Differentiation of inflammatory stages: prepartum blood data patterns detect at least five inflammatory stages that differ in their birth-related adverse outcomes

Differentiation of inflammatory stages: prepartum blood data patterns detect at least five inflammatory stages that differ in their birth-related adverse outcomes | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Brief Communication Differentiation of inflammatory stages: prepartum blood data patterns detect at least five inflammatory stages that differ in their birth-related adverse outcomes Hari Shankar, Yash Gupta, Neeta Kumar, Vinoth Gnana Chellaiyan Devanbu, and 8 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6335868/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Inflammation is a multi-stage process that affects all humans. While differentiating inflammatory stages may have broad medical applications, no real-time, easily implemented method currently differentiates early (or acute) from late (or chronic) stages. Here, an algorithm expected to detect unlimited inflammatory phases and associated outcomes was explored within reproductive health –a large medical field that lacks personalized prognostics. Maternal (pre-partum) blood data were utilized to prognosticate adverse birth outcomes: low birth weight and pre-term birth (double risks). Immunological and/or clinical non-overlapping data patterns distinguished five inflammatory stages: ‘early inflammation’ (one neutrophil-predominant subset, no double risks), ‘late inflammation’ (two mononuclear-predominant subsets, one with 16.6%, the other with 83.3% double risks), and ‘no inflammation’ (two lymphocyte-predominant subsets with no double risks). Because construct, internal, external and statistical validity aspects were documented, this method may support earlier and personalized interventions of pregnancies at risk. To that end, prospective studies should determine its generalizability. Health sciences/Health care/Prognosis/Pregnancy outcome Health sciences/Biomarkers/Prognostic markers Health sciences/Medical research/Translational research Inflammatory stages Immune complexity Adverse birth outcomes Low birth weight Preterm birth Pattern recognition Prognosis Reductionism Figures Figure 1 Figure 2 Main Inflammation is a multi-stage process that affects all humans. While differentiating inflammatory stages may have broad medical applications, no real-time, easily implemented method currently differentiates early (or acute) from late (or chronic) stages. Here, an algorithm expected to detect an unlimited number of inflammatory phases and associated outcomes was explored within reproductive health –a large medical field that lacks personalized prognostics. Maternal (pre-partum) blood data were compared to birth-related low birth weight and pre-term birth (double risks). Immunological and/or clinical non-overlapping data patterns distinguished five inflammatory stages: ‘early inflammation’ (one neutrophil-predominant subset, no double risks), ‘late inflammation’ (two mononuclear-predominant subsets, one with 16.6%, the other with 83.3% double risks), and ‘no inflammation’ (two lymphocyte-predominant subsets with no double risks). Because construct, internal, external and statistical validity aspects were documented, this method may support earlier and personalized interventions of pregnancies at risk. To that end, prospective studies should determine its generalizability. Inflammation affects all living vertebrates. It is a process that includes several temporal stages. 1 While numerous tests differentiate absence of inflammation from earlier inflammatory responses, no real-time, low-cost test currently identifies later responses. Tests that distinguish late inflammations that evolve into recovery (those including the resolution phase of inflammation) from those that promote chronicity could have numerous applications in biomedical research and clinical practices. 2 To develop such tests, methodological paradigms and validation contexts need to be considered. Self-limiting (reductionist) paradigms are not appropriate because pre-established assumptions –e.g., cutoffs claimed to apply to everybody, everywhere– induce non-trivial errors. 3 However, combinatorial theory and reproductive health offer a methodological alternative and a relevant testing context. That is so because immunity, like human language, is a system in which a few primary elements (whether cell types or letters) produce quasi-infinite combinations. Even with small datasets, the combinatorial nature of the immune system, integrated within a non-reductionist method that focuses on pattern recognition, may uncover hidden information. 4 Such a testing approach could be evaluated in India –where more births occur than in any other country and many of them result in preterm birth (PTB) and/or low birth weight (LBW). 5 Accordingly, this study was designed to elucidate: (1) whether a systems-oriented, assumption-free, computerized combinatorial method that captures complex interactions among leukocytes and non-cellular analytes in pre-partum maternal blood can predict birth-related adverse outcomes; (2) whether observed patterns are immunologically and clinically valid; and (3) whether validated findings may support biomedical research and earlier medical interventions. This method was evaluated in 192 pregnancies from India (Supplementary Data Table 1), where pre-partum maternal blood leukocytes, C-Reactive protein (CRP) and ferritin were investigated for their ability to differentiate inflammation stages and anticipate LBW and PTB (here described as ‘double risks’). Leukocyte percentages were not informative when each cell type was measured separately (Fig. 1 a). When an algorithm partitioned the data into four classes (Figs. 1 b-f), neither leukocyte (lymphocyte [L], neutrophil [N]) percentages nor the phagocyte/lymphocyte (P/L) ratio fully discriminated: two of such classes overlapped (Fig. 1 b). In contrast, all data classes were differentiated when complex indicators generated by the same algorithm were considered (Figs. 1 c-f). Data partitioning was validated with directly interpretable indicators (Figs. 1 a, b, and 2 a-d). While classes ‘A’ and ‘B’ did not differ clinically, they differed immunologically: (a) high lymphocyte percentages and low P/L ratio values characterized class ‘A’ (regarded to reflect absence of inflammation ); and (b) given its opposite pattern, class ‘B’ seemed to express early inflammation (Fig. 1 b). The remaining data points –which separated classes ‘A’ and ‘B’– were named ‘ late inflammations’ (oval, Fig. 1 b). Clinical data also informed: 3.3% or 9.6% double risks were associated with classes ‘C’ or ‘D’, respectively, Figs. 1 b-f). Additional analyses that focused on interactions (ratios among cell percentages) showed that the highest N/L ratios occurred in the ‘B’ class (Fig. 2 a). Supporting the view that it expresses late inflammations evolving into chronicity , the ‘D’ class displayed the highest monocyte percentages (Fig. 2 b). Because class ‘C’ data points partially overlapped with ‘A’ ones, ‘C’ class appeared to represent late inflammations under recovery (Fig. 2 c). Because complete differentiation (no data overlapping among classes) was only detected when non-lymphocyte, multi-cellular interactions were evaluated (Fig. 2 d), earlier reports that suggest systems-level information emerges from interactions among variables (not from any one primary variable measured in isolation) were supported as well as those that associate chronic inflammations with monocytes. 