Circadian and sleep-wake dysregulation have been implicated in episode recurrence and clinical heterogeneity in bipolar disorder (BD), supporting peripheral circadian markers as candidates for further investigation. This study examined the peripheral expression of selected circadian rhythm-related gene mRNAs and antisense long non-coding RNAs (lncRNAs), alongside corresponding serum protein levels and self-report-based rhythm phenotyping, in euthymic patients with BD. Sixty euthymic patients with BD (20 BD-1, 20 BD-2, and 20 with antidepressant-associated hypo/manic switch [ASHH], treated as an exploratory group) and 20 healthy controls were enrolled. Euthymia was verified by HAM-D < 7 and YMRS < 6, and chronotype and rhythm phenotypes were assessed using the MEQ, SRM-5, and BRIAN. Morning fasting blood was collected (08:00-10:00) with the morning medication dose withheld; because sampling was confined to a single morning window, the reported values represent morning transcript abundance rather than circadian rhythmicity, amplitude, or phase position. Three core clock gene mRNAs (CIART, ARNTL, CLOCK) and two antisense lncRNAs (CRY1AS, PER2AS) were quantified by qRT-PCR (2^-ΔΔCt, β-actin reference), and serum CIART, CRY1, PER2, and ARNTL proteins were measured by ELISA. Between-group comparisons used nonparametric tests (Mann-Whitney U; Kruskal-Wallis H with Bonferroni post hoc), and correlations used Spearman's coefficient with Benjamini-Hochberg FDR correction. Compared with healthy controls, patients showed statistically significantly higher expression of the lncRNAs PER2AS (r = 0,57) and CRY1AS (r = 0,49) and the mRNAs CIART (r = 0,35) and CLOCK (r = 0,29), all p ≤ 0,007, whereas ARNTL mRNA did not differ (p = 0,654). In exploratory subgroup analyses, differences were concentrated in BD-2 and ASHH: BD-2 showed the highest expression and BD-1 the lowest, with BD-1 not differing from controls for any marker. Serum protein levels did not differ between groups or subgroups, and MEQ, BRIAN, and SRM-5 scores did not differ; however, a sensitivity analysis excluding shift workers revealed a chronotype difference masked in the full sample. Across 174 correlational tests, none survived FDR correction (lowest q = 0,063); nominal uncorrected associations are reported as preliminary only. In euthymic BD, peripheral expression of selected circadian rhythm-related markers-particularly the antisense lncRNAs PER2AS and CRY1AS and the mRNAs of CIART and CLOCK - differed from healthy controls, whereas corresponding serum protein levels and behavioral rhythm phenotypes did not. This divergence across measurement levels can be considered a substantive finding, and the subgroup patterns are exploratory and hypothesis-generating. Longitudinal, episode-inclusive, and subtype-stratified studies incorporating objective circadian phase markers and repeated sampling are needed to establish the significance of peripheral transcript-level differences in BD.HIGHLIGHTS Peripheral circadian gene mRNAs and antisense lncRNAs are profiled in euthymic bipolar subtypes.PER2AS and CRY1AS lncRNAs and CIART and CLOCK mRNAs are higher in patients than controls.Exploratory analyses: transcript differences concentrate in BD-2 and ASHH; not BD-1.Serum CIART, CRY1, PER2, and ARNTL protein levels by ELISA do not differ by group.Behavioral rhythm scores do not differ; correlations do not survive FDR correction.
