Subclinical Leptospiral Infections and Antibody Waning During Pregnancy: A Population-Based Cohort Study from Rural...

Leptospirosis is a zoonotic infection with significant mortality, yet its burden is underestimated due to diagnostic limitations. The dynamics of naturally acquired anti-Leptospira antibodies, particularly during pregnancy, remain poorly understood. This study examined the seroincidence of leptospiral infection among a cohort of pregnant women and characterized longitudinal antibody kinetics in an endemic setting. Paired sera were collected from a population-based pregnancy cohort in Anuradhapura, Sri Lanka. Women were enrolled before 12 weeks of gestation and followed up between 25-28 weeks. Data on sociodemographics, self-reported febrile illness, and hospital admissions were collected. Microscopic agglutination testing (MAT) was performed against 11 Leptospira serovars. Seroincidence was defined as seroconversion from non-reactive (<1:50) to reactive (≥1:50). Antibody kinetics were assessed using Gaussian accelerated failure time and proportional odds logistic regression models. Among 3,374 recruited women, 1,338 had paired samples. Of 1,245 seronegative at baseline, 22 (1.8%) seroconverted, yielding a seroincidence rate of 52.8/1,000 person-years (95% CI: 31.2-76.8), using the ≥1:50 definition; using a ≥1:100 cutoff, the seroincidence rate was 14.4 per 1,000 person-years (95% CI: 5.3-31.4). No significant differences in pregnancy outcomes were observed between seroconverts and non-seroconverts. Among 93 women with baseline antibody titers ≥1:50, 14 (15.1%) seroreverted to <1:50, 12 (12.9%) showed a decline in titers but remained ≥1:50, 56 (60.2%) remained unchanged, and 11 (11.8%) showed an increase in antibody titers. Antibody decay rates ranged from +0.353 to -0.789 log2 units/30 days (mean: -0.074), with faster decay at higher baseline titers. This study reveals frequent, asymptomatic Leptospira exposure and rapid antibody decline during pregnancy, highlighting substantial silent transmission and the limitations of current surveillance tools in endemic settings.