associated with NF1. Accordingly, the present study utilized Z-score normalization to define the erythrocyte profile in patients with NF1, aiming to characterize the integrated physiological phenotype associated with NF1, while minimizing the influence of age- and sex-related physiological variation. Results Data validity No significant differences in age distribution were observed between patients with NF1 and the non-NF1 reference population, nor between male and female patients within the NF1 cohort (Tables 1 and 2). In addition, phenotypic characteristics based on the DNB classification (dermatological, neurological and bone manifestations) did not differ significantly by sex (Table 3), supporting the comparability of the analyzed subgroups. Table 1Comparison of Age Distribution Between Patients with NF1 and Non-NF1 Reference GroupsAge (y)NF1ReferenceP-valueMale populationn = 78n = 501.00020–297430–3910640–49201350–59191260–691410≥7085Female populationn = 147n = 37.99320–2912330–3924640–49431250–5942960–69174≥7093Abbreviation: NF1, neurofibromatosis type 1. Comparison of Age Distribution Between Patients with NF1 and Non-NF1 Reference Groups Abbreviation: NF1, neurofibromatosis type 1. Table 2Comparison of Age Distribution Between Male and Female Patients with NF1Age (y)Male (n = 78)Female (n = 147)P-ValueOverall.59620–2971230–39102440–49204350–59194260–691417≥7089Abbreviation: NF1, neurofibromatosis type 1. Comparison of Age Distribution Between Male and Female Patients with NF1 Abbreviation: NF1, neurofibromatosis type 1. Table 3Comparison of Phenotypes Between Male and Female Patients with NF1Mean Age(y) and DNB ClassificationMale (n = 78)Female (n = 147)P-ValueMean age ± SD50.3 ± 14.448.4 ± 12.7.264D-class.102D107D22844D33046D42050N-class.255N04165N13675N217B-class.714B050101B11929B2917Abbreviation: NF1, neurofibromatosis type 1. Comparison of Phenotypes Between Male and Female Patients with NF1 Abbreviation: NF1, neurofibromatosis type 1. Erythrocyte-related parameters in NF1 One-sample Wilcoxon signed-rank tests demonstrated that Z-scores for height and body weight were significantly below zero in patients with NF1 (Table 4 and Figure 1a). Table 4Z-score Deviations and Effect Sizes of Clinical Parameters in Patients with NF1Clinical ParametersMedian ValueP-ValueEffect Size in Cohen’s dHeight−0.695< .001−0.659Body weight−0.475< .001−0.659RBC count0.355< .0010.267Hemoglobin0.261.0020.053Hematocrit0.326< .0010.232MCV0.170.0220.192MCH0.120.235−0.041MCHC−0.241< .001−0.417Abbreviations: MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; NF1, Neurofibromatosis type 1; RBC, red blood cell. Z-score Deviations and Effect Sizes of Clinical Parameters in Patients with NF1 Abbreviations: MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; NF1, Neurofibromatosis type 1; RBC, red blood cell. Z-scores represented by box chart. The vertical and horizontal axes represent each parameter and the Z-score, respectively. In the box plots, the right and left edges of the box indicate the upper and lower quartiles, while the vertical bar inside represents the median. The whiskers extend to the upper and lower extremes. Outliers are shown as individual dots, and the X-mark indicates the mean. Gray and white boxes represent male and female data, respectively. (a) Z-scores for body size. (b) Z-scores for erythrocyte parameters. MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume. For erythrocyte-related parameters, Z-scores for red blood cell (RBC) count, hemoglobin, hematocrit, and MCV were modestly but significantly shifted toward higher values relative to the reference population, whereas the mean corpuscular hemoglobin concentration (MCHC) was significantly lower. In contrast, the MCH did not show a significant deviation from zero (Figure 1b). Effect size analysis based on Cohen’s d revealed small magnitudes for individual parameters (RBC count: 0.267; hemoglobin: 0.053; hematocrit: 0.232; MCV: 0.192; MCH: −0.014; MCHC: −0.417). The composite erythrocyte phenotype score showed a moderate effect size (Cohen’s d = 0.43), exceeding that observed for individual parameters, indicating a consistent directional shift across erythrocyte-related indices. MCH was excluded from the composite score due to its minimal deviation and lack of statistical significance. A significant inverse correlation was observed between MCV and RBC count (r = −0.390, P < .001) (Figure 2), consistent with the physiological relationship between erythrocyte size and number. Negative correlation of MCV with RBC count. The vertical and horizontal axes represent MCV and RBC count, respectively. Each open circle corresponds to an individual patient's data point, with regression lines superimposed. MCV, mean corpuscular