Reappraisal of the optimal fasting time for insulin tolerance tests in mice

Deborah Carper, Marine Coué, Claire Laurens, Dominique Langin, Cedric Moro



Most studies routinely use overnight or 6 h of fasting before testing metabolic glucose homeostasis in mice. Other studies used empirically shorter fasting times (<6 h). We attempted to determine the shortest fasting time required for optimal insulin responsiveness while minimizing metabolic stress.


A course of fasting for up to 24 h (0, 2, 4, 6, 12, and 24 h) was conducted in C57Bl/6J male mice. Body weight, metabolic parameters, and insulin tolerance were measured in each experimental group. The organs were collected at the same time on separate occasions and glycogen and metabolic gene expression were measured in the liver and skeletal muscle.


Our data show that blood glucose levels do not significantly change during a 6 h fast, while plasma insulin levels decrease to similar levels between 2 h and 6 h of fasting. During overnight (12 h) and 24 h fasts, a robust decrease in blood glucose and plasma insulin was observed along with a profound depletion in liver glycogen content. Insulin tolerance was comparable between baseline and 6 h fasts while 4 h and 6 h fasts were associated with a greater depletion of liver glycogen than 2 h fasts, impacting the glucose counter-regulatory response. Fasting induced progressive weight loss that was attenuated at thermoneutrality. Fasting longer than 4 h induced major body weight loss (>5%) and significant changes in catabolic gene expression in the liver and skeletal muscle.


Collectively, these data suggest that 2 h of fasting appears optimal for the assessment of insulin tolerance in mice as this duration minimizes major metabolic stress and weight loss.