Bulimia nervosa (BN) is characterized by dysregulated eating behaviour and present data suggest adipokines may regulate food intake. We investigated a possible association between BN and adipokine levels and hypothesized that plasma (P)-adiponectin would be elevated and P-leptin and P-leptin-adiponectin-ratio would be reduced in women with BN.

The study was designed as a cross-sectional study with a longitudinal arm for patients with BN. Plasma-adiponectin and leptin was measured in 148 female patients seeking psychiatric ambulatory care and 45 female controls. Fifteen patients were diagnosed with BN and the remaining with other affective and anxiety disorders. P-adiponectin and P-leptin levels were compared between patients with BN, patients without BN and controls. At follow-up 1–2years later, adipokines were reassessed in patients with BN and the Eating Disorder Examination Questionnaire was used to assess symptom severity.

P-adiponectin was elevated in patients with BN at baseline and at follow-up when compared to patients without BN and controls (P<0.004 and <0.008 respectively). The difference remained significant after controlling for body mass index. P-adiponectin was correlated to symptom severity at follow-up in patients with BN without morbid obesity (ρ=0.72, P<0.04). P-leptin-adiponectin-ratio was significantly lower in patients with BN compared to controls (P<0.04) and P-leptin non-significantly lower.

Findings indicate a stable elevation of P-adiponectin in women with BN. P-adiponectin at follow-up correlates to eating disorder symptom severity in patients without morbid obesity, indicating that P-adiponectin should be further investigated as a possible potential prognostic biomarker for BN.