Vitamin D and hypophosphatemia in patients with anorexia nervosa and avoidant/restrictive food intake disorder: a case control study

Background Refeeding hypophosphatemia (RH) is a common complication of nutritional restoration in malnourished individuals, yet clear risk stratification remains elusive. Individuals with anorexia nervosa (AN) and avoidant/restrictive food intake disorder (ARFID) may be deficient in vitamin D, an important component of dietary phosphorus absorption in the gut. The relationship between vitamin D and RH in AN and ARFID is unknown. Therefore, the aims of this study were to (1) report rates of low serum 25-hydroxy vitamin D and RH in AN and ARFID; (2) describe associations between phosphorus and variables associated with RH identified in extant literature; (3) examine the relationship between 25-hydroxy vitamin D and RH and (4) investigate moderation by vitamin D between variables of interest and phosphorus level. Method Analyses included retrospective chart review of 307 individuals admitted to the ACUTE Center for Eating Disorders and Severe Malnutrition with a diagnosis of AN or ARFID. Variables of interest included admission laboratory values (vitamin D level, comprehensive metabolic panel, hemoglobin, point-of-care blood glucose), anthropometric measures (weight, body mass index [BMI], % ideal body weight [IBW]), age, duration of illness, length of stay, feeding method, and serum phosphorus nadir. Pearson and Spearman rank correlation, one-way ANOVA, and regression analyses were used to determine the relationship between variables and serum phosphorus. Results Over 1/3 of the sample (35.3%) had serum phosphorus levels ≤ 2.9 mg/dL. There were no significant differences between groups in phosphorus nadir ( p = .17, η 2 = 0.12) or hypophosphatemia ( p = .16, ϕc = 0.11). Thirty-five (35%) of individuals with ARFID were either deficient or insufficient in vitamin D, compared to 29% of individuals with AN. Individuals with AN had significantly higher mean vitamin D levels compared to those with ARFID ( p = .03; η 2 = 0.015). Nadir phosphorus showed a positive association with weight, BMI, %IBW, potassium, and calcium on admission, and a negative association with length of stay, hemoglobin, and total number of tube-fed days. Higher levels of 25-hydroxy vitamin D moderated the relationship between serum phosphorus nadir and weight on admission ( p = .0004). Conclusion Individuals diagnosed with ARFID are as nutritionally fragile as those with AN regarding vitamin D and RH. The negative feedback loop involving vitamin D that maintains phosphorus homeostasis may play a role in the development of RH in AN and ARFID.