Ntial confounding factors Patients’ psychological scores and nutritional status indicators were correlated with AN subtype, type of treatment, age, and Table 1. Patient (n = 155) characteristics at inclusion.Mean Age (years) 20.90 14.43 13.05 20.08 5.SD 6.16 1.46 1.55 3.24 3.22 4.71 0.8 1.14 0.16 0.Minimum 13.16 10.72 8.59 13.15 0 0.12 10.68 20.53* 0.70 0.Maximum 43.76 18.51 18.51 32.47 17.94 24.39 15.03 5.10 1.53 1.Statistical AnalysisStatistical analysis was performed on SPSS 17.0. First descriptive statistics were produced. Results are presented as means (SD). The relationships between the different 15900046 psychological scores (BDI, HAD depression, HAD anxiety, MOCI and LSAS) and the nutritional status indicators (BMI, FFMI, FMI, severity of weight loss, albumin and prealbumin levels) were tested using Pearson’s correlations. A statistical significance was considered when p was ,0.05. Potential confounding factors (minimum lifetime BMI, age and duration of illness) that might affect psychological symptoms and nutritional status, were also included in the analyses Student ttests were performed in order to check for differences in nutritional status in relation to potential confounding factors such as restrictive and binge/purging AN, presence/absence of drug treatment (antidepressants, anxiolytics). Following this, multiple linear regression models were run in order to test the link between psychiatric scale scores and the nutritional status indicators, adjusting on all confounding factors identified by univariateInclusion BMI (kg/m2) Minimum lifetime BMI (kg/m2) Maximum lifetime BMI (kg/m2) Severity of weight loss FFMI+ FMI+ Albumin level# Prealbumin level#Duration of illness (years) 4.29 12.54 1.92 1.02 0.BMI: Body Mass Index; SD: Standard Deviation. *Negative values are due to negative body fat predicted values because of the severe 1418741-86-2 chemical information emaciation of AN patients. + FFMI and FMI are obtained for 146 patients. # Albumin and prealbumin were measured for 132 patients. doi:10.1371/journal.pone.0049380.tAnorexia NervosaTable 2. Psychlogical scores at inclusion.Mean BDI HAD depression HAD anxiety MOCI LSAS 26.80 9.43 12.58 12.53 57.SD 11.73 4.52 4.35 5.80 15.Minimum 3.00 .00 2.00 1.00 26.Maximum 53.00 
18.00 21.00 28.00 95.Finally only age and medication, but not subtype of AN nor minimum lifetime BMI, were introduced into the multivariate analysis as confounding factors because they were the only 2 factors that affected the psychological scores.Multivariate analysisNone of the 
psychological scores were HIV-RT inhibitor 1 explained by the nutritional indicators except for a negative correlation between albumin level and the LSAS fear scale (p = 0.024; beta = 20.225), as in the univariate analysis. Only antidepressants explained the variability in BDI scores (p = 0.029; beta = 0.228) and anxiolytics explained the variability in the HADs depression scores (p = 0.037; beta = 0.216).SD: standard deviation; BDI : Beck Depression Inventory, HAD: Hospital Anxiety and Depression scale, MOCI : Maudsley Obsessive-Compulsive Inventory, LSAS: Liebowitz social anxiety scale. doi:10.1371/journal.pone.0049380.tDiscussionThe present study has adopted a very novel approach to investigating the relationship between nutritional status and psychological symptoms in AN, since it takes into account body composition and biological markers, and not solely weight or BMI, for the nutritional assessment, also adjusting for age and psychotropic treatment. To our knowledge it is the largest study.Ntial confounding factors Patients’ psychological scores and nutritional status indicators were correlated with AN subtype, type of treatment, age, and Table 1. Patient (n = 155) characteristics at inclusion.Mean Age (years) 20.90 14.43 13.05 20.08 5.SD 6.16 1.46 1.55 3.24 3.22 4.71 0.8 1.14 0.16 0.Minimum 13.16 10.72 8.59 13.15 0 0.12 10.68 20.53* 0.70 0.Maximum 43.76 18.51 18.51 32.47 17.94 24.39 15.03 5.10 1.53 1.Statistical AnalysisStatistical analysis was performed on SPSS 17.0. First descriptive statistics were produced. Results are presented as means (SD). The relationships between the different 15900046 psychological scores (BDI, HAD depression, HAD anxiety, MOCI and LSAS) and the nutritional status indicators (BMI, FFMI, FMI, severity of weight loss, albumin and prealbumin levels) were tested using Pearson’s correlations. A statistical significance was considered when p was ,0.05. Potential confounding factors (minimum lifetime BMI, age and duration of illness) that might affect psychological symptoms and nutritional status, were also included in the analyses Student ttests were performed in order to check for differences in nutritional status in relation to potential confounding factors such as restrictive and binge/purging AN, presence/absence of drug treatment (antidepressants, anxiolytics). Following this, multiple linear regression models were run in order to test the link between psychiatric scale scores and the nutritional status indicators, adjusting on all confounding factors identified by univariateInclusion BMI (kg/m2) Minimum lifetime BMI (kg/m2) Maximum lifetime BMI (kg/m2) Severity of weight loss FFMI+ FMI+ Albumin level# Prealbumin level#Duration of illness (years) 4.29 12.54 1.92 1.02 0.BMI: Body Mass Index; SD: Standard Deviation. *Negative values are due to negative body fat predicted values because of the severe emaciation of AN patients. + FFMI and FMI are obtained for 146 patients. # Albumin and prealbumin were measured for 132 patients. doi:10.1371/journal.pone.0049380.tAnorexia NervosaTable 2. Psychlogical scores at inclusion.Mean BDI HAD depression HAD anxiety MOCI LSAS 26.80 9.43 12.58 12.53 57.SD 11.73 4.52 4.35 5.80 15.Minimum 3.00 .00 2.00 1.00 26.Maximum 53.00 18.00 21.00 28.00 95.Finally only age and medication, but not subtype of AN nor minimum lifetime BMI, were introduced into the multivariate analysis as confounding factors because they were the only 2 factors that affected the psychological scores.Multivariate analysisNone of the psychological scores were explained by the nutritional indicators except for a negative correlation between albumin level and the LSAS fear scale (p = 0.024; beta = 20.225), as in the univariate analysis. Only antidepressants explained the variability in BDI scores (p = 0.029; beta = 0.228) and anxiolytics explained the variability in the HADs depression scores (p = 0.037; beta = 0.216).SD: standard deviation; BDI : Beck Depression Inventory, HAD: Hospital Anxiety and Depression scale, MOCI : Maudsley Obsessive-Compulsive Inventory, LSAS: Liebowitz social anxiety scale. doi:10.1371/journal.pone.0049380.tDiscussionThe present study has adopted a very novel approach to investigating the relationship between nutritional status and psychological symptoms in AN, since it takes into account body composition and biological markers, and not solely weight or BMI, for the nutritional assessment, also adjusting for age and psychotropic treatment. To our knowledge it is the largest study.