Controversy persists
re accepting the new guidelines for the screening and diagnosis of gestational
diabetes (see http://www.adips.org/). The Hyperglycaemia and Adverse Pregnancy
Outcome (HAPO) study clearly identifies a continuous risk gradient between
glucose levels achieved during a 75 g GTT in the latter part of pregnancy and a
variety of maternal and foetal outcomes. In assessing any treatment strategy, a
risk-benefit analysis is essential. The HAPO data provide very strong evidence
for the risk part of this analysis. The new diagnostic criteria recognise equal
risk levels for the fasting, one hour and two hour glucose values in the GTT.
What is less certain is the benefit of treatment particularly relating to the
fasting plasma glucose. We have good evidence from both RCTs and widespread
clinical practice for treating to targets of 5.3 mmol/L fasting and 6.7 mmol/L at
2 hours. Less well recognised or ignored, is the fact that we have for RCTs (1)
that demonstrate the benefit of treating to much lower fasting values (4.4
mmol/L) in GDM pregnancies in which the foetus is identified as being obese
based on abdominal circumference. Reference 1 reviews the evidence linking
abdominal circumference to foetal adiposity and foetal hyperinsulinaemia. Can this evidence be applied to all GDM
pregnancies? I do not think it can. Tightening glycaemic control for foetuses
who are not at risk of becoming obese may do more harm than good (foetuses with
the glucokinase mutation are an excellent example). This applies regardless of
the diagnostic criteria used to diagnose GDM. An abstract presented at last
year's reach meeting using very tight glycaemic targets demonstrated a
clinically important but non-statistically significant reduction in weight in
the intensively treated group compared to normal women. This study was
underpowered to identify this difference as significant, if it were real (3).
I think we need to stop arguing about how to diagnose GDM and begin to examine
how to optimise treatment for all the women and foetuses involved. This will
involve detailed analysis of individual neonates and not simply group counting
of large, normal or small babies.