6 , 7 It was also observed that interactions in which monocytes predominate expressed kurtosis (‘L’-shaped, non-linear, non-unimodal distribution, Extended Data Fig. 1 ). Data partitioning was also validated by independent (non-cellular) variables. Supporting the hypothesis that birth-related adverse outcomes risks may be associated with inflammation, the highest concentrations of ferritin were found in some ‘D’ observations (Extended Data Fig. 2 a). Some ‘A’ observations with high concentrations of ferritin differed from a subset of ‘B’ and ‘D’ data points that also presented with high concentrations of ferritin (Extended Data Fig. 2 b). Supporting the notion that CRP may not be highly informative in reproductive health, data points with high CRP concentrations represented less than half the number of data points with high concentrations of ferritin (Extended Data Figs. 2 c, d ). Pattern recognition did not depend on any one variable, numerical value or cutoff but patterns that displayed distinct, visually observable, non-overlapping features –such as orthogonal data inflections and bifurcations observed in three-dimensional space. While overlapping total leukocyte counts prevented differentiations across classes of data (Extended Data Fig. 3a), perpendicular data inflections differentiated two groups of data (Extended Data Figs. 3b-d). Data bifurcatons also facilitated additional discoveries. For example, one bifurcation helped separate ‘A’ class data into two (‘A1’ and ‘A2’) distinct segments (Extended Data Fig. 3e). Other data structures grouped ‘A1’ and ‘A2’ observations into distinct, non-overlapping data groups (Extended Data Fig. 3f). The fact that five –not only four– data classes were distinguished, may prompt new studies to explore whether classes ‘A1’ and ‘A2’ differ in their associated outcomes. If corroborated, this proof-of-concept may promote error prevention and new (personalized and/or earlier) medical interventions. While CRP data did not detect double risks (Extended Data Fig. 4a) and interventions based on high concentrations of ferritin values identified 50% of the double risks (Extended Data Fig. 4b), spatial patterns identified more (83.3%) birth-related adverse outcomes (Extended Data Figs. 4c, d) and may prevent errors. Type I errors (claiming as different what is not so or false-positives) may occur when double risk-negatives show ferritin concentrations within the range exhibited by double risk-positives (Extended Data Fig. 4b). However, if tested together, ferritin and leukocyte-derived indicators could prevent such errors (Extended Data Figs. 4c-f). Because some ‘A2’ observations seemed to be associated with late inflammations that express birth-related double risks (oval, Extended Data Figs. 4g, h), future studies may explore whether false-negative errors may be prevented with early monitoring of ‘A2’ observations. An overall validation showed that reductionist assessments did not separate double risk-positives from -negatives (Extended Data. Figure 5a ). In contrast, the partitioning method distinguished five classes of data, which differed clinically and/or immunologically while showing construct, internal, external and statistical dimensions of validity 8 (Extended Data Figs. 5b, c and 6a, b). Such findings expand the literature on reproductive health (where gestation was previously viewed as composed of three inflammatory stages) and demonstrate that combinatorial immunology applies, at least, to reproduction and infectious diseases. 9 – 13 Yet, more discoveries are still possible. Alternative partitioning may derive from relationships that appear to be linearly correlated but, in fact, hide L-shaped patterns (Extended Data Figs. 7a-c). This classification differed from the previous ones because (a) two ‘A2’ data points were grouped together with all ‘A1’ observations, and (b) some ‘C’ observations were grouped together with ‘D’ data points ( Extended Data Fig. 7d ) . While some findings appeared to be contradictory (e.g., a statistically significant linear relationship was later shown to be non-linear, Extended Data Figs. 7a-d), it is suggested that they are not so but the result of a method that assesses interactions that regulate immunological functions. 14 As such, it explores complex (combinatorial) processes that result in not just one (and static) ‘solution’ but multiple (multi-dimensional and, possibly, dynamic) perspectives that may reflect different functions that occur at the same time. 4 , 14 If corroborated with prospective studies, this approach could be instrumental to the Sustainable Development Goals promoted by the United Nations, which seek a major reduction of infant mortality ( https://sdgs.un.org/sites/default/files/publications/21252030%20Agenda%20for%20Sustainable%20Development%20web.pdf ). Because this approach has already been validated in infectious diseases 4 , 10 – 13 , its generic applicability could be further explored by investigating cancer. For example, while subsets of acute myeloid leukemia seem responsive to therapies 15 , new (blood-based) methods are needed to partition patients into subsets and monitor their responsiveness to personalized treatments. In conclusion, birth-related adverse outcomes were associated with late (not with early) inflammations. Because data on blood leukocytes are abundantly available on many disease-related processes, this method can be rapidly corroborated and potentially applied without new technologies or training programs. Declarations Contributors: HS, YG and ALR conceptualized this study. NK and VGCD managed the study subject enrollment and project administration. DNR acquired funding, facilitated laboratory investigations and overall study management. ALR and HS performed retrospective data curation and analysis. ALR wrote the first draft, HS and YG did review and editing. WMT, MJI and AY have provided critical comments on validation and analysis. ALH, JMF and RBSK edited the manuscript and provided intellectual inputs. All the authors have read and approved the final version of the manuscript. Data Sharing Statement: All data relevant to this retrospective analysis are available in the Supplementary Material. Declaration of interests: AR and AH are co-inventors of the software package here utilized. All other authors declare no conflicts of interest. Role of the funding source The study was funded by the Indian Council of Medical Research. The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. Acknowledgements: The authors wish to acknowledge the support of the Human Reproduction Research Centre staff of AIIMS, New Delhi, and PGIMER, Chandigarh, who were involved in the screening, enrolment, and follow-up of the subjects. References Beck KD et al. Quantitative analysis of cellular inflammation after traumatic spinal cord injury: evidence for a multiphasic inflammatory response in the acute to chronic environment. Brain 133 , 433-447 (2010). Sugimoto MA, Sousa LP, Pinho V, Perretti M, Teixeira MM. Resolution of inflammation: What Controls Its Onset? Front. Immunol. 