Premenarchal anorexia nervosa: clinical features, psychopharmacological interventions, and rehospitalization analysis in a 1-year follow-up, controlled study

**Background** Anorexia nervosa (AN) is a mental health condition with a point prevalence of 2.5% in females. Premenarchal AN (onset before menarche) is rare, representing only 5% of AN cases. Previous studies have used non-univocal definitions and small samples. This study aimed to provide naturalistic evidence on psychopharmacological and clinical interventions in premenarchal vs. postmenarchal AN, and to compare rehospitalization rates over a 12-month follow-up.

**Methods** This was a case-control, observational retrospective study conducted at a third-level Italian Regional center for Feeding and Eating Disorders (FED) in Children and Adolescents. Consecutive patients hospitalized (inpatient or day-hospital) between 01/01/2016 and 31/12/2020 were included. Inclusion criteria: female cisgender, DSM-5 diagnosis of AN, informed consent. Exclusion: insufficient clinical documentation. Premenarchal AN was defined as primary amenorrhea at hospital admission. The control group had secondary amenorrhea or resumed menses. A total of 234 participants were analyzed (43 premenarchal, 191 postmenarchal). Data on demographics, clinical variables, psychopathology (SAFA scales, EDI-3, BDI-II, SCL-90-R, BUT), endocrine measures (LH, estrogen, progesterone, prolactin), treatments (SSRI, atypical antipsychotics, nasogastric tube feeding), and hospitalization details were collected. Rehospitalizations for AN within 12 months were recorded. Statistical analyses included descriptive statistics, repeated-measures ANOVA, Kaplan-Meier survival analysis, log-rank test, and Cox regression.

**Key Results** - Premenarchal AN patients were younger (mean age 12.9 vs. 16.2 years, p<0.001), had shorter duration of untreated illness (11.6 vs. 16.4 months, p=0.024), and were more likely to have restrictive AN (97.7% vs. 85.9%, p=0.093). - Psychopathology: Premenarchal patients had lower SAFA-D (depression) scores (57.0 vs. 67.2, p=0.010), lower SAFA-O (obsessive-compulsive) scores (48.1 vs. 54.3, p=0.017), and lower SAFA-P (eating disorder) scores (51.3 vs. 57.6, p=0.035). - Endocrine: LH levels were lower in premenarchal patients (0.4 vs. 2.3, p=0.009). Estrogen, progesterone, and prolactin were not significantly different. - Treatments: Premenarchal patients received SSRIs less frequently (62.8% vs. 80.1%, p=0.015). Atypical antipsychotic use was similar (66.7% vs. 66.7%, p=1.000). Nasogastric tube feeding was used in 37.2% vs. 32.5% (p=0.551). - Hospitalization: Premenarchal patients had shorter mean hospitalization (99.1 vs. 126.6 days, p=0.026) and were all treated as inpatients (100% vs. 69.1%, p<0.001). - Weight: Admission %BMI (70.7 vs. 73.1, p=0.221) and discharge %BMI (82.0 vs. 79.8, p=0.091) were not significantly different. %BMI increased significantly over time in both groups (F=51.8, p<0.001). - Rehospitalization: Cumulative freedom from rehospitalization at 12 months was 66.1% (95% CI 45.5-86.7) for premenarchal and 62.9% (95% CI 52.1-73.7) for postmenarchal (log-rank p=0.457). Cox regression showed that higher age at admission (HR=1.686, p=0.020) and day-hospital treatment level (HR=52.312, p=0.007) significantly predicted rehospitalization, but premenarchal status did not (HR=0.667, p=0.653).

**Clinical Implications** Premenarchal AN presents with distinct clinical features including lower depressive symptoms and less SSRI use, but similar weight outcomes and rehospitalization risk compared to postmenarchal AN. The higher rehospitalization risk associated with day-hospital treatment may reflect unplanned transitions between care levels. Clinicians should consider age-specific guidelines and the role of family involvement. Further research with standardized definitions is needed to develop tailored treatment protocols for premenarchal AN.