volume; RBC, red blood cell. Kruskal–Wallis analysis demonstrated that hemoglobin and MCHC Z-scores were significantly associated with D-class (Table 5 and Figure 3). In particular, MCHC showed a stepwise decrease with increasing D-class. Table 5Correlation Between NF1 Phenotypes and Z-scores for Each ParameterDNB ClassificationRBCHbHtMCVMCHMCHCMV for D-classesD10.3550.3230.2770.2570.1850.546D20.3290.4890.4090.1620.2140.071D30.3230.1050.3550.2140.185−0.154D40.3640.1360.2890.102−0.100−0.602P-value.485.037.152.662.084.005MV for N-classesN00.3830.3850.4000.0930.107−0.445N10.3170.1360.2750.3430.167−0.167N20.2010.0430.2400.1620.050−0.204P-value.153.211.309.308.410.793MV for B-classesB00.3550.3230.3260.1360.146−0.286B10.2660.0230.0070.177−0.028−0.341B20.3590.3540.5490.2600.223−0.167P-value.812.341.552.825.479.228Abbreviations: Hb, hemoglobin; Ht, hematocrit; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; MV, median value; NF1, Neurofibromatosis type 1; RBC, red blood cell. Correlation Between NF1 Phenotypes and Z-scores for Each Parameter Abbreviations: Hb, hemoglobin; Ht, hematocrit; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; MV, median value; NF1, Neurofibromatosis type 1; RBC, red blood cell. Erythrocyte parameters associated with phenotype. Distribution of standardized scores for blood parameters across different D-classes. The vertical and horizontal axes represent the D-classes and the Z-scores for hemoglobin and MCHC, respectively. In the box plots, the right and left edges of the box indicate the upper and lower quartiles, while the vertical bar inside represents the median. The whiskers extend to the upper and lower extremes. Outliers are shown as individual dots, and the X-mark indicates the mean. MCHC, mean corpuscular hemoglobin concentration. Linear regression analysis showed no significant association between body size (height or weight) and erythrocyte-related parameters (Table 6). Table 6Correlation Between Body Size and Erythrocyte Parameters in NF1RBC ParametersHeight r-ValueHeight P-ValueBody Weight r-ValueBody Weight P-ValueRBC count0.008.9060.093.181Hemoglobin0.051.4060.018.791Hematocrit0.048.4830.014.844MCV0.059.3900.099.152MCH0.057.4030.062.374MCHC0.044.5260.012.858 Correlation Between Body Size and Erythrocyte Parameters in NF1 Sex-based comparison Mann–Whitney U testing revealed that hemoglobin and hematocrit Z-scores were significantly more positively deviated in male patients than in female patients (P = .009 and .014, respectively), whereas no significant sex differences were observed in RBC count, MCV, MCH, or MCHC (Table 7). Table 7Comparison in Z-scores for Each Parameter Between Male and Female PatientsClinical ParametersMale (n = 78) – Median ValueFemale (n = 147) – Median ValueP-ValueHeight−0.650−0.730.915Body weight−0.610−0.450.204RBC count0.3600.350.919Hemoglobin0.5400.140.009Hematocrit0.6100.250.014MCV0.2500.140.067MCH0.1700.110.266MCHC−0.340−0.170.442Abbreviations: MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; NF1, Neurofibromatosis type 1; RBC, red blood cell. Comparison in Z-scores for Each Parameter Between Male and Female Patients Abbreviations: MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; NF1, Neurofibromatosis type 1; RBC, red blood cell. Discussion This study demonstrates that erythrocyte-related indices in adult patients with NF1 exhibit subtle but coordinated shifts, rather than marked abnormalities in individual parameters. The observed pattern is characterized by modest increases in erythrocyte mass–related indices and cell size without a corresponding increase in hemoglobin content per cell, resulting in a relative reduction in hemoglobin concentration. This is further supported by the composite analysis, which captures the coordinated directional changes across multiple erythrocyte indices. The mechanisms underlying these findings remain unclear; however, several non-mutually exclusive explanations should be considered. First, relative hemoconcentration may contribute to increased RBC count, hemoglobin, and hematocrit. However, this alone is unlikely to explain the observed increase in MCV or its inverse relationship with RBC count. Second, NF1 haploinsufficiency may influence erythropoiesis. Neurofibromin regulates RAS signaling, which plays a key role in hematopoietic stem and progenitor cell function, and altered RAS pathway activity could therefore affect erythroid differentiation, although direct evidence in NF1 remains limited. Third, nutritional factors such as vitamin B12 deficiency may contribute, as previously reported in NF1 (Aydin and Bucak, 2021). In the present study, height and body weight were reduced, consistent with previous reports in NF1 (Souza et al, 2016; Vassilopoulou-Sellin et