7 , 160 (2016). Naggara O, et al. Analysis by or dichotomizing continuous variables is inadvisable: an example from the natural history of unruptured aneurysms. Am J Neuroradiol . 32, 437-440 (2011). Rivas AL, et al. Nature and consequences of biological reductionism for the immunological study of infectious diseases. Front Immunol 8, 612 (2017). Kuppusamy P, Prusty RK, Kale DP. High-risk pregnancy in India: Prevalence and contributing risk factors - a national survey-based analysis. J Glob Health 13 , 04116 (2023). Pensotti A, Bertolaso M, Bizzarri M. Is cancer reversible? Rethinking carcinogenesis models-a new epistemological tool. Biomolecules 13 , 733 (2023). Italiani P, et al. Profiling the course of resolving vs. persistent inflammation in human monocytes: the role of IL-1 family molecules. Front. Immunol. 11 , 1426 (2020). Nanda SK, Rivas AL, Trochim WM, JD Deshler. Emphasis on Validation in Research: A Meta-Analysis. Scientometrics 48 , 45–64 (2000). Mor G. Introduction to the immunology of pregnancy. Immunol Rev . 308 , 5-8 (2022). Kempaiah P, et al. Decoding immuno-competence: a novel analysis of complete blood cell count data in COVID-19 outcomes. Biomedicines 12 , 871 (2024). Pappa T, et al. Personalized, disease-stage specific, rapid identification of immunosuppression in sepsis. Front. Immunol 15 , 1430972 (2024). Verma JS, et al. Multicellular immunological interactions associated with COVID-19 infections. Front Immunol 13 , 794006 (2022) Libertin CR, et al. Data structuring may prevent ambiguity and improve personalized medical prognosis. Mol Aspects Med 91 , 101142 (2023). Cappuccio A, et al. Deciphering the combinatorial landscape of immunity. Elife 9 , e62148 (2020). Bunaciu RP et al. Potential for subsets of wt-NPM1 primary AML blasts to respond to retinoic acid treatment. Oncotarget. 9 , 4134-4149 (2018). Methods Inclusion and ethics In compliance with the 1964 Helsinki Declaration and later amendments, this study was approved by the Ethics Committees of the two participating institutions: (1) All India Institute of Medical Sciences (AIIMS), New Delhi, India vide approval no. IEC/NP-339/2010; and (2) the Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh, Punjab, India vide approval No. 10/4815. Study design and participants This was a is a multicenter, investigator-initiated, retrospective, longitudinal, exploratory study conducted in (a) AIIMS, New Delhi and (b) PGIMER, Punjab. This study comprised lymphocyte counts at two time-points (i.e. early second trimester and 3 months after first sampling) during pregnancy and pregnancy outcomes (newborn weight and gestational length). Capture and description of leukocyte-derived complex (combinatorial) indicators A retrospective longitudinal analysis of pregnancy and birth-related data was performed with blood samples collected from pregnant females attending the antenatal clinics of the Obstetrics and Gynecology departments of either the AIIMS, New Delhi, India, or PGIMER, Chandigarh, Punjab, India. To be selected, candidates should: be / present with / belong to: (a) 19–30 years old primigravida, (b) not later than in their second trimester, (c) 8.0–13.0 g/dl of hemoglobin; (d) 18-22 body mass index, (e) middle socioeconomic indicators, and (f) provide informed consent to participate in a study approved by the Institutional Ethical Committees. Exclusion criteria included: (a) any metabolic disease, (b) malignancy, (c) heart disease, (d) any infectious disease, and (e) any chronic morbidity. Out of 429 candidate pregnancies (which included 858 observations), 558 observations (269 candidate pregnancies) were rejected due to incomplete data. The remaining 300 observations were gathered from 160 pregnancies that contributed up to two temporal observations. However, 108 of such observations lacked information on neonates. Thus, the prognostic evaluation was conducted with 131 pregnancies, which generated 192 prepartum observations that helped determine, retrospectively, whether immunological (prepartum) maternal observations predicted birth-related adverse outcomes. Low birth weight (< 2500 g) and preterm birth (gestation length < 259 days) were described as “double risks”. Supplementary Data Table 1 reports the data of this retrospective study, indicating the blood total leukocyte count (TLC/mm 3 ) and leukocyte percentages for each observation as well as serum ferritin and C-reactive protein (CRP) concentrations. Only the 192 prepartum observations were considered to conduct prognostic evaluations. Methodological procedures Two methods were compared with the same data: (1) a reductionist one, which investigated each variable (cellular or not) in isolation; and (2) a non-reductionist alternative, which explored interactions among variables. 4 Laboratory procedures Five milliliter blood samples were collected and analyzed using an automated cell counter (Sysmex XN1000, Japan). Aliquots were prepared to conduct (i) complete blood cell counts and differentials, and (ii) serological studies (including the concentrations of ferritin and CRP). Validation To explore the construct validity of the method, immunology-based data partitioning was compared to independent variables (ferritin and CRP) –an exploration that also considered statistical validity. To explore the internal and external validity (repeatability across metrics and across populations), blood samples collected from individuals residing in two (New Delhi and Punjab) territories of India were investigated with several data structures. Detection and evaluation of complex immunological data patterns Complex data structures were created and analyzed as described elsewhere. 10-13 Interactions among blood leukocyte were explored with indicators created by a proprietary algorithm (US patent 10,429,389 B2), which reveals patterns that facilitate data partitioning into three or more data groups that (a) partially or totally express non-overlapping intervals of continuous data, and (b) such intervals relate to qualitatively different outcomes, such as PTB and LBW. Sample-size estimation Because immunological responses are known to be non-linear, no previous study on the dynamic and complex immunological interactions associated with pregnancy was available to estimate sample size. However, previous studies on numerous (non-pregnancy-related) conditions have shown that, when the internal complexity and dynamics of immunological responses are evaluated, they can reveal three or more data groups that display non-overlapping data intervals (and, consequently, achieve statistically significant differences) even when the sample size is low. Because abundant, biologically interpretable, longitudinal data patterns have been observed even when n 130 was likely to show distinct data patterns, if they existed, and also offer abundant opportunities to personalized patterns of temporal data directionality. Statistics Statistical tests included comparisons among proportions and medians, which were conducted with the chi-square and Mann-Whitney tests, respectively. Such analyses were performed with a commercial package (Minitab LLC, 2024). P-values less than 0.05 were considered statistically significant. Additional Declarations Yes there is potential Competing Interest. AR and AH are co-inventors of the software package here utilized. All other authors declare no conflicts of interest. Supplementary Files SupplDateTable1.docx Complete blood cell counts and percentages of pregnant females, birth-related outcomes and inflammation-related data classes (n=192) SUPPLEMENTARYFIGURES.