Methods:  Children fulfilling 4 or more ACR classification criteria for systemic lupus erythematosus who were between the ages of 8 and 12 years (girls) or 11 and 14 years (boys) whose pubertal development was Tanner 1 or early Tanner 2 were consented and enrolled in a prospective longitudinal observational study to assess what pubertal events and hormone changes affected lupus disease activity as measured by the SELENA SLEDAI. They were evaluated every 3 months until after the completion of puberty. Flares were defined as mild to moderate (MMF) if SLEDAI increased by 3 to 5 points, and severe (SvF) if 6 or more compared to the previous 3 monthly assessment. If SLEDAI decreased by at least 3 points, the visit was considered improved (IMP). Multivariate fitting after stepwise model slelection was used to correlate hormone changes with disease activity. Generalized equalizing equations were used to compare changes in disease activity over time and multiple regression analysis to compare factors potentially influencing disease activity. In some analyses, girls with pre-pubertal onset SLE (PreP) were compared to girls with post-pubertal (age 15 or more, PostP) SLE from two contemporaneously collected cohorts (HI Brunner).

Results:  Of 55 girls with pre-pubertal onset lupus 100% had ds-DNA antibody vs 70% with SLE onset age >15 (p=0.001). PreP were twice as likely to have anemia (37 vs 18%, p=0.01) as PostP, and they had more nephritis at onset (60 vs 40%, p<0.02). Serum concentration of FSH, LH and estradiol did not correlate temporally with flare. In 270 visits from 65 pubertal children with SLE (55 F, 10 M), 48 had MMF (17.8%) and 19 had SvF (7.04%); 24.8% had any flare. In 63 visits (23.3%) SLEDAI decreased >3 points (IMP). 49% of visits had change in activity sufficient to consider alterring treatment. 16 entered puberty and 17 transitioned from Tanner 2-3. Flares were associated with Tanner stages 2 and 3 (p=0.008) vs Tanner 1. No severe flares happened in/after Tanner 4. Flares correlated with rapid falls in visfatin (OR 0.072, p=0.002), high adiponectin (OR 1.20, p=0.0046), and neither estradiol nor testosterone correlated with flare. IMP correlated with testosterone (p=0.023) and followed a drop in adiponectin (0.008) and rise in prolactin (p=0.003) and resistin (p=0.026).

Conclusion:  Trasition from pre-puberty to early puberty is associated with flares in SLE, which falls after Tanner 4 is reached. Puberty is a tumultuous time for children with SLE with half the visits requiring medication adjustment because of flares and improvements. Adipokine changes in serum are temporally related to disease activity. Higher concentrations of testosterone appear to exert a calming effect on lupus disease activity.