al, 1995). Some studies have also suggested that patients with NF1 may have suboptimal nutritional status (Souza et al, 2019), which might be expected to reduce erythrocyte mass. However, we observed modest but significant shifts toward increased erythrocyte-related indices, including RBC count. In addition, no significant associations were found between body size (height or weight) and erythrocyte-related parameters. These findings suggest that the observed erythrocyte phenotype is likely multifactorial and that erythropoiesis in NF1 may be regulated independently of general somatic growth, potentially reflecting systemic alterations associated with the disease. Increased MCV has been associated with reduced erythrocyte deformability and impaired microcirculatory flow in previous studies (Franczuk et al, 2015; Prudinnik et al, 2024; von Tempelhoff et al, 2016), and has also been linked to both cognitive impairment and vascular events in the general population (Gamaldo et al, 2013; Hsieh et al, 2017; Xu et al, 2024). Given that NF1 is associated with cognitive impairment and vascular complications, it is conceivable that even subtle alterations in erythrocyte properties could influence microvascular physiology. However, the present data do not establish a causal relationship, and this interpretation should be considered hypothesis-generating. Neurofibroma burden, mainly represented by D-class, was negatively associated with MCHC. Given that MCHC is typically tightly regulated (Higgins, 2015), this trend, although modest in magnitude, may suggest a link between disease burden and erythrocyte hemoglobin concentration. One possible explanation is mild iron-restricted erythropoiesis associated with chronic inflammatory activity in neurofibromas, including the enrichment of mast cells and macrophages. Inflammatory cytokines may influence systemic iron metabolism through hepcidin-mediated pathways (Nemeth et al, 2004). However, this remains speculative and requires direct validation. Sex-based analysis showed that hemoglobin and hematocrit deviations were more pronounced in males, whereas erythrocyte size–related indices were comparable between sexes. These findings suggest that the impact of NF1 on erythrocyte mass may differ by sex, with a greater shift observed in males, potentially reflecting an interaction between NF1-related pathways and sex-dependent regulation of erythropoiesis. Because NF1 is a systemic disorder affecting multiple physiological processes, including metabolism, endocrine regulation, and organ function, variables such as body size, nutritional status, and renal function may themselves represent manifestations of the disease rather than purely external confounding factors. Accordingly, the erythrocyte phenotype observed in the present study should not necessarily be interpreted as an isolated hematological abnormality but rather as 1 component of an integrated physiological phenotype associated with NF1. From this perspective, our findings reflect the overall systemic consequences of NF1, incorporating the combined influence of multiple physiological processes, rather than the effect of any single determinant. Although this integrative approach does not eliminate the possibility of residual confounding, it may provide a more clinically relevant representation of the biological state of adults with NF1. This concept may also explain why coordinated but individually modest alterations were observed across multiple erythrocyte indices, as such changes likely reflect the cumulative influence of systemic physiological alterations rather than disruption of a single hematological pathway. This study has several limitations. First, the retrospective design and lack of formal power calculation raise the possibility of insufficient statistical power, particularly for parameters that show only trends. Second, the single-institute design may limit generalizability and raises the possibility of institute-specific biases. Third, potential confounders, such as nutritional status (eg, vitamin B12 or iron levels), smoking status, and renal function, were not systematically assessed. In particular, concerning renal function, patients with NF1 often exhibit lower serum creatinine levels and, consequently, higher eGFR (Yasuda et al, 2022), which likely reflects reduced muscle mass rather than true renal hyperfiltration. Given that renal function can influence RBC phenotypes, the present data should be interpreted with caution. In addition, height and body weight data were unavailable for the control group, preventing a direct comparison of body mass index or the evaluation of potential confounding effects of obesity on erythropoiesis. Although some of these factors may themselves constitute