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6335868","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Brief Communication","associatedPublications":[],"authors":[{"id":436444189,"identity":"d2c3d932-1d57-4455-9c25-bc2025ac237f","order_by":0,"name":"Hari Shankar","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABB0lEQVRIiWNgGAWjYBACAyA+wNjAwCAB4iUwHJAD0QcekKLFGKwlgYAWBrgWoOLEBohe3MCc/ezDgz932OVJtp999uBBxZ30+WGHHwJtsZPTbcCuxbIn3eAw75nkYmmedHODhDPPcjfeTjMAakk2NjuAw2EH0hgOM7YxJ85jSGOTSGw7nLtxdgJIy4HEbbi0nH/GcPBnW33iPP5nYC3phrPTP+DXciON4QBv2+HE2RIQWxLkpXMI2HLjGcNh3rbjiTNnAG1JOHPYcIN0TsGBBAM8fjmfxvzxZ1t14ozzaWySPyoOy8vPTt/84UOFnRwuLdgCBEwSqxwE5BtIUT0KRsEoGAUjAQAA9B9qb0fGpFoAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-1736-368X","institution":"Indian Council of Medical Research","correspondingAuthor":true,"prefix":"","firstName":"Hari","middleName":"","lastName":"Shankar","suffix":""},{"id":436444190,"identity":"a270706a-26b1-4b07-8a2f-961a3c29f6d7","order_by":1,"name":"Yash Gupta","email":"","orcid":"","institution":"Penn State College of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yash","middleName":"","lastName":"Gupta","suffix":""},{"id":436444191,"identity":"8c3fbfee-aec0-4774-ac24-cd2ef10b374e","order_by":2,"name":"Neeta Kumar","email":"","orcid":"","institution":"Indian Council of Medical Research","correspondingAuthor":false,"prefix":"","firstName":"Neeta","middleName":"","lastName":"Kumar","suffix":""},{"id":436444192,"identity":"4bc36161-a241-4fd4-9e6b-52d7bbceab20","order_by":3,"name":"Vinoth Gnana Chellaiyan Devanbu","email":"","orcid":"","institution":"Chettinad Hospital \u0026 Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Vinoth","middleName":"Gnana Chellaiyan","lastName":"Devanbu","suffix":""},{"id":436444193,"identity":"eb3fd114-c19a-4abf-918b-311a8be1bcea","order_by":4,"name":"William Trochim","email":"","orcid":"","institution":"Cornell University","correspondingAuthor":false,"prefix":"","firstName":"William","middleName":"","lastName":"Trochim","suffix":""},{"id":436444194,"identity":"7d70965b-e383-4da6-90dd-387cc5fcf98f","order_by":5,"name":"Almira L. Hoogesteijn","email":"","orcid":"","institution":"Cinvestav","correspondingAuthor":false,"prefix":"","firstName":"Almira","middleName":"L.","lastName":"Hoogesteijn","suffix":""},{"id":436444195,"identity":"ed6d64cc-3bbe-4dbf-9dbc-a963b893c5b3","order_by":6,"name":"Jeanne M. Fair","email":"","orcid":"","institution":"Los Alamos National Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Jeanne","middleName":"M.","lastName":"Fair","suffix":""},{"id":436444196,"identity":"91716278-a500-4f48-b996-fcd8d4bc11b2","order_by":7,"name":"Raja Babu Singh Kushwah","email":"","orcid":"","institution":"Texas A\u0026M University","correspondingAuthor":false,"prefix":"","firstName":"Raja","middleName":"Babu Singh","lastName":"Kushwah","suffix":""},{"id":436444197,"identity":"9bff78ff-aaf7-4fbb-bb88-0907e86d9597","order_by":8,"name":"Michelle J. Iandiorio","email":"","orcid":"","institution":"University of New Mexico","correspondingAuthor":false,"prefix":"","firstName":"Michelle","middleName":"J.","lastName":"Iandiorio","suffix":""},{"id":436444198,"identity":"fa3a6fa1-c343-458f-9cf6-59283ff654ea","order_by":9,"name":"Andrew Yen","email":"","orcid":"","institution":"Cornell University","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"","lastName":"Yen","suffix":""},{"id":436444199,"identity":"5f19f21c-308a-42f1-8835-9ef320b4f9af","order_by":10,"name":"Donthamsetty Nageswara Rao","email":"","orcid":"","institution":"All India Institute of Medical Research","correspondingAuthor":false,"prefix":"","firstName":"Donthamsetty","middleName":"Nageswara","lastName":"Rao","suffix":""},{"id":436444200,"identity":"ad8e5574-0c58-4d7c-bbf1-31abadcc2c53","order_by":11,"name":"Ariel L. Rivas","email":"","orcid":"","institution":"University of New Mexico","correspondingAuthor":false,"prefix":"","firstName":"Ariel","middleName":"L.","lastName":"Rivas","suffix":""}],"badges":[],"createdAt":"2025-03-29 20:55:07","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6335868/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6335868/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":79658110,"identity":"c1331602-7bec-4fa8-97be-7f74f27c61c2","added_by":"auto","created_at":"2025-04-01 09:11:36","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":115353,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFrom no differentiation to recognition of four inflammation-related data classes. \u003c/strong\u003eLeukocyte percentages did not reveal patterns when outcomes (low weight birth and preterm birth or double risks) were explored (\u003cstrong\u003ea\u003c/strong\u003e). When a proprietary algorithm partitioned the data into four classes and two percentages and one ratio were considered, data overlapping prevented the differentiation of two classes (oval, \u003cstrong\u003eb\u003c/strong\u003e). In contrast, when complex indicators created by the same algorithm were utilized, four data classes were distinguished (here labeled A-D, \u003cstrong\u003ec-f\u003c/strong\u003e). Italics (e.g., \u003cem\u003eAX\u003c/em\u003e, \u003cem\u003eBBK\u003c/em\u003e) refer to descriptors created by the algorithm to identify data combinations derived from leukocyte data.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6335868/v1/a3bef6b0384b90b1da30ecb8.png"},{"id":79656787,"identity":"f0ea873a-354e-42a7-ae6d-54c77157c7ee","added_by":"auto","created_at":"2025-04-01 09:03:36","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":143992,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eImmuno-clinical validation of inflammatory stages. \u003c/strong\u003eThe four-class partitioning was immunologically justified because: (a) class ‘A’ (no inflammation) displayed the highest lymphocyte percentage and the lowest neutrophil percentage (i.e., the lowest N/L ratio, \u003cstrong\u003ea\u003c/strong\u003e); (b) class “B” (early inflammation) showed the highest N/L and N/M ratios (\u003cstrong\u003ea, b\u003c/strong\u003e); (c) class ‘C’ (late inflammation leading to recovery) partially overlapped with class ‘A’ (no inflammation, \u003cstrong\u003ec, d\u003c/strong\u003e); and (d) class ‘D’ was suspected to express late inflammation leading to chronicity because it showed the highest monocyte percentage and substantially overlapped with class ‘B’ (\u003cstrong\u003eb-d\u003c/strong\u003e).