Jul 19, 1986 - Division of Chemical Pathology,. United Medical and Dental Schools .... the Royal College of Pathologists is seriously considering proposals for ...
204
BRITISH MEDICAL JOURNAL
measurements of relative blood flow changes are valid. The "peak flow" data shown in the paper should therefore be converted to percentage increases over baseline values; when re-rexpressed in this way, are there still significant differences between the diabetic and non-diabetic groups? Secondly, although plethysmography (and presumably laser Doppler flowmetry) can be used to monitor blood flow at a single site for two to three hours, repeated readings over "the next two to three days" are probably meaningless. Thirdly, Impaired microvascular hyperaemic although needle insertion seems a reasonable form response to minor skin trauma in of trauma, the same cannot really be said of gentle type I diabetes warming to 44°C, which after all is the temperature SIR,-Dr G Rayman and colleagues made some of the skin probe used to measure transcutaneous interesting observations on the local hyperaemic oxygen tension.6 GARETH WILLIAMS response to minor skin trauma (needle insertion and warming) in insulin dependent diabetes (17 Endocrine Unit, May, p 1295). However, their literature search Hammersmith Hospital, may have been incomplete, as they do not mention London W12 OHS JOHN PICKUP our reports on the local hyperaemic responses to needle insertion and superficial injection of small Division of Chemical Pathology, United Medical and Dental Schools, volumes of bland liquids in diabetic and non- Guy's Hospital Campus, diabetic subjects. 14 London SEI 9RT Dr Rayman's protocol was similar to our own, except that we used photoelectric plethysmography, a I Williams G, Pickup JC, Clark AJL, Bowcock S, Cooke E, Keen H. Changes in blood flow close to subcutaneous insulin non-invasive optical technique which yields similar injection sites in stable and brittle diabetics. Diabetes 1983;32: results to the laser Doppler flowmetry method used by 466-73. the authors'; the two methods have been compared (1 2 Williams G, Pickup JC, Keen H. The use of photoelectric March, p 620). Injection depth was fixed at 7-8 mm plethysmography to assess local vasoactivity of drug&-C1ia Sci under the centre of the photoelectric plethysmo183;28:77-8. graphy probe by inserting the needle along an angled 3 Williams G, Pickup JC, Bowcock 5, Cooke E, Keen H. Subcutaneous aprotinin causes hyperaemia: a possible mechtrack drilled through a perspex block cemented to the anism by which aprotinin improves control in some diabetic side of the probe. Injection of 0-1-0 15 ml of saline or patients. Diabetologia 1983;24:914. insulin diluent in normal subjects (n=6, mean age 24 Williams G. Blood flow at insulin injection sites. In: Pickup JC, 4 years) produced a local hyperaemic "flare," which ed. Briuk diabetes. Oxford: Blackwell, 1985:132-53. reached a peak representing a 1600/o increase over 5 Nijboer JA, Dorlas JC, Mahieu HF. Photoelectric plethysmobaseline flow at 2 minutes and had largely faded by 30 graphy-some fundamental aspects of the reflection and minutes. Simple needle insertion without injection transmission method. Clm PhysPPysiolMeas 1981;2:205-15. produced a response of similar amplitude and dura- 6 Railton R, Newman P, Hislop J, Harrower ADB. Reduced transcutaneous oxygen tension and impaired vascular response tion, suggesting that this acute, transient hyperaemia in Type 1 (insulin-dependent) diabetes. Diabetologia 1983;25: was a non-specific response to injury. In contrast, 340-2. there was prolonged and considerable hyperaemia after injection of local vasodilators such as prostaglandin El,l 2nicotinicacid,2aprotinin,3andinsulin.. In stable insulin dependent diabetics (n=7, mean Oral contraceptives and hepatoceliular age 22 years, mean duration of diabetes 9- years) the carcinoma response was slightly but not significantly less than in normal subjects, but in severely brittle diabetics with chronic metabolic instability (n= 8, mean age 18 years, SIR,-Your recent papers reporting an association mean duration of diabetes 10 years) the hyperaemic of long term use of oral contraceptives with liver flare was significantly flatter and shorter than in both cancer (24 May, p 1355, 1357) might be criticised other groups (figure). These findings, unlike those of for their size, their choice of control groups, and Dr Rayman and colleagues, therefore suggested that their methods of ascertainment of previous oral impaired local hyperaemic responses to needle inser- contraceptive use. On the whole, and given the tion were related to long term metabolic control rather biological plausibility of the reported associations, than the presence of diabetes itself; differences in the studies are convincing. patient selection may explain these divergences. However, death rates from primary malignant Other points should be mentioned. Firstly, as neoplasms of the liver (ICD (8th revision) 155-0; discussed (1 March, p 620), neither photoelectric ICD (9th revision) CM 155-0) have not changed in plethysmography nor laser Doppler flowmetry can women in the United States (table). This suggests measure absolute units of blood flow, although that the risk of liver cancer associated with oral contraceptive use in the United States may be lower than in Great Britain. It is my own observa. 200 tion that continuous long term use of oral contra00Stable ceptives in the United States is unusual because of 0 o ..V Brittle a widespread clinical policy of giving women "a vacation" from oral contraceptive use every two or three years. If the risk of liver cancer is related to continuous long term oral contraceptive use and if this policy of "vacations" has not been followed as
finished giving evidence, as we reported briefly in the journal (24 May, p 1400). We apologise to Professor Stewart if our report has caused him any personal distress, but we thought it right that we should report his examination as fully as we could on a matter of such great public importance that has received so much media attention over the last decade. -ED, BMJ.
Normal
.
VOLUME 293
19 JULY 1986
widely in Great Britain as in the US then oral contraceptive users in the United States would be at lower risk of liver cancer than in Great Britain. Perhaps the authors of the two papers on this topic could examine this possibility. DIANA B PETITTI School of Medicine, University of California, San Francisco, Ca 94143, USA
Bone mineral content in Polynesian and white New Zealand women SIR,-We read this paper by Dr I R Reid and others (14 June, p 1547) with considerable interest and feel able to advance an explanation for racial differences in bone mineral content. Recent work using a similar densitometer with 69 female volunteers (mean age 59 (SD 19) years) confirmed an inverse correlation between bone density and age and between strength and age (p
