The article quoted by 'Geezer' repeats more simply much that is written in this paper: It's a good explanation as to the why people can put on weight though the latter part is a promotion for the efficacy of levimir as a background insulin.
http://dmh.mo.gov/docs/medicaldirector/ ... rticle.pdf
Here's part of the the abstract: with a my comments in blue
. Insulin-associated weight gain may result from a reduction of blood glucose to levels below the renal threshold without a compensatory reduction in calorie intake
, (ie once levels are reduced below about 10 glucose is no longer spilled from the urine but people have got unconsciously used to eating more in an attempt to fuel the body so when their levels are normalised they tend to carry on doing so)
a defensive or unconscious increase in calorie intake caused by the fear or experience of hypoglycaemia,
( The fear of hypos causes some people eat when there levels begin to fall, they may also eat far more than necessary to treat hypos )
or the ‘unphysiological’ pharmacokinetic and metabolic profiles that follow subcutaneous administration.
(normal insulin is released through the portal vein, not by absorption through the skin. Theres a world of difference between injecting some insulin when we think we need it and the finely balanced feedback loop of insulin/glucagon in a working system, the paper discusses why this aspect can lead to weight gain)
There is, however, scope for limiting insulin-associated weight gain. Strategies include limiting dose by increasing insulin sensitivity through diet (
(bit of a catch 21 here, the author describes how weight loss through calorie reduction increases insulin sensitivity. For my self it's a matter of watching the scales and intervening by reducing calories before any weight gain becomes significant) and
exercise (i
exercise in particular improves insulin sensitivy both short term and long term, indeed the body can get glucose into the muscle cells without insulin during exercise) or by using adjunctive anorectic or insulin-sparing pharmacotherapies such as pramlintide or metformin.
( seems to help some people on insulin ; more so Met in the UK)
Insulin replacement regimens that attempt to mimic physiological norms should also enable insulin to be dosed with
maximum efficiency
(MDI and even more so pumps try to do this but they are a long way from perfect replication, even a pump together with a continuous monitoring system is not an artificial pancreas : http://en.wikipedia.org/wiki/Artificial_pancreas