Abstract
1. The absorption of insulin in constant and varied volumes and concentration was studied in healthy volunteers. 2. Disappearance curves of radioactivity after subcutaneous injection of labelled insulin were fitted with three models: a diffusion rate-limited, a dissolution rate-limited and a capacity-limited model. 3. For each curve and model goodness of fit (r2) and sum of weighted squared deviations (SWSD) were calculated. With constant volume and concentration r2 was greater than 0.097 for each curve. SWSDs with the second model were less than those with the first (NS) and third (P less than 0.01). 4. When volume and concentration were varied, the parameters of the first and third model also varied, arguing against the validity of these models. 5. It is concluded that each of the three models, although yielding an accurate description of the observed data is too simple conceptually. Suggestions for more appropriate modelling are discussed.