Serum serotonin and its association with glycaemic indices and haemoglobin in adults with migraine: a comparative cross-sectional study Authors: Anusha Jaganathan, Kalpana Radhakrishnan, Mohamed Azarudeen Musthak Ali, Robert Wilson Sundaram, Arunan Subbiah Published: 2026 ABSTRACT AbstractBackgroundMigraine is a neurovascular disorder involving serotonergic, vascular, metabolic and haematological pathways. Peripheral serotonin has been implicated in migraine biology, but its relationship with glycaemic and haematological markers remains unclear. This study evaluated serum serotonin in migraine and examined associations with clinical, glycaemic and haematological variables.MethodsThis comparative cross-sectional study included 74 adults with migraine diagnosed using International Classification of Headache Disorders, 3rd edition criteria and 82 non-migraine controls from a tertiary-care teaching hospital. Demographic data, anthropometry, blood pressure, pulse rate, fasting glucose, HbA1c, estimated average glucose, haemoglobin, thyroid-stimulating hormone and serotonin were recorded. Group comparisons used t-tests and age-adjusted analysis of covariance. Pearson and partial correlations assessed serotonin associations. Multiple regression used estimated average glucose in the primary model and HbA1c in a sensitivity model, as estimated average glucose is derived from HbA1c.ResultsAmong 156 participants, the migraine group was older and had higher blood pressure, pulse rate, fasting glucose and HbA1c/estimated average glucose, with lower haemoglobin. Serum serotonin was lower in migraine than control participants (10.72 ± 5.42 vs. 88.43 ± 6.26 ng/mL; p < 0.001), remaining significant after age adjustment. Serotonin correlated inversely with HbA1c/estimated average glucose, fasting glucose, blood pressure and pulse rate, and positively with haemoglobin. In multivariable analysis, estimated average glucose, haemoglobin and age were independently associated with serotonin.ConclusionsAdults with migraine showed lower peripheral serotonin and altered glycaemic, haemodynamic and haematological profiles. These exploratory findings require validation in longitudinal studies with detailed migraine phenotyping.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12883-026-05088-w.Keywords: Migraine, Serotonin, Peripheral serotonin, Estimated average glucose, HbA1c, Haemoglobin, Cross-sectional study Introduction Migraine is a common and disabling neurovascular disorder characterized by recurrent attacks of moderate-to-severe headache, frequently accompanied by nausea, photophobia and phonophobia, with or without aura [1, 2]. Contemporary models conceptualize migraine as a disorder of interacting neural, vascular and inflammatory networks, involving cortical hyperexcitability, trigeminovascular activation, altered brainstem modulation and changes in sensory processing [1–4]. Among the mediators implicated in these pathways, serotonin and calcitonin gene-related peptide have received particular attention because of their biological relevance and therapeutic implications [5, 6]. Serotonin has a long-standing but complex relationship with migraine [5, 6]. Multiple serotonin receptor subtypes are expressed in trigeminovascular and brainstem pathways, and several antimigraine therapies act through serotonergic mechanisms [5, 6]. However, peripheral serotonin is not a direct measure of central serotonergic neurotransmission. Serum serotonin represents a peripheral measure of 5-hydroxytryptamine and may serve as a peripheral correlate of vascular, metabolic or inflammatory changes rather than a specific marker of central serotonin activity. Increasing evidence also suggests that migraine may be influenced by systemic metabolic and haematological factors. Glycaemic dysregulation, insulin resistance, obesity and altered energy metabolism have been linked to migraine susceptibility and severity in previous studies [7, 8]. Serotonin is also involved in metabolic regulation, including body-weight regulation, insulin secretion and glycaemic control [9, 10]. Similarly, lower haemoglobin, iron deficiency and nutritional factors may contribute to headache burden through impaired oxygen delivery, altered monoamine synthesis and changes in cortical excitability [11, 12]. These observations support the possibility that peripheral serotonergic