Our study is one of the first to explore the association of multiple restrictive ED characteristics (malnutrition, duration of ED symptoms, and pre-morbid BMI status) with co-morbid anxiety and depression symptoms. The majority of the study participants met or exceeded the cutoffs for clinical anxiety and depression, demonstrating a high rate of clinically significant anxiety and depression in adolescents and young adults with restrictive EDs. We did not find that degree of malnutrition was associated with symptoms of anxiety and depression, after controlling for age, sex assigned at birth, sexual orientation, ED diagnosis, and use of psychiatric medication. Though this was not as we hypothesized, this highlights the importance of other factors that contribute to depression and anxiety in this vulnerable population. Longer duration of ED illness was associated with clinically meaningful and significantly worse depression and clinically meaningful worse anxiety symptoms, supporting the need for rapid identification and treatment of EDs and its co-morbid mental health disorders. However, contrary to our hypothesis, those who had a pre-morbid BMI ≥ 75th percentile had lower anxiety and depression scores than those who had BMIs < 75th percentile prior to weight loss. Though the association with anxiety was not statistically significant, the trend noted is clinically relevant, and an important factor to study further.
It is critical for us to determine whether weight restoration should remain the top priority in ED treatment, or whether optimal treatment includes balanced focus on physical and psychological components of EDs. Some studies of patients with AN have found an association between severity of malnutrition and anxiety/depressive symptoms,19 and that weight restoration led to improved mental health.19–23 However, like us, others have found no statistically significant relationship between degree of malnutrition and depressive symptoms24,25 or anxiety, 25 and found that depression and anxiety may persist despite weight restoration.19,39,40 Given the dangers associated with struggling with both an ED and comorbid psychiatric illness,14,16,17 these inconsistent findings highlight the need for more comprehensive ED treatment that focuses on both nutritional rehabilitation and mental health treatment.
Our study found that longer chronicity of ED symptoms was related to more severe depression symptoms and identified a potentially similar trend with anxiety. This finding is important to consider in treatment for EDs as negative affect and internalizing symptoms can contribute to more severe ED symptoms,15 and depression is also associated with lower rates of ED recovery over time.42,43 Longer duration of ED illness is overall associated with worse prognosis and treatment outcomes (e.g., weight gain and ED recovery) among individuals diagnosed with AN.13,15,41,44,45 However, research on the effect of duration of illness on co-morbid anxiety and depression for individuals with restrictive EDs is generally limited. Our finding is in line with a previous study which showed that depression and anxiety scores were significantly higher for females with AN who had a longer duration of illness.41 And findings on duration of ED illness consistently suggest that longer duration raises the risk for worse psychological symptoms, worse prognosis, treatment outcomes, and even mortality.46 Our study highlights the need for early identification and intervention for individuals with restrictive EDs due to the possibility of worsening depression and anxiety with a longer duration of illness. Our findings also suggest that adapting therapeutic approaches based on the illness duration may be necessary given the association between the duration of illness and psychological correlates of restrictive EDs.
Lastly, our cohort of adolescents and young adults with higher pre-morbid BMIs actually had lower anxiety and depression scores at the time of their baseline surveys. Of note, we used a cut off of 75th percentile due to sample size and clinical suspicion that those with a BMI 75th to 85th percentile had similar weight-stigma experiences compared to those who met CDC criteria for overweight/obese (≥ 85th percentile). A sensitivity analysis using the cutoff of 85th percentile had similar findings as the cutoff of 75th percentile (though it was not statistically significant, possibly due to sample size). Potentially, this finding is actually a reflection of improved anxiety and depression after weight loss, as those with higher weights tend to experience more weight stigma and elevated mental health concerns prior to weight loss.47,48 Youth with overweight or obesity often experience weight stigma as weight-based victimization, teasing, and bullying,49 and those with restrictive EDs with a higher pre-morbid weight status may experience social reinforcement of their disordered eating behaviors; often patients receive praise as they begin their weight loss.50 Other studies have shown that patients with severe obesity experience improved anxiety and depression with a psychosocial weight loss intervention51 or bariatric surgery.52,53 However, we were not able to find studies that have observed the effects of malnutrition, ED treatment, or weight gain on the levels of anxiety and depression in this unique population of patients with higher pre-morbid weights and a restrictive ED. The temporal relationship of the mental health comorbidities in these patients could be enlightening and provide insight on effective ED treatment approaches and priorities for this population.
Our cohort included participants with not just AN, but also other EDs affected by malnutrition, such as OSFED and ARFID. Additionally, our use of eBMI, based on each patient’s previous growth trajectory, allowed us to capture the effects of relative malnutrition. This is vital for capturing the spectrum of patients who are at different weights. However, our study was not without limitations. The sample size was not large enough to include all variables that could be of clinical significance, so for our third model, a stepwise approach was used to include the most statistically influential variables. There was also large variability in degree of malnutrition, length of time under our care when recruited, and previous experiences with intensive ED treatment. This may have a role in why a cross-sectional association between malnutrition and anxiety and depression was not found. Additionally, as a cross-sectional study, we were unable to capture individuals’ changes in anxiety and depression in regards to changes in weight during treatment, a critical question in need of clarification. Therefore, a longitudinal study could be vital for answering many questions. It could demonstrate how weight restoration and rapidity of treatment aid in the treatment of anxiety and depression symptoms in patients with EDs. Additionally, a longitudinal study would allow us to explore how anxiety and depressive symptoms change throughout treatment for all patients, and especially those with high pre-morbid weight statuses.