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6335868/v1/7685960e17c49f970352d962.png"},{"id":79792821,"identity":"96cdef06-e4e0-40ac-8879-a57e5e201a04","added_by":"auto","created_at":"2025-04-02 20:13:35","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":757268,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6335868/v1/e3b2a58a-e370-4e7f-a229-ca7b71b6f18c.pdf"},{"id":79656790,"identity":"a546332f-89a0-4580-b50e-8c894b84302a","added_by":"auto","created_at":"2025-04-01 09:03:36","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":80739,"visible":true,"origin":"","legend":"Complete blood cell counts and percentages of pregnant females, birth-related outcomes and inflammation-related data classes (n=192)","description":"","filename":"SupplDateTable1.docx","url":"https://assets-eu.researchsquare.com/files/rs-6335868/v1/441bd58075dae4a783cd9a35.docx"},{"id":79656805,"identity":"c7275eca-0174-4ba5-aeda-3a3ba422f3ae","added_by":"auto","created_at":"2025-04-01 09:03:37","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2588833,"visible":true,"origin":"","legend":"","description":"","filename":"SUPPLEMENTARYFIGURES.docx","url":"https://assets-eu.researchsquare.com/files/rs-6335868/v1/95fbd9e86b0359577a3b44e7.docx"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential Competing Interest.\nAR and AH are co-inventors of the software package here utilized. All other authors declare no conflicts of interest.","formattedTitle":"Differentiation of inflammatory stages: prepartum blood data patterns detect at least five inflammatory stages that differ in their birth-related adverse outcomes","fulltext":[{"header":"Main","content":"\u003cp\u003eInflammation is a multi-stage process that affects all humans. While differentiating inflammatory stages may have broad medical applications, no real-time, easily implemented method currently differentiates early (or acute) from late (or chronic) stages. Here, an algorithm expected to detect an unlimited number of inflammatory phases and associated outcomes was explored within reproductive health \u0026ndash;a large medical field that lacks personalized prognostics. Maternal (pre-partum) blood data were compared to birth-related low birth weight and pre-term birth (double risks). Immunological and/or clinical non-overlapping data patterns distinguished five inflammatory stages: \u0026lsquo;early inflammation\u0026rsquo; (one neutrophil-predominant subset, no double risks), \u0026lsquo;late inflammation\u0026rsquo; (two mononuclear-predominant subsets, one with 16.6%, the other with 83.3% double risks), and \u0026lsquo;no inflammation\u0026rsquo; (two lymphocyte-predominant subsets with no double risks). Because construct, internal, external and statistical validity aspects were documented, this method may support earlier and personalized interventions of pregnancies at risk. To that end, prospective studies should determine its generalizability.\u003c/p\u003e \u003cp\u003eInflammation affects all living vertebrates. It is a process that includes several temporal stages.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e While numerous tests differentiate absence of inflammation from earlier inflammatory responses, no real-time, low-cost test currently identifies later responses. Tests that distinguish late inflammations that evolve into recovery (those including the resolution phase of inflammation) from those that promote chronicity could have numerous applications in biomedical research and clinical practices.\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eTo develop such tests, methodological paradigms and validation contexts need to be considered. Self-limiting (reductionist) paradigms are not appropriate because pre-established assumptions \u0026ndash;e.g., cutoffs claimed to apply to everybody, everywhere\u0026ndash; induce non-trivial errors.\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e However, combinatorial theory and reproductive health offer a methodological alternative and a relevant testing context.\u003c/p\u003e \u003cp\u003eThat is so because immunity, like human language, is a system in which a few primary elements (whether cell types or letters) produce quasi-infinite combinations. Even with small datasets, the combinatorial nature of the immune system, integrated within a non-reductionist method that focuses on pattern recognition, may uncover hidden information.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e Such a testing approach could be evaluated in India \u0026ndash;where more births occur than in any other country and many of them result in preterm birth (PTB) and/or low birth weight (LBW).\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eAccordingly, this study was designed to elucidate: (1) whether a systems-oriented, assumption-free, computerized combinatorial method that captures complex interactions among leukocytes and non-cellular analytes in pre-partum maternal blood can predict birth-related adverse outcomes; (2) whether observed patterns are immunologically and clinically valid; and (3) whether validated findings may support biomedical research and earlier medical interventions. This method was evaluated in 192 pregnancies from India (Supplementary Data Table\u0026nbsp;1), where pre-partum maternal blood leukocytes, C-Reactive protein (CRP) and ferritin were investigated for their ability to differentiate inflammation stages and anticipate LBW and PTB (here described as \u0026lsquo;double risks\u0026rsquo;).\u003c/p\u003e \u003cp\u003eLeukocyte percentages were not informative when each cell type was measured separately (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). When an algorithm partitioned the data into four classes (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb-f), neither leukocyte (lymphocyte [L], neutrophil [N]) percentages nor the phagocyte/lymphocyte (P/L) ratio fully discriminated: two of such classes overlapped (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). In contrast, all data classes were differentiated when complex indicators generated by the same algorithm were considered (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec-f).\u003c/p\u003e \u003cp\u003eData partitioning was validated with directly interpretable indicators (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, b, and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea-d). While classes \u0026lsquo;A\u0026rsquo; and \u0026lsquo;B\u0026rsquo; did not differ clinically, they differed immunologically: (a) high lymphocyte percentages and low P/L ratio values characterized class \u0026lsquo;A\u0026rsquo; (regarded to reflect \u003cem\u003eabsence of inflammation\u003c/em\u003e); and (b) given its opposite pattern, class \u0026lsquo;B\u0026rsquo; seemed to express \u003cem\u003eearly inflammation\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). The remaining data points \u0026ndash;which separated classes \u0026lsquo;A\u0026rsquo; and \u0026lsquo;B\u0026rsquo;\u0026ndash; were named \u0026lsquo;\u003cem\u003elate inflammations\u0026rsquo;\u003c/em\u003e (oval, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Clinical data also informed: 3.3% or 9.6% double risks were associated with classes \u0026lsquo;C\u0026rsquo; or \u0026lsquo;D\u0026rsquo;, respectively, Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb-f). Additional analyses that focused on interactions (ratios among cell percentages) showed that the highest N/L ratios occurred in the \u0026lsquo;B\u0026rsquo; class (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Supporting the view that it expresses late inflammations evolving into \u003cem\u003echronicity\u003c/em\u003e, the \u0026lsquo;D\u0026rsquo; class displayed the highest monocyte percentages (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). Because class \u0026lsquo;C\u0026rsquo; data points partially overlapped with \u0026lsquo;A\u0026rsquo; ones, \u0026lsquo;C\u0026rsquo; class appeared to represent late inflammations under \u003cem\u003erecovery\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). Because complete differentiation (no data overlapping among classes) was only detected when non-lymphocyte, multi-cellular interactions were evaluated (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed), earlier reports that suggest systems-level information emerges from interactions among variables (not from any one primary variable measured in isolation) were supported as well as those that associate chronic inflammations with monocytes.