variation in migraine may coexist with systemic metabolic and haematological changes. Despite these developments, relatively few studies have examined serum serotonin together with routine glycaemic indices and haemoglobin in the same migraine cohort. Such routinely available markers may help contextualize peripheral serotonergic changes in adults with migraine, although they cannot establish causality in a cross-sectional design. Therefore, this study aimed to compare serum serotonin levels between adults with migraine and non-migraine controls and to explore the associations of serum serotonin with glycaemic indices, haemodynamic variables and haemoglobin. Methods Study design and setting This was a comparative cross-sectional study conducted in a tertiary-care teaching hospital. Adults attending neurology and general outpatient clinics were screened and allocated to migraine or non-migraine control groups based on clinical assessment. The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Ethics Committee of SRM Medical College Hospital and Research Centre. Participants and eligibility criteria Adults aged 18–60 years were eligible for inclusion. The migraine group included participants with a diagnosis of migraine according to the International Classification of Headache Disorders, 3rd edition criteria, confirmed by a neurologist after clinical evaluation [13, 14]. The control group included adults without a history of recurrent primary or secondary headache disorders. The control group is described as non-migraine controls rather than age-matched controls because the final sample showed a significant age difference between groups. The exclusion criteria were known diabetes mellitus or chronic kidney disease; thyroid dysfunction under treatment; anaemia or other haematological disorders already receiving therapy; chronic inflammatory or autoimmune disease; pregnancy or lactation; current use of drugs known to alter serotonin signalling, including antidepressants, antipsychotics or serotonergic analgesics; acute infection; recent surgery; or hospitalization within the preceding 4 weeks. Participants receiving major serotonin-modifying drugs were excluded. Detailed information on acute migraine medication use, preventive migraine therapy, antihypertensive medication and other drugs that may influence serotonin, blood pressure or pulse rate was not uniformly recorded; this was treated as residual confounding in interpretation. Migraine clinical characterization Migraine diagnosis was confirmed according to ICHD-3 criteria by a neurologist. The available clinical records captured the diagnosis of migraine but did not uniformly capture migraine subtype, aura status, chronic versus episodic classification, monthly headache days, attack frequency, attack severity, migraine duration, acute/preventive medication exposure or ictal/interictal sampling phase. The manuscript therefore avoids phenotype-specific conclusions. The availability of these migraine-characterization variables is summarized in Supplementary Table S1. Clinical and anthropometric assessment Demographic data, anthropometry and clinical variables were recorded using a structured pro forma. Height and weight were measured using calibrated equipment, and body mass index was calculated as kg/m². Blood pressure was measured in the sitting position after at least 5 min of rest using a validated automated device, consistent with office blood-pressure measurement recommendations [15]. Three readings were obtained at 1-2-minute intervals, and the average was used for analysis. Pulse rate was recorded at the time of blood-pressure measurement. Blood sampling and laboratory measurements Venous blood samples were collected from all participants between 08:00 and 10:00 h after an overnight fast of 8–10 h. Therefore, the glucose variable is reported as fasting blood glucose throughout the manuscript. Samples were processed in an ISO-accredited clinical laboratory according to standard operating procedures and internal quality-control protocols. Fasting blood glucose and thyroid-stimulating hormone were measured using an automated chemistry analyser. Haemoglobin was measured using an automated haematology analyser. HbA1c was determined by high-performance liquid chromatography. Estimated average glucose was calculated using the American Diabetes Association/ADAG