All classifications not built on pathogenesis become inadequate sooner or later
- First Online:
- Cite this article as:
- Iafusco, D., Galderisi, A., Lombardo, F. et al. Diabetologia (2011) 54: 1583. doi:10.1007/s00125-011-2113-9
KeywordsGestational diabetes Glucokinase MODY-2
Hyperglycemia Adverse Pregnancy Outcome
Large for gestational age
Small for gestational age
To the Editor: We read with interest the paper by E. A. Ryan  concerning the new criteria suggested by the Hyperglycemia Adverse Pregnancy Outcome (HAPO) study for the diagnosis of gestational diabetes . Ryan showed that applying these criteria would result in a doubling of the number of pregnant women diagnosed with gestational diabetes without a clear demonstration of the benefits derived from this new classification. As he observes, maternal obesity represents a stronger predictor of large-for-gestational-age (LGA) babies than glucose levels in all except the highest glucose category [1, 3]. This suggests that not all gestational hyperglycaemia has the same aetiology. As was the case in paediatrics, where for many decades we wrongly diagnosed all children with hyperglycaemia as having type 1 diabetes , we risk including in this generic categorisation of gestational diabetes the undetected monogenic forms that are often underdiagnosed [5, 6]. For example, in those patients belonging to the lowest glucose categories of HAPO those with MODY-2 obtain no benefit from the treatment of their hyperglycaemia. In the 11 patients with MODