Albumin and pre-albumin levels do not reflect the nutritional status of female adolescents with restrictive eating disorders.

AIM: Albumin and pre-albumin are frequently used as nutritional markers in clinical practice. We examined whether serum albumin and pre-albumin were predicted by body mass index (BMI), hydration and/or inflammation in female adolescents with a recently diagnosed restrictive eating disorder (RED).

METHODS

This was a retrospective study of female adolescents with RED from 2002 to 2011. Low albumin and pre-albumin levels were defined as <3.5 g/dL and <20 mg/dL, respectively. We assessed inflammation using the erythrocyte sedimentation rate (ESR) and dehydration using the haematocrit levels.

RESULTS

We included 75 females with a mean age of 15.2 years and 64% had a BMI Z score of <-2. The mean albumin and pre-albumin levels were 4.8 g/dL and 22.2 mg/dL, respectively, with 24% of the children having low pre-albumin and none having low albumin levels. The stepwise multiple regression for albumin identified ESR and haematocrit as significant predictors, which explained 14.8% of the variance. Age was the only significant predictor for pre-albumin, which explained 15.3% of the variance.

CONCLUSION

Albumin, but not pre-albumin, levels were primarily predicted by low-grade inflammation and hydration, but not by BMI. These markers should not be used to assess nutritional status in adolescents with RED.