formula: . Because eAG is a positive deterministic linear transformation of HbA1c, correlation coefficients between serum serotonin and HbA1c are mathematically identical to those between serum serotonin and eAG. Therefore, HbA1c and eAG were interpreted as equivalent glycaemic indices for correlation-based analyses and were not treated as independent predictors in the same regression model. Serum was separated and stored at -20 °C until serotonin analysis. Serum 5-hydroxytryptamine concentrations were quantified using a commercially available human serotonin/5-HT enzyme-linked immunosorbent assay kit based on a competitive immunoassay format, following the manufacturer’s protocol. Standards, quality-control material and participant samples were run in duplicate, absorbance was read according to the kit protocol, and concentrations were calculated from the standard curve and reported in ng/mL after applying the relevant dilution factor. Laboratory personnel were blinded to group allocation. The serotonin units and group values reported in this manuscript were verified against the original assay calculation sheets. Data management and statistical analysis Data were entered into a password-protected participant-level database and analysed using IBM SPSS Statistics, version 27. All inferential analyses reported in this manuscript, including Pearson correlations, partial correlations, age-adjusted analysis of covariance and multivariable regression, were performed from the participant-level dataset. No inferential analysis was computed from aggregated summary statistics. Where an earlier descriptive output had exported dispersion as standard error of the mean, SD = SEM × √n was used only to express descriptive values as mean ± standard deviation for manuscript presentation; this display conversion was not used to create participant-level observations or to calculate correlations, ANCOVA or regression results. Continuous variables are presented as mean ± standard deviation. Normality was assessed using the Shapiro-Wilk test and visual inspection of histograms. Between-group comparisons were initially performed using independent-samples t-tests. Because age differed significantly between groups, age-adjusted analysis of covariance was applied with migraine status as the fixed factor and age as the covariate. Pearson correlation coefficients were calculated to assess associations between serum serotonin and clinical-biochemical variables. Partial correlations were performed after adjustment for pulse rate. Multiple linear regression was used to identify independent predictors of serum serotonin, following general reporting principles for regression analysis in medical research [16]. Because estimated average glucose is mathematically derived from HbA1c, eAG and HbA1c were not entered together in the primary model. The primary model included nine predictors: age, BMI, fasting blood glucose, estimated average glucose, haemoglobin, thyroid-stimulating hormone, systolic blood pressure, diastolic blood pressure and pulse rate. A sensitivity model replaced estimated average glucose with HbA1c. Model assumptions were assessed using residual plots, tolerance, variance inflation factor and Durbin-Watson statistics. Model fit was reported using R², adjusted R², F statistic and overall model p-value. A two-tailed p-value < 0.05 was considered statistically significant. Results Descriptive analysis A total of 156 participants were included, comprising 74 adults with migraine and 82 non-migraine controls. The descriptive comparison of demographic, anthropometric, haemodynamic and biochemical variables is presented in Table 1. Participants with migraine were significantly older than controls (37.81 ± 11.24 vs. 30.13 ± 11.56 years, p < 0.001). Height, weight and BMI did not differ significantly between groups. The height, weight and BMI descriptive values were verified against the participant-level data. Table 1Descriptive analysis of study participantsVariableMigraine (n = 74) Mean ± SDNon-migraine (n = 82) Mean ± SDMean difference (95% CI)p-valueAge (years)37.81 ± 11.2430.13 ± 11.567.68 (4.07 to 11.29)< 0.001Height (cm)149.19 ± 18.24150.11 ± 18.28-0.92 (-6.70 to 4.86)0.74Weight (kg)65.36 ± 13.4465.13 ± 13.630.23 (-4.05 to 4.52)0.91BMI (kg/m²)30.30 ± 7.9929.58 ± 8.530.71 (-1.90 to 3.33)0.52Systolic BP (mmHg)125.59 ± 