Physiologic mechanisms can predict hematologic responses to iron supplements in growing children: a computer...

Area of Science:

• Nutritional Science
• Biomedical Modeling
• Pediatric Health

Background:

• Iron deficiency is a widespread, preventable nutritional issue in developing nations.
• Numerous randomized clinical trials (RCTs) have assessed various iron supplementation strategies.
• Understanding iron metabolism is crucial for managing deficiency and its consequences.

Purpose of the Study:

• To assess if current knowledge of iron metabolism can predict hemoglobin and serum ferritin (SF) levels using a computer model.
• To determine if this model can serve as an alternative to conducting new RCTs.

Main Methods:

• Developed a computer model based on iron absorption and regulation physiology, using data from RCTs on iron Sprinkles.
• Utilized data from two RCTs in Ghana to calibrate the model for iron absorption and hemoglobin changes.
• Validated the model by predicting hemoglobin and SF concentrations in a Chinese RCT, comparing predictions with observed data.

Main Results:

• Model-predicted hemoglobin means closely matched actual values (within +/-2 g/L).
• Model-predicted SF medians were highly consistent with actual values (within +/-3 microg/L).
• Quantile-quantile plots showed strong agreement between predicted and actual hemoglobin and SF values.

Conclusions:

• The developed iron metabolism model accurately predicts hemoglobin and SF concentrations in children receiving iron Sprinkles.
• This computational model can potentially replace the necessity of conducting repetitive RCTs across diverse settings.
• The findings support the use of predictive modeling in nutritional research and intervention.