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e It was also observed that interactions in which monocytes predominate expressed kurtosis (\u0026lsquo;L\u0026rsquo;-shaped, non-linear, non-unimodal distribution, Extended Data Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eData partitioning was also validated by independent (non-cellular) variables. Supporting the hypothesis that birth-related adverse outcomes risks may be associated with inflammation, the highest concentrations of ferritin were found in some \u0026lsquo;D\u0026rsquo; observations (Extended Data Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Some \u0026lsquo;A\u0026rsquo; observations with high concentrations of ferritin differed from a subset of \u0026lsquo;B\u0026rsquo; and \u0026lsquo;D\u0026rsquo; data points that also presented with high concentrations of ferritin (Extended Data Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). Supporting the notion that CRP may not be highly informative in reproductive health, data points with high CRP concentrations represented less than half the number of data points with high concentrations of ferritin (Extended Data Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec, d\u003cb\u003e).\u003c/b\u003e\u003c/p\u003e \u003cp\u003ePattern recognition did not depend on any one variable, numerical value or cutoff but patterns that displayed distinct, visually observable, non-overlapping features \u0026ndash;such as orthogonal data inflections and bifurcations observed in three-dimensional space. While overlapping total leukocyte counts prevented differentiations across classes of data (Extended Data Fig.\u0026nbsp;3a), perpendicular data inflections differentiated two groups of data (Extended Data Figs.\u0026nbsp;3b-d).\u003c/p\u003e \u003cp\u003eData bifurcatons also facilitated additional discoveries. For example, one bifurcation helped separate \u0026lsquo;A\u0026rsquo; class data into two (\u0026lsquo;A1\u0026rsquo; and \u0026lsquo;A2\u0026rsquo;) distinct segments (Extended Data Fig.\u0026nbsp;3e). Other data structures grouped \u0026lsquo;A1\u0026rsquo; and \u0026lsquo;A2\u0026rsquo; observations into distinct, non-overlapping data groups (Extended Data Fig.\u0026nbsp;3f). The fact that five \u0026ndash;not only four\u0026ndash; data classes were distinguished, may prompt new studies to explore whether classes \u0026lsquo;A1\u0026rsquo; and \u0026lsquo;A2\u0026rsquo; differ in their associated outcomes.\u003c/p\u003e \u003cp\u003eIf corroborated, this proof-of-concept may promote error prevention and new (personalized and/or earlier) medical interventions. While CRP data did not detect double risks (Extended Data Fig.\u0026nbsp;4a) and interventions based on high concentrations of ferritin values identified 50% of the double risks (Extended Data Fig.\u0026nbsp;4b), spatial patterns identified more (83.3%) birth-related adverse outcomes (Extended Data Figs.\u0026nbsp;4c, d) and may prevent errors. Type I errors (claiming as different what is not so or false-positives) may occur when double risk-negatives show ferritin concentrations within the range exhibited by double risk-positives (Extended Data Fig.\u0026nbsp;4b). However, if tested together, ferritin and leukocyte-derived indicators could prevent such errors (Extended Data Figs.\u0026nbsp;4c-f). Because some \u0026lsquo;A2\u0026rsquo; observations seemed to be associated with late inflammations that express birth-related double risks (oval, Extended Data Figs.\u0026nbsp;4g, h), future studies may explore whether false-negative errors may be prevented with early monitoring of \u0026lsquo;A2\u0026rsquo; observations.\u003c/p\u003e \u003cp\u003eAn overall validation showed that reductionist assessments did not separate double risk-positives from -negatives (Extended Data. \u003cb\u003eFigure\u0026nbsp;5a\u003c/b\u003e). In contrast, the partitioning method distinguished five classes of data, which differed clinically and/or immunologically while showing construct, internal, external and statistical dimensions of validity\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e (Extended Data Figs.\u0026nbsp;5b, c and 6a, b). Such findings expand the literature on reproductive health (where gestation was previously viewed as composed of three inflammatory stages) and demonstrate that combinatorial immunology applies, at least, to reproduction and infectious diseases. \u003csup\u003e\u003cspan additionalcitationids=\"CR10 CR11 CR12\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eYet, more discoveries are still possible. Alternative partitioning may derive from relationships that appear to be linearly correlated but, in fact, hide L-shaped patterns (Extended Data Figs.\u0026nbsp;7a-c). This classification differed from the previous ones because (a) two \u0026lsquo;A2\u0026rsquo; data points were grouped together with all \u0026lsquo;A1\u0026rsquo; observations, and (b) some \u0026lsquo;C\u0026rsquo; observations were grouped together with \u0026lsquo;D\u0026rsquo; data points \u003cb\u003e(\u003c/b\u003eExtended Data Fig.\u0026nbsp;7d\u003cb\u003e)\u003c/b\u003e.\u003c/p\u003e \u003cp\u003eWhile some findings appeared to be contradictory (e.g., a statistically significant linear relationship was later shown to be non-linear, Extended Data Figs.\u0026nbsp;7a-d), it is suggested that they are not so but the result of a method that assesses \u003cem\u003einteractions that regulate immunological functions.\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e As such, it explores complex (combinatorial) processes that result in not just one (and static) \u0026lsquo;solution\u0026rsquo; but multiple (multi-dimensional and, possibly, dynamic) perspectives that may reflect different functions that occur at the same time.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eIf corroborated with prospective studies, this approach could be instrumental to the Sustainable Development Goals promoted by the United Nations, which seek a major reduction of infant mortality (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://sdgs.un.org/sites/default/files/publications/21252030%20Agenda%20for%20Sustainable%20Development%20web.pdf\u003c/span\u003e\u003cspan address=\"https://sdgs.un.org/sites/default/files/publications/21252030%20Agenda%20for%20Sustainable%20Development%20web.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Because this approach has already been validated in infectious diseases\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan additionalcitationids=\"CR11 CR12\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, its generic applicability could be further explored by investigating cancer. For example, while subsets of acute myeloid leukemia seem responsive to therapies\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, new (blood-based) methods are needed to partition patients into subsets and monitor their responsiveness to personalized treatments.