15.97116.35 ± 9.289.24 (5.04 to 13.44)< 0.001Diastolic BP (mmHg)81.86 ± 9.7977.32 ± 7.904.55 (1.71 to 7.38)0.002Pulse rate (bpm)84.93 ± 10.8178.39 ± 7.436.54 (3.57 to 9.51)< 0.001Fasting blood glucose (mg/dL)104.93 ± 30.0087.82 ± 11.2017.12 (9.76 to 24.47)< 0.001HbA1c (%)5.49 ± 0.984.76 ± 0.370.73 (0.49 to 0.97)< 0.001eAG (mg/dL)110.96 ± 28.0989.89 ± 10.6321.07 (14.18 to 27.96)< 0.001Haemoglobin (g/dL)12.71 ± 1.9414.24 ± 1.61-1.53 (-2.09 to -0.96)< 0.001Serotonin (ng/mL)10.72 ± 5.4288.43 ± 6.26-77.71 (-79.56 to -75.87)< 0.001HbA1c values are presented in percentage units consistent with the ADA/ADAG eAG calculation. Descriptive values were checked against the participant-level dataset. Where earlier exported output displayed SEM, SD conversion was used only for manuscript presentation and not for inferential analyses Descriptive analysis of study participants HbA1c values are presented in percentage units consistent with the ADA/ADAG eAG calculation. Descriptive values were checked against the participant-level dataset. Where earlier exported output displayed SEM, SD conversion was used only for manuscript presentation and not for inferential analyses The migraine group had higher systolic blood pressure, diastolic blood pressure, pulse rate, fasting blood glucose, HbA1c and estimated average glucose than controls. Haemoglobin was lower in the migraine group. Serum serotonin was markedly lower in participants with migraine compared with controls (10.72 ± 5.42 vs. 88.43 ± 6.26 ng/mL; p < 0.001). Age-adjusted group comparison Age-adjusted analysis of covariance was performed because the migraine group was older than the non-migraine control group. After age adjustment, the lower serum serotonin level in the migraine group remained statistically significant (Table 2). The direction of the group differences in glycaemic, haemodynamic and haematological variables also remained unchanged after adjustment for age. Table 2Age-adjusted analysis of covariance for key study variablesVariableAge-adjusted mean difference95% CIAdjusted p-valuePartial eta-squaredSystolic BP (mmHg)7.803.70 to 11.90< 0.0010.081Diastolic BP (mmHg)3.700.90 to 6.500.0100.043Pulse rate (bpm)5.502.50 to 8.50< 0.0010.078Fasting blood glucose (mg/dL)14.807.40 to 22.20< 0.0010.091HbA1c (%)0.630.38 to 0.88< 0.0010.127eAG (mg/dL)18.0010.80 to 25.20< 0.0010.126Haemoglobin (g/dL)-1.33-1.90 to -0.76< 0.0010.117Serotonin (ng/mL)-72.03-74.20 to -69.85< 0.0010.887The large partial eta-squared for serum serotonin indicates a marked group difference in this sample; this result is interpreted cautiously because biomarker separation of this magnitude requires independent replication and analytical verification Age-adjusted analysis of covariance for key study variables The large partial eta-squared for serum serotonin indicates a marked group difference in this sample; this result is interpreted cautiously because biomarker separation of this magnitude requires independent replication and analytical verification Correlation between serotonin and clinical-biochemical variables The bivariate relationships between serum serotonin and clinical-biochemical parameters are summarized in Table 3. Serum serotonin showed significant inverse correlations with glycaemic and haemodynamic markers. Because eAG is calculated directly from HbA1c using a positive linear equation, the Pearson correlation coefficients for HbA1c and eAG are identical. HbA1c/eAG showed the strongest inverse glycaemic association with serotonin (r = -0.441, p < 0.001), followed by fasting blood glucose (r = -0.351, p < 0.001), systolic blood pressure (r = -0.338, p < 0.001), pulse rate (r = -0.330, p < 0.001) and diastolic blood pressure (r = -0.229, p = 0.003). Haemoglobin showed a positive association with serotonin (r = + 0.387, p < 0.001). BMI was not significantly correlated with serotonin. Table 3Pearson correlation matrix for serum serotonin and clinical-biochemical variablesVariablePearson rp-valueInterpretationFasting blood glucose-0.351< 0.001Higher glucose associated with lower serum serotoninHbA1c-0.441< 0.001Same coefficient as eAG because eAG is derived from HbA1ceAG-0.441< 0.001Mathematically equivalent glycaemic association to HbA1cHaemoglobin+ 0.387< 0.001Higher haemoglobin associated with higher serum serotoninSystolic BP-0.338< 0.001Inverse haemodynamic associationDiastolic BP-0.2290.003Significant inverse associationPulse rate-0.330< 0.001Significant inverse associationBMI-0.0590.46Not significanteAG is a positive linear transformation of HbA1c; therefore, Pearson correlation coefficients with serum serotonin are identical for HbA1c and eAG Pearson correlation matrix for serum serotonin and clinical-biochemical variables eAG is a positive linear transformation of HbA1c; therefore, Pearson correlation coefficients with serum serotonin are identical for HbA1c and eAG Partial correlation analysis After controlling for pulse rate, the inverse associations between serum serotonin and glycaemic markers persisted. Because eAG is derived from HbA1c, the pulse-rate-adjusted partial correlation coefficients for HbA1c and eAG are identical. HbA1c/eAG remained the strongest adjusted glycaemic correlate of serotonin (partial r = -0.409, p < 0.001). Fasting blood glucose continued to show a significant inverse association, and haemoglobin remained positively associated with serotonin. The results of the partial correlation analysis are presented in Table 4. Table 4Partial correlation analysis controlling for pulse rateVariablePartial rp-valueInterpretationFasting blood glucose-0.312< 0.001Still significant after pulse-rate adjustmentHbA1c-0.409< 0.001Same coefficient as eAG because eAG is derived from HbA1ceAG-0.409< 0.001Mathematically equivalent adjusted association to HbA1cHaemoglobin+ 0.366< 0.001Positive adjusted associationeAG is a positive linear transformation of HbA1c; therefore, partial correlation coefficients adjusted for pulse rate are identical for HbA1c and eAG Partial correlation analysis controlling for pulse rate eAG is a positive linear transformation of HbA1c; therefore, partial correlation coefficients adjusted for pulse rate are identical for HbA1c and eAG Regression analysis for predictors of serum serotonin In the multivariable regression model, estimated average glucose, haemoglobin and age were independently associated with serum serotonin (Table 5). Estimated average glucose was negatively associated with serotonin, haemoglobin was positively associated with serotonin, and age was negatively associated with serotonin. HbA1c was evaluated only in a sensitivity model because estimated average glucose is mathematically derived from HbA1c. The primary model demonstrated acceptable collinearity diagnostics, with all variance inflation factor values below 2.0. Model fit statistics for the primary model were R² = 0.392, adjusted R² = 0.355, F(9,146) = 10.46, p < 0.001, and Durbin-Watson statistic = 1.93. Table 5Multiple linear regression model for serum serotoninPredictorBSE95% CI for BStandardized betap-valueToleranceVIFeAG-0.470.13-0.73 to -0.21-0.2780.00070.701.42Haemoglobin+ 5.321.41+ 2.54 to + 8.10+ 0.2600.00030.811.24Age-0.740.22-1.17 to -0.31-0.2260.0010.761.31Fasting blood glucose-0.090.05-0.19 to + 0.01-0.0540.080.621.62BMI-0.120.19-0.50 to + 0.26-0.0250.520.931.08TSH+ 2.731.90-1.02 to + 6.48+ 0.0910.150.951.05Systolic BP-0.180.11-0.40 to + 0.04-0.0640.100.671.48Diastolic BP+ 0.080.14-0.20 to + 0.36+ 0.0180.570.661.52Pulse rate-0.210.14-0.49 to + 0.07-0.0520.140.721.39Primary model: dependent variable = serum serotonin. The primary model included nine predictors; therefore, the corrected model fit is reported as F(9,146). HbA1c was excluded from the primary model because eAG is derived from HbA1c. Sensitivity model replacing eAG with HbA1c showed a similar inverse direction for glycaemic status without altering the interpretation of haemoglobin and age Multiple linear regression model for serum serotonin Primary model: dependent variable = serum serotonin. The primary model included nine predictors; therefore, the corrected model fit is reported as F(9,146). HbA1c was excluded from the primary model because eAG is derived from HbA1c. Sensitivity model replacing eAG with HbA1c showed a similar inverse direction for glycaemic status without altering the interpretation of haemoglobin and age Discussion In this comparative cross-sectional study, adults with migraine showed lower serum serotonin levels and differences in selected glycaemic, haemodynamic and haematological variables compared with non-migraine controls. Serum serotonin was inversely correlated with HbA1c/eAG, fasting blood glucose, blood pressure and pulse rate, and positively correlated with haemoglobin. In the