The search for a reliable post-ictal biomarker has spanned nearly five decades. Serum prolactin, first proposed by Trimble in 1978,1 requires sampling within 20 min, can also be elevated after syncope, and achieves only moderate sensitivity.2 Creatine kinase peaks too late (48–72 h) for acute decision-making, though may have value in an outpatient setting.3 Lactate, though informative normalizes within 1–2 h and is clinically nonspecific,4 inflammatory markers such as WBC and temperature are also elevated5 but again are clinically nonspecific. The identification of phosphate as the most profoundly altered electrolyte after tonic–clonic seizures by Nass et al. in 20196 opened a new avenue and was subsequently validated in emergency department cohorts in Switzerland, Portugal, and even in canine seizure presentations.7-9 We therefore read with interest the study by Binks and colleagues (Volume 10, Issue 6, December 2025, Pages 2025–2031) extending this evidence into first seizure clinics and neurology consults,10 and sought to review this finding within our own service. We retrospectively reviewed 96 cases presenting to the first seizure clinic at our hospital who had bloods taken at the initial presentation to our Emergency Department. Of these, 89 had a diagnosis made in clinic: 72 epileptic seizures (50 epilepsy, 22 single seizures), 11 non-epileptic attacks, 5 syncopal episodes, and 1 other. Phosphate was available in 77.5% of diagnosed cases, higher than the 49% reported by Binks et al.,10 suggesting routine measurement is both feasible and may already often occur. Mean phosphate was 0.92 mmol/L (SD 0.30) in epileptic seizures and 0.96 mmol/L (SD 0.38) in non-epileptic events. A phosphate below 0.8 mmol/L was present in 37.5% (21/56) of epileptic seizures and 23.1% (3/13) of non-epileptic events. This difference did not reach statistical significance in our cohort (Fisher's exact test, p = 0.52). Mean lactate values (2.93 vs. 1.81 mmol/L) were comparable to those reported by Binks et al. Logistic regression using lactate alone yielded an AUC of 0.793 (95% CI 0.311–0.945) for discriminating epileptic seizures, while the addition of phosphate did not improve discrimination (AUC 0.731, 95% CI 0.527–0.948), likely reflecting the small number of non-epileptic cases with both lactate and phosphate available for model fitting (n = 11). This contrasts with the larger cohort in Binks et al. where a combined lactate–phosphate–calcium model achieved an AUC of 0.825. Exploratory analysis of diagnostic thresholds in our cohort suggested that a lower cutoff of <0.75 mmol/L may optimize the trade-off between sensitivity and specificity, achieving 92.3% specificity and 93.8% PPV (Youden's J = 0.191) compared with 76.9% specificity at the <0.8 mmol/L threshold used by Binks et al., while the <0.6 mmol/L cutoff proposed by Barras et al. offered no further gain in specificity but halved sensitivity (12.5% vs. 26.8%). When pooled with the data from Binks et al., the combined estimate showed low phosphate in 32.8% (38/116) of epileptic seizures versus 11.6% (5/43) of non-epileptic events (Fisher's exact test, p = 0.009). This combined effect size is consistent with the odds ratios of 3.5–5.5 reported across the literature.7, 9, 10 The proposed mechanisms behind post ictal hypophosphatemia, catecholamine-driven cellular uptake, respiratory alkalosis from post-ictal hyperventilation, and glycolytic phosphate consumption during intense muscular contraction6 predict that hypophosphatemia should be largely specific to convulsive seizures and absent in non-epileptic attacks (NEA), where these processes do not occur. The low rate of hypophosphatemia in NEA observed by both Binks et al. (1/15) and in our cohort (3/13) supports this. A limitation of our analysis is the absence of ictal onset timing in many cases, preventing replication of the 6-h diagnostic window identified by Binks et al. and earlier by Nass et al. We also did not categorize seizures as convulsive or not, which likely increased our mean phosphate levels; however, it is true to routine clinical practice. Nevertheless, this data strengthens the case for adding serum phosphate to initial investigations in TLoC. Given its low cost, widespread availability, and a practical diagnostic window, phosphate represents an accessible addition to the TLoC assessment toolkit—particularly where episodes are unwitnessed or the history is unclear. Prospective studies should look to evaluate combined biomarker panels incorporating phosphate, lactate, calcium, and CK focusing on evaluating thresholds. TA and AG have no conflicts of interest to disclose. RHT has received honoraria from Angelini, Bial, Biocodex, Eisai, Jazz, LivaNova, Paladin, Neuraxpharm, Sanofi, Takeda, and UCB Pharma. Meeting support from Angelini, UCB Pharma. Unrestricted research funding from Angelini. Joint working partnership with UCB Pharma. CI on BioHaven and Lundbeck RCTs. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Atkinson et al. (Mon,) studied this question.