\u003c/p\u003e \u003cp\u003eIn conclusion, birth-related adverse outcomes were associated with late (not with early) inflammations. Because data on blood leukocytes are abundantly available on many disease-related processes, this method can be rapidly corroborated and potentially applied without new technologies or training programs.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eContributors:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHS, YG and ALR conceptualized this study. NK and VGCD managed the study subject enrollment and project administration. DNR acquired funding, facilitated laboratory investigations and overall study management. ALR and HS performed retrospective data curation and analysis. ALR wrote the first draft, HS and YG did review and editing. WMT, MJI and AY have provided critical comments on validation and analysis. ALH, JMF and RBSK edited the manuscript and provided intellectual inputs. All the authors have read and approved the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Sharing Statement:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data relevant to this retrospective analysis are available in the Supplementary Material.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of interests:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAR and AH are co-inventors of the software package here utilized. All other authors declare no conflicts of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRole of the funding source\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was funded by the Indian Council of Medical Research. The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors wish to acknowledge the support of the Human Reproduction Research Centre staff of AIIMS, New Delhi, and PGIMER, Chandigarh, who were involved in the screening, enrolment, and follow-up of the subjects.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eBeck KD et al. Quantitative analysis of cellular inflammation after traumatic spinal cord injury: evidence for a multiphasic inflammatory response in the acute to chronic environment. \u003cem\u003eBrain\u003c/em\u003e\u003cstrong\u003e133\u003c/strong\u003e, 433-447 (2010).\u003c/li\u003e\n \u003cli\u003eSugimoto MA, Sousa LP, Pinho V, Perretti M, Teixeira MM. Resolution of inflammation: What Controls Its Onset? \u003cem\u003eFront. Immunol.\u003c/em\u003e\u003cstrong\u003e7\u003c/strong\u003e, 160 (2016).\u003c/li\u003e\n \u003cli\u003eNaggara O, et al. Analysis by or dichotomizing continuous variables is inadvisable: an example from the natural history of unruptured aneurysms. \u003cem\u003eAm J Neuroradiol\u003c/em\u003e. \u003cstrong\u003e32,\u003c/strong\u003e 437-440 (2011).\u003c/li\u003e\n \u003cli\u003eRivas AL, et al. Nature and consequences of biological reductionism for the immunological study of infectious diseases. \u003cem\u003eFront Immunol\u003c/em\u003e\u003cstrong\u003e8,\u003c/strong\u003e 612 (2017). \u0026nbsp;\u003c/li\u003e\n \u003cli\u003eKuppusamy P, Prusty RK, Kale DP. High-risk pregnancy in India: Prevalence and contributing risk factors - a national survey-based analysis. \u003cem\u003eJ Glob Health\u0026nbsp;\u003c/em\u003e\u003cstrong\u003e13\u003c/strong\u003e, 04116 (2023).\u003c/li\u003e\n \u003cli\u003ePensotti A, Bertolaso M, Bizzarri M. Is cancer reversible? Rethinking carcinogenesis models-a new epistemological tool. \u003cem\u003eBiomolecules\u003c/em\u003e\u003cstrong\u003e13\u003c/strong\u003e, 733 (2023).\u003c/li\u003e\n \u003cli\u003eItaliani P, et al. Profiling the course of resolving vs. persistent inflammation in human monocytes: the role of IL-1 family molecules. \u003cem\u003eFront. Immunol.\u003c/em\u003e\u003cstrong\u003e11\u003c/strong\u003e, 1426 (2020).\u003c/li\u003e\n \u003cli\u003eNanda SK, Rivas AL, Trochim WM, JD Deshler. Emphasis on Validation in Research: A Meta-Analysis. \u003cem\u003eScientometrics\u003c/em\u003e\u003cstrong\u003e48\u003c/strong\u003e, 45\u0026ndash;64 (2000).\u003c/li\u003e\n \u003cli\u003eMor G. Introduction to the immunology of pregnancy. \u003cem\u003eImmunol Rev\u003c/em\u003e. \u003cstrong\u003e308\u003c/strong\u003e, 5-8 (2022).\u003c/li\u003e\n \u003cli\u003eKempaiah P, et al. Decoding immuno-competence: a novel analysis of complete blood cell count data in COVID-19 outcomes. \u003cem\u003eBiomedicines\u003c/em\u003e\u003cstrong\u003e12\u003c/strong\u003e, 871 (2024).\u003c/li\u003e\n \u003cli\u003ePappa T, et al. Personalized, disease-stage specific, rapid identification of immunosuppression in sepsis. \u003cem\u003eFront. Immunol\u003c/em\u003e\u003cstrong\u003e15\u003c/strong\u003e, 1430972 (2024).\u003c/li\u003e\n \u003cli\u003eVerma JS, et al. Multicellular immunological interactions associated with COVID-19 infections. \u003cem\u003eFront Immunol\u003c/em\u003e\u003cstrong\u003e13\u003c/strong\u003e, 794006 (2022)\u003c/li\u003e\n \u003cli\u003eLibertin CR, et al. Data structuring may prevent ambiguity and improve personalized medical prognosis. \u003cem\u003eMol Aspects Med\u003c/em\u003e\u003cstrong\u003e91\u003c/strong\u003e, 101142 (2023).\u003c/li\u003e\n \u003cli\u003eCappuccio A, et al.\u0026nbsp;Deciphering the combinatorial landscape of immunity. \u003cem\u003eElife\u003c/em\u003e\u003cstrong\u003e9\u003c/strong\u003e, e62148 (2020).\u003c/li\u003e\n \u003cli\u003eBunaciu RP et al. Potential for subsets of wt-NPM1 primary AML blasts to respond to retinoic acid treatment. Oncotarget. \u003cstrong\u003e9\u003c/strong\u003e, 4134-4149 (2018).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Methods","content":"\u003cp\u003eInclusion and ethics\u003c/p\u003e\n\u003cp\u003eIn compliance with the 1964 Helsinki Declaration and later amendments, this study was approved by the Ethics Committees of the two participating institutions: (1) All India Institute of Medical Sciences (AIIMS), New Delhi, India vide approval no. IEC/NP-339/2010; and (2) the Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh, Punjab, India vide approval No. 10/4815.\u003c/p\u003e\n\u003cp\u003eStudy design and participants\u003c/p\u003e\n\u003cp\u003eThis was a is a multicenter, investigator-initiated, retrospective, longitudinal, exploratory study conducted in (a) AIIMS, New Delhi and (b) PGIMER, Punjab. This study comprised lymphocyte counts at two time-points (i.e. early second trimester and 3 months after first sampling) during pregnancy and pregnancy outcomes (newborn weight and gestational length).\u003c/p\u003e\n\u003cp\u003eCapture and description of leukocyte-derived complex (combinatorial) indicators\u003c/p\u003e\n\u003cp\u003eA retrospective longitudinal analysis of pregnancy and birth-related data was performed with blood samples collected from pregnant females attending the antenatal clinics of the Obstetrics and Gynecology departments of either the AIIMS, New Delhi, India, or PGIMER, Chandigarh, Punjab, India. To be selected, candidates should: be / present with / belong to: (a) 19\u0026ndash;30 years old primigravida, (b) not later than in their second trimester, (c) 8.0\u0026ndash;13.0 g/dl of hemoglobin; (d) 18-22 body mass index, (e) middle socioeconomic indicators, and (f) provide informed consent to participate in a study approved by the Institutional Ethical Committees. Exclusion criteria included: (a) any metabolic disease, (b) malignancy, (c) heart disease, (d) any infectious disease, and (e) any chronic morbidity.