multivariable model, estimated average glucose, haemoglobin and age were independently associated with serum serotonin. These findings suggest that peripheral serotonergic variation in migraine may coexist with systemic metabolic and haematological differences; however, the cross-sectional design does not allow causal or mechanistic conclusions. The observed lower serum serotonin levels in participants with migraine are consistent with the long-recognized involvement of serotonergic pathways in migraine biology [17, 18]. Serotonin receptors are expressed in trigeminovascular and brainstem pathways, and serotonergic mechanisms remain clinically relevant through the therapeutic actions of triptans, ditans and related agents [5, 6, 18]. Alterations across tryptophan-serotonin-related pathways have also been discussed in migraine-related mechanisms [19]. Nevertheless, serum serotonin requires cautious interpretation. Peripheral 5-hydroxytryptamine is a serum-based peripheral measure and may not directly reflect central serotonergic neurotransmission; therefore, the present findings represent peripheral serotonergic associations rather than direct evidence of central serotonin depletion [20, 21]. The magnitude of the between-group serum serotonin difference was large. Although the units and group values were checked against the original assay calculation sheets, a peripheral biomarker separation of this degree should be interpreted cautiously and replicated in independent cohorts, ideally using detailed assay reporting, external quality control and complementary analytical platforms such as liquid chromatography-mass spectrometry or platelet-poor plasma serotonin measurement. Accordingly, the present manuscript avoids presenting serum serotonin as a diagnostic biomarker and interprets the result as an exploratory group difference requiring validation. The associations between serotonin and glycaemic indices are biologically plausible but exploratory. Serotonin participates in peripheral metabolic regulation, including glucose handling and insulin secretion [9, 10], while migraine has been linked to altered energy metabolism, insulin resistance and metabolic comorbidity [7, 8, 22, 23]. In the present study, HbA1c/eAG showed the strongest inverse glycaemic association with serum serotonin. Because estimated average glucose is derived from HbA1c, the two markers were interpreted as mathematically equivalent glycaemic indicators rather than independent biological constructs in correlation analyses or in the same primary regression model. The positive association between haemoglobin and serotonin may reflect an interaction between haematological status and headache biology. Lower haemoglobin may contribute to headache susceptibility through reduced oxygen delivery, altered monoamine synthesis or nutritional factors, as suggested by previous work linking iron-deficiency anaemia and chronic headache disorders [11, 12]. This interpretation remains speculative because iron indices, ferritin, transferrin saturation, vitamin B12, folate and dietary intake were not measured. Although participants receiving treatment for known anaemia were excluded, undiagnosed iron deficiency or nutritional deficiency may have contributed to the observed haemoglobin difference. The migraine group was older than the control group. This age pattern may reflect the hospital-based tertiary-care recruitment setting, in which adults with persistent, recurrent or complicated headache presentations may be more likely to seek specialist evaluation. It may also reflect coexistence of metabolic or haemodynamic comorbidities in adults attending neurology services. Therefore, the age distribution in this sample does not represent population prevalence patterns of migraine, which commonly peak in younger and middle adult age groups. Age was retained as an adjustment variable, and the lower serum serotonin level in the migraine group remained significant after age adjustment. This study has several limitations. First, the single-centre cross-sectional design prevents causal inference and limits generalizability. Second, migraine phenotyping was limited; data on aura status, disease duration, monthly headache days, attack frequency, headache severity, chronic versus episodic migraine status and sampling during ictal or interictal phase were not uniformly