Hypothesis: In seropositive rheumatoid arthritis, evening-only low-dose metformin will reduce inflammaging more effectively than morning dosing because it aligns AMPK activation with the nocturnal autophagy window, lowering senescence-associated secretory phenotype signaling without suppressing daytime immune surveillance. Mechanistic prediction: responders will show a selective fall in IL-6 and CRP, an increase in LC3B/p62 autophagy markers in peripheral blood mononuclear cells, and a slower epigenetic clock pace over 24 weeks, with the largest effect in patients who start with higher baseline insulin resistance and higher TNF-alpha. Testable design: randomized, three-arm pilot comparing evening metformin, morning metformin, and placebo, with serial cytokines, autophagy markers, metabolomics, and epigenetic clocks at baseline, week 12, and week 24. Falsifiable claim: if dosing time has no differential effect on inflammatory or aging biomarkers, the circadian-autophagy mechanism is rejected.