\u003c/p\u003e\n\u003cp\u003eOut of 429 candidate pregnancies (which included 858 observations), 558 observations (269 candidate pregnancies) were rejected due to incomplete data. The remaining 300 observations were gathered from 160 pregnancies that contributed up to two temporal observations. However, 108 of such observations lacked information on neonates. Thus, the prognostic evaluation was conducted with 131 pregnancies, which generated 192 prepartum observations that helped determine, retrospectively, whether immunological (prepartum) maternal observations predicted birth-related adverse outcomes. Low birth weight (\u0026lt; 2500 g) and preterm birth (gestation length \u0026lt; 259 days) were described as \u0026ldquo;double risks\u0026rdquo;. Supplementary Data Table 1 reports the data of this retrospective study, indicating the blood total leukocyte count (TLC/mm\u003csup\u003e3\u003c/sup\u003e) and leukocyte percentages for each observation as well as serum ferritin and C-reactive protein (CRP) concentrations. Only the 192 prepartum observations were considered to conduct prognostic evaluations.\u003c/p\u003e\n\u003cp\u003eMethodological procedures\u003c/p\u003e\n\u003cp\u003eTwo methods were compared with the same data: (1) a reductionist one, which investigated each variable (cellular or not) in isolation; and (2) a non-reductionist alternative, which explored interactions among variables.\u003csup\u003e4\u0026nbsp;\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eLaboratory procedures\u003c/p\u003e\n\u003cp\u003eFive milliliter blood samples were collected and analyzed using an automated cell counter (Sysmex XN1000, Japan). Aliquots were prepared to conduct (i) complete blood cell counts and differentials, and (ii) serological studies (including the concentrations of ferritin and CRP).\u003c/p\u003e\n\u003cp\u003eValidation\u003c/p\u003e\n\u003cp\u003eTo explore the construct validity of the method, immunology-based data partitioning was compared to independent variables (ferritin and CRP) \u0026ndash;an exploration that also considered statistical validity. To explore the internal and external validity (repeatability across metrics and across populations), blood samples collected from individuals residing in two (New Delhi and Punjab) territories of India were investigated with several data structures.\u003c/p\u003e\n\u003cp\u003eDetection and evaluation of complex immunological data patterns\u003c/p\u003e\n\u003cp\u003eComplex data structures were created and analyzed as described elsewhere.\u003csup\u003e10-13\u003c/sup\u003e Interactions among blood leukocyte were explored with indicators created by a proprietary algorithm (US patent 10,429,389 B2), which reveals patterns that facilitate data partitioning into three or more data groups that (a) partially or totally express non-overlapping intervals of continuous data, and (b) such intervals relate to qualitatively different outcomes, such as PTB and LBW. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSample-size estimation\u003c/p\u003e\n\u003cp\u003eBecause immunological responses are known to be non-linear, no previous study on the dynamic and complex immunological interactions associated with pregnancy was available to estimate sample size. However, previous studies on numerous (non-pregnancy-related) conditions have shown that, when the internal complexity and dynamics of immunological responses are evaluated, they can reveal three or more data groups that display non-overlapping data intervals (and, consequently, achieve statistically significant differences) even when the sample size is low. Because abundant, biologically interpretable, longitudinal data patterns have been observed even when \u003cem\u003en\u003c/em\u003e\u0026lt;101,\u003csup\u003e12-13\u003c/sup\u003e it was expected that a sample with \u003cem\u003en\u003c/em\u003e\u0026gt;130 was likely to show distinct data patterns, if they existed, and also offer abundant opportunities to personalized patterns of temporal data directionality. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eStatistics\u003c/p\u003e\n\u003cp\u003eStatistical tests included comparisons among proportions and medians, which were conducted with the chi-square and Mann-Whitney tests, respectively. Such analyses were performed with a commercial package (Minitab LLC, 2024). P-values less than 0.05 were considered statistically significant.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Inflammatory stages, Immune complexity, Adverse birth outcomes, Low birth weight, Preterm birth, Pattern recognition, Prognosis, Reductionism ","lastPublishedDoi":"10.21203/rs.3.rs-6335868/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6335868/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eInflammation is a multi-stage process that affects all humans. While differentiating inflammatory stages may have broad medical applications, no real-time, easily implemented method currently differentiates early (or acute) from late (or chronic) stages. Here, an algorithm expected to detect unlimited inflammatory phases and associated outcomes was explored within reproductive health \u0026ndash;a large medical field that lacks personalized prognostics. Maternal (pre-partum) blood data were utilized to prognosticate adverse birth outcomes: low birth weight and pre-term birth (double risks). Immunological and/or clinical non-overlapping data patterns distinguished five inflammatory stages: \u0026lsquo;early inflammation\u0026rsquo; (one neutrophil-predominant subset, no double risks), \u0026lsquo;late inflammation\u0026rsquo; (two mononuclear-predominant subsets, one with 16.6%, the other with 83.3% double risks), and \u0026lsquo;no inflammation\u0026rsquo; (two lymphocyte-predominant subsets with no double risks). Because construct, internal, external and statistical validity aspects were documented, this method may support earlier and personalized interventions of pregnancies at risk. To that end, prospective studies should determine its generalizability.\u003c/p\u003e","manuscriptTitle":"Differentiation of inflammatory stages: prepartum blood data patterns detect at least five inflammatory stages that differ in their birth-related adverse outcomes","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-01 09:03:31","doi":"10.21203/rs.3.rs-6335868/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"59477061-5ae4-4451-ad7f-4d813cf50379","owner":[],"postedDate":"April 1st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":46470200,"name":"Health sciences/Health care/Prognosis/Pregnancy outcome"},{"id":46470201,"name":"Health sciences/Biomarkers/Prognostic markers"},{"id":46470202,"name":"Health sciences/Medical research/Translational research"}],"tags":[],"updatedAt":"2025-04-02T20:45:27+00:00","versionOfRecord":[],"versionCreatedAt":"2025-04-01 09:03:31","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6335868","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6335868","identity":"rs-6335868","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}