available, although such clinical characterization is important for interpreting migraine biomarker studies [1, 13, 14]. Third, serum serotonin was measured at a single time point, and temporal fluctuations across migraine phases could not be assessed. Fourth, total serum serotonin was measured rather than platelet-poor plasma serotonin, central serotonin metabolites, tryptophan-kynurenine metabolites, CGRP or inflammatory markers. Fifth, unmeasured confounders including diet, physical activity, sleep, menstrual status, psychiatric symptoms, inflammatory phenotype, iron indices and nutritional status may have influenced the results. Sixth, although major serotonin-modifying drug exposure was excluded, detailed acute migraine medication use, preventive migraine therapy, antihypertensive medication and other drugs that could influence serotonin, blood pressure or pulse rate were not uniformly recorded. Finally, the unusually large serum serotonin group difference and the variability in anthropometric measures require confirmation in independent samples with full assay documentation and prospectively harmonized data capture. Overall, the findings are exploratory and hypothesis-generating. They support further investigation of peripheral serotonin in relation to metabolic and haematological markers in migraine, but they do not establish serotonin as a diagnostic biomarker or prove a mechanistic pathway. Future multicentre longitudinal studies with age-matched controls, detailed migraine phenotyping, ictal/interictal sampling, medication documentation and broader biomarker panels are warranted to validate these observations. Conclusions Adults with migraine showed lower serum serotonin levels and differences in glycaemic, haemodynamic and haematological markers compared with non-migraine controls. Serum serotonin was associated with estimated average glucose, haemoglobin and age in the multivariable model. Because estimated average glucose is a mathematical transformation of HbA1c, these two glycaemic indices were interpreted as equivalent measures rather than independent predictors in the same model. Because of the cross-sectional design, age imbalance, single-time-point serotonin measurement, limited migraine phenotyping and the need for independent verification of the large serum serotonin difference, these findings are best interpreted as exploratory and hypothesis-generating. Further longitudinal, multicentre studies are required to clarify whether peripheral serotonin is consistently associated with metabolic and haematological profiles in migraine. Supplementary Information Supplementary Material 1: Table S1. Migraine phenotype and medication-characterization variables. Acknowledgements The authors would like to thank SRM Medical College Hospital and Research Centre, SRM Institute of Science and Technology, Kattankulathur, for providing technical resources. Clinical trial registration Not applicable. This was an observational comparative cross-sectional study and did not involve a healthcare intervention. Authors' contributions Concept and design : Anusha Jaganathan, Kalpana Radhakrishnan, Arunan Subbiah Manuscript drafting : Anusha Jaganathan, Statistical analysis and administrative support : Mohamed Azarudeen Musthak Ali, Robert Wilson Sundaram Critical revision of the manuscript for important intellectual content : Kalpana Radhakrishnan, Mohamed Azarudeen Musthak Ali, Robert Wilson Sundaram Supervision : Kalpana Radhakrishnan All authors have reviewed the final version of the manuscript , approved it for submission , and agreed to be accountable for all aspects of the work. Funding Open access funding provided by SRM Institute of Science and Technology for SRMIST – Medical & Health Sciences. This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Data availability The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. Ethical approval was obtained from the Institutional Ethics Committee of SRM Medical College Hospital and Research Centre, SRM Institute of Science and Technology, Tamil Nadu, India. IEC/IRB No.: 9003. Written informed consent was obtained from all participants before enrolment. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1: Table S1. Migraine phenotype and medication-characterization variables. Data Availability Statement The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.