Insulin injection technique.

For many years after the introduction of insulin the recommended injection technique was to raise a skin fold and insert the needle at an angle of 45°. With the advent of shorter 12-13 mm needles these two instructions were no longer thought necessary; the belief was that a full depth perpendicular injection would allow consistent deposition of insulin into subcutaneous adipose tissue.' This technique is shown in educational material provided by the syringe manufacturers for patients and nurses and indeed is the recommended technique for use of the recently introduced and increasingly popular pen shaped injectors. Further support for the method recently appeared in the patients' magazine of the British Diabetic Association.2 Nevertheless, evidence is increasing that the perpendicular technique may be far from ideal for many patients; it may be contributing to the variability of absorption that so impedes attempts to improve blood glucose control further. Frid and Linden have shown by computed tomography that in non-obese patients the subcutaneous fat in the thigh and abdomen is often less than 10mm thick.3 This meant that patients using the recommended injection technique had been giving themselves painless intramuscular injections. In a further study that used computed tomography in non-obese patients similar results were found with a mean (SE) depth of adipose tissue in the upper lateral thigh of 5 (1) mm in men and 11 (2)mm in women, albeit in a relatively small number of patients.4 Spraul et al showed with ultrasonography that the deltoid and abdominal subcutaneous fat layer was below 10 mm in all but two of 13 male volunteers, including some who were 6verweight.5 It may be concluded, therefore, that many injections given by patients using the perpendicular injection technique will deposit insulin into muscle at least on an intermittent basis. Is this of clinical importance? If the pharmacokinetic behaviour of injected insulin varies with the nature of the tissue into which it is deposited-that is, fat, muscle, or peritoneal cavity-then this would clearly be reflected in different rates of appearance of insulin in the blood and thereby in control of blood glucose. There have been conflicting conclusions from previous studies on the relative absorption rates of soluble (regular) insulin from subcutaneous and intramuscular injection sites, several showing no difference6 7 and others showing more rapid absorption from muscle.89 These earlier studies did not control for depth of injection by direct measurement of subcutaneous tissue, but some more recent work has used modern imaging techniques to allow accurate placement of the insulin. This has shown a 50% greater absorption of soluble insulin from an injection into a superficial thigh muscle when compared with injection into a subcutaneous site, but no such difference could be shown with injections into these two tissues in the abdomen.4 In this study the insulin depots were placed by using computed tomography to control the depth ofinjection. In another study comparing the rates of absorption of soluble insulin after true subcutaneous, superficial intramuscular, and deep intramuscular injections ultrasonography was used to define tissue depths.5 Similar results were found in that even superficial intramuscular injection resulted in more rapid absorption of the insulin. The new evidence shows, then, that if insulin is injected accidentally and intermittently into the superficial layers of muscle this will contribute to day to day variability in the control of blood glucose. Furthermore, the true rate of absorption of soluble insulin from subcutaneous fat is slower and even more unphysiological than was previously thought. Thus the time for the plasma insulin concentration to reach its peak is about 60-80 minutes,5 while the time to 50% absorption of an insulin depot as assessed by isotopic methods was greater than 180 minutes.4 Does the more rapid absorption of insulin from muscle confer a metabolic advantage over subcutaneous delivery? Theoretically hypoglycaemia excursions of the blood glucose at meal times and the risk of faster absorption would reduce between meals. Vaag et al recently showed more physiological absorption profiles at meal times with lower rates of absorption at five hours after injection when intramuscular injection was compared with true subcutaneous injection.'0 Perhaps of greater importance was their observation that there was less variability of blood glucose when it was monitored over a more prolonged period as evidenced by mean coefficients of variation of 33% after intramuscular injection and 43% after subcutaneous injection. But exercise had a greater potentiating and therefore unphysiological effect on intramuscular insulin absorption, an effect that must be of some concern if this route was used regularly by patients. Less attention has been paid to the pharmacokinetics of the intermediate acting insulin preparations with respect to their rate of absorption from different tissues. This is somewhat surprising because as a group these preparations constitute more than 60% of all insulin prescribed and are the origin of much of the day to day variability in the rate of insulin absorption." A recent study has shown that isophane insulin is more rapidly absorbed from muscle than from subcutaneous tissue and that variability from day to day and between patients was greater after intramuscular injection.'2 Overnight insulin absorption is often too rapid to provide adequate blood glucose control before breakfast, and thus true subcutaneous injection of isophane insulin results in an absorption profile that is better suited to providing overnight basal insulin delivery. '3 The coefficient of variation of the rate of absorption from subcutaneous tissue was lower than in previous studies at around 18%; possibly the lack of control for depth with resultant intermittent intramuscular injection may have been the cause of higher variability in past studies.9 Clearly the intramuscular route cannot be recommended for the intermediate acting insulins. Might pain from intramuscular injection mitigate against its long term clinical use? In practice this would seem not to be the case, at least in most patients. Patients in recent studies have noted almost uniformly that the superficial intramuscular route was no more uncomfortable than subcutaneous injection.35 In addition it would seem likely that many diabetic patients have been unwittingly injecting intramuscularly for many years without recognised problems.3 So we need to re-evaluate the recommendations for insulin injection. In order to reduce variability of absorption and to provide an adequate basal insulin supply overnight the extended acting insulins should be injected at an angle into a raised skin fold. The currently available 12-13 mm needles would be acceptable for this purpose. On the basis of present evidence the deliberate injection of soluble insulin into muscle