integral components of the systemic physiological alterations associated with NF1, their absence made it difficult to completely distinguish disease-specific effects from secondary influences. Fourth, the absence of data on red cell distribution width limits the detailed evaluation of erythrocyte heterogeneity and deformability. Finally, the modest effect sizes across parameters indicate that these findings should be interpreted cautiously. In conclusion, adult patients with NF1 exhibit subtle but consistent shifts in erythrocyte indices, characterized by an increased cell size without a proportional hemoglobin gain, leading to a relative reduction in intracellular hemoglobin concentration. In addition, neurofibroma burden, mainly represented by D-class, was negatively associated with MCHC, suggesting that erythrocyte hemoglobin concentration may further decline with increasing disease burden. Although the magnitude of these changes is modest and their mechanisms remain unclear, their consistency and association with disease burden suggest potential biological relevance. These findings should be considered exploratory and hypothesis-generating and warrant further investigation. Materials and Methods Patients The study protocol was reviewed and approved by the Ethics Committee of The Jikei University School of Medicine, Tokyo, Japan (Approval No. 32-073). Because this was a non-interventional, retrospective, observational study that utilized completely anonymized data, it posed no risk or distress to the study participants. Consequently, the requirement for written informed consent was formally waived by the committee. Instead, comprehensive information regarding the study was publicly disclosed, and all subjects were provided with the opportunity to opt-out, strictly in accordance with the Ethical Guidelines for Medical and Health Research Involving Human Subjects in Japan. Two hundred twenty-five Japanese patients with NF1 (78 males and 147 females) and 87 Japanese non-NF1 individuals (50 males and 37 females; these included patients suffering from lipoma [28 males and 17 females], atheroma [15 males and 8 females], and melanocytic nevus [7 males and 12 females]) were retrospectively examined after matching for age, in which ± 5% was allowed in the proportion of each generation by 10-year age group (Table 1 and Supplementary Tables S1 and S2). The patients included in this study met the following criteria: (i) referral to The Jikei University School of Medicine; (ii) for patients with NF1, fulfillment of the revised diagnostic criteria for NF1 proposed by Legius et al (2021); (iii) for patients with NF1, surgical excision of neurofibromas between January 2018 and February 2025, and for non-NF1 individuals, surgical excision of lipoma, atheroma, or melanocytic nevus between January 2017 and December 2018; (iv) availability of blood test results from preoperative evaluations performed within 6 weeks prior to surgery; (v) absence of acute illness at the time of the preoperative evaluation or surgery; and (vi) no history of systemic treatment, including glucocorticoids or MAPK/ERK kinase inhibitors. Data on body size of Japanese population Reference data for body size, including height and body weight, in the general population were obtained from e-Stat, the official portal site for Japanese government statistics (https://www.e-stat.go.jp/dbview?sid=0003224177). Examination The clinical phenotypes were categorized according to the DNB severity classification (Ehara et al, 2017), which systematically evaluates the involvement of dermatological (D), neurological (N), and skeletal (B) systems in patients with NF1. Each D-class was defined as follows: D1, pigmented macules and a few neurofibromas; D2, pigmented macules and many neurofibromas; D3, numerous neurofibromas (>1000 in number, >1 cm in size); D4, severe plexiform neurofibromas or a malignant peripheral nerve sheath tumor. Each N-class was defined as follows: N0, no neurological symptoms; N1, neurological symptoms (eg, paralysis or pain) and/or abnormal neurological findings; N2, severe or progressive neurological symptoms. Each B-class was defined as follows: B0, no bone lesions; B1, a mild or moderate bone lesion (deformity in the spine or extremities that does not require treatment); B2, a severe bone lesion (dystrophic type or spine deformity that requires surgery (eg, scoliosis or kyphosis), severe bone deformity in the extremities (eg, pseudarthrosis and fracture), or a defect of the skull or facial bone). Hematological parameters were evaluated in individuals with and without NF1 using an automated cell counter (XE-5000; Sysmex, Kobe, Japan). Calculation and statistical analysis The sample size was not determined by a formal power calculation, as this was a retrospective observational study. Instead, all eligible patients who met the predefined inclusion criteria during the study period were consecutively included. To reduce potential bias and improve comparability between groups, non-NF1 individuals were selected using an age-matching strategy, allowing a ±5% difference in the proportion of each 10-year age category. The final sample size was therefore determined by the number of available cases that satisfied the inclusion criteria and the matching conditions. Despite the absence of a priori power calculation, the sample size was considered sufficient to detect clinically meaningful differences based on the observed effect sizes and consistency across analyses. For each sex, Z-scores were calculated by subtracting the sex-specific control mean from individual NF1 values and dividing by the corresponding control SD. Preoperative blood test results from non-NF1 individuals who underwent surgical excision of a solitary lipoma, an atheroma, or a melanocytic nevus were used as control reference data. Linear regression analysis was used to assess the relationship between 2 continuous variables. Differences among phenotypic groups, as defined by the DNB severity classification system, were evaluated using the Kruskal–Wallis test. Sex-related differences in continuous and categorical variables were analyzed using the Mann–Whitney U test and the chi-square test, respectively. To assess whether the distribution of Z-scores significantly deviated from zero, the 1-sample Wilcoxon signed-rank test was applied. Effect sizes were quantified using Cohen’s d. For standardized Z-scores, Cohen's d corresponds to the mean Z-score, as the SD is 1 by definition. Effect sizes for the composite scores were calculated as Cohen's d, defined as the mean of the composite Z-scores divided by their SD. Unlike individual standardized Z-scores (which have a SD of 1), the composite Z-scores do not necessarily retain unit variance; therefore, standardization by the sample SD was applied. MCH was not included because it showed minimal deviation and lacked a consistent directional pattern. To ensure directional consistency across parameters, the Z-scores for MCHC were sign-inverted prior to constructing the composite score, as MCHC showed changes in the opposite direction compared to the other parameters. All statistical analyses were performed using IBM Statistical Package for the Social Sciences Statistics version 22 (IBM Corporation, Armonk, NY), and a 2-sided P value < .05 was considered statistically significant. Ethics Statement The study protocol was reviewed and approved by the Ethics Committee of The Jikei University School of Medicine, Tokyo, Japan (Approval No. 32-073). Because this was a non-interventional, retrospective, observational study that utilized completely anonymized data, it posed no risk or distress to the study participants. Consequently, the requirement for written informed consent was formally waived by the committee. Instead, comprehensive information regarding the study was publicly disclosed, and all subjects were provided with the opportunity to opt-out, strictly in accordance with the Ethical Guidelines for Medical and Health Research Involving Human Subjects in Japan. Data Availability Statement The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials. ORCIDs Yoshimasa Nobeyama: http://orcid.org/0000-0002-0125-8930 Satoko Imai: http://orcid.org/0000-0001-7223-5579 Akihiko Asahina: http://orcid.org/0000-0001-6127-3182 Conflict of Interest The authors state no conflict of interest. Acknowledgments This study was supported by institutional funds from The Jikei University School of Medicine to cover the research execution, authors' time, administrative support, and publishing costs. Author Contributions Conceptualization: YN; Data curation: SI; Formal analysis: YN; Investigation: YN; Project administration: YN, AA; Resources: SI; Supervision: AA; Validation: YN; Writing - Original Draft Preparation: YN, SI; Writing - Review and Editing: AA Declaration of Generative Artificial Intelligence (AI) or Large Language Models (LLMs) During the preparation of this work, the authors used GPT-4o mini to improve the English language and grammatical correctness of the manuscript. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Footnotes Cite this article as: JID Innovations 2026;X:100527. Supplementary material is linked to the online version of the paper at www.jidonline.org, and at 10.1016/j.xjidi.2026.100527. Supplementary Material Supplementary Tables References Associated Data Supplementary Materials Supplementary Tables Data Availability Statement The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.