The Gene That Makes GLP-1 Drugs Less Effective
High GLP-1 levels should mean GLP-1 drugs work better. For carriers of a PAM loss-of-function allele, the data suggest otherwise.
About 10% of people carry a hypomorphic variant in PAM, the gene encoding the only enzyme that amidates bioactive hormones like GLP-1. Carriers of the rarer p.S539W allele showed 52% lower PAM enzyme activity — and paradoxically, higher postprandial GLP-1 levels. Yet, despite that surplus of circulating hormone, they showed an 18% reduction in GLP-1 sensitivity. The incretin effect measured by isoglycemic clamp was statistically indistinguishable from non-carriers — the resistance only became apparent under pharmacological challenge.
In a meta-analysis of 1,119 patients across three cohorts, p.S539W carriers achieved only a −0.69% HbA1c reduction on GLP-1RAs versus −1.24% in non-carriers. Only 11.5% hit the HbA1c less than 7% target, compared to 25.3% of non-carriers. In PAM-knockout mice, gastric emptying was accelerated and refractory to exendin-4, with impaired cAMP signaling in the pylorus — pointing toward a post-receptor signaling defect. The authors describe findings "consistent with PAM LoF allele carriers exhibiting GLP-1 resistance," while noting PAM's effects on GLP-1 clearance and tissue-specific contributions remain incompletely characterized.
Metformin, sulphonylureas, and DPP-4 inhibitors showed no genotype-based response differences in the same cohorts.
Source: Genome Medicine
Two Years Gone and the Nose Might Still Come Back
In a small retrospective cohort, half of COVID smell-loss patients who started treatment more than 2 years after symptom onset still hit the threshold for clinically meaningful recovery — though with no control group, the authors caution that spontaneous recovery can't be ruled out.
A University of Alabama at Birmingham team tracked 104 patients at a tertiary smell and taste clinic from 2023 to 2025. The COVID group (n=47) showed significant TDI score gains — from 19.7 to 24.0 — while the non-COVID group showed no statistically significant change. Worth noting: the non-COVID group was a heterogeneous mix of idiopathic, traumatic, and postsurgical cases, so that comparison is less clean than it looks. The treatment protocol was olfactory training plus twice-daily budesonide nasal irrigation.
"Clinically meaningful olfactory recovery remains achievable with long-standing CRSL, even when treatment is initiated more than 2 years after onset."
— Hunsicker et al., Int Forum Allergy Rhinol, 2026
The sneaky part: The COVID group arrived with higher baseline TDI scores than the non-COVID group, suggesting milder dysfunction at the start — which means some of that recovery gap may reflect who these patients were, not just what they were treated with.
Adherence mattered — a lot, apparently. Among the delayed-treatment subgroup, 76% of compliant patients hit minimal clinically important difference versus 26% of noncompliant ones, a striking gap that needs prospective confirmation, given self-reported adherence and a wide confidence interval. The mechanism remains speculative: postviral olfactory loss may preserve more regenerative capacity in the olfactory neuroepithelium than traumatic or idiopathic causes, but that's still an open question.
Clinical takeaway: Don't write off the patient who shows up 2 years late about COVID smell loss. A trial of olfactory training plus budesonide irrigation is reasonable to offer — just counsel them that the evidence base is still retrospective, spontaneous recovery is real in this population, and sticking with it matters.
Source: International Forum of Allergy & Rhinology
Better Sleep May Depend on How Patients Relate to Time
Stronger interoceptive awareness — the ability to notice and regulate internal bodily signals — is associated with better sleep quality. That part isn't entirely surprising. What is: the relationship runs partly through how a person orients to time.
A new cross-sectional study in Frontiers in Psychology gave 152 nonclinical adults validated measures of interoceptive awareness (MAIA-2) and time perspective (ZTPI), plus single-item self-ratings of sleep and digestion quality. Among MAIA subscales, Trusting — the sense of connection and safety with one's own body — showed the strongest association with sleep (r = 0.38, p < 0.001). Digestion quality correlated positively with all eight interoceptive subscales.
Here's the sneaky part: for sleep specifically, a balanced time perspective statistically mediated the effect. High interoceptive self-regulation predicted lower deviation from temporal balance, which in turn predicted better sleep — a partial mediation for self-regulation and full mediation for attention regulation. The digestion pathway was different and more limited: it ran through past-negative orientation, not a broadly balanced time perspective, and only partially.
"Interoception provides present-moment physiological access, while time perspective situates these signals within cognitive-affective narratives of past and future."
— Klamut & Weissenberger, 2026
The proposed mechanism is the anterior insula, implicated in both interoceptive integration and subjective time perception, within the broader neurovisceral integration framework. These remain theoretical — no neural or physiological measures were collected, and the cross-sectional, self-report design means causality cannot be established in either direction.
Clinical takeaway: Mindfulness-based approaches that build interoceptive self-awareness may carry sleep benefits partly by reorienting how patients relate to their past and future — not just by calming the nervous system directly. The sample was small, predominantly female, and relied on single-item outcome measures, so treat these findings as hypothesis-generating rather than practice-changing.
Source: Frontiers
A Urine Test for Schizophrenia Might Not Be as Crazy as It Sounds
Turns out, the humble urine sample may hold candidate clues to some of psychiatry's hardest-to-diagnose conditions — and no one's really been looking there.
Researchers at Cambridge ran a two-sample Mendelian randomization analysis across seven psychiatric disorders, using genetic data as proxies for urinary metabolite levels. The result: 67 statistically significant metabolite-disorder associations, 21 of which were exclusive to a single condition.
The standout findings: lower urinary tyrosine was genetically linked to schizophrenia risk, and lower creatine tracked with bipolar disorder(BD) — both associations held up across multiple independent studies. Meanwhile, N,N-dimethylglycine emerged as a solo flag for ADHD, and pyridoxal + ferulic acid 4-sulfate showed associations with anorexia nervosa.
Here's the sneaky part: BD and schizophrenia — which get misdiagnosed as each other roughly 24% to 33% of the time — showed 22 overlapping markers plus distinct exclusive ones. The authors suggest a differential panel could eventually be worth exploring.
On mechanism, the tyrosine-schizophrenia link is consistent with the dopamine hypothesis, while creatine points toward disrupted brain energy metabolism in BD. Importantly, the authors themselves flag that these are genetically predicted associations, not validated biomarkers — calling explicitly for "experimental validation in independent cohorts."
Clinical takeaway: Nothing actionable yet, but this MR-based approach is a genuinely cost-effective way to generate testable hypotheses — and urinary metabolomics in psychiatry deserves a closer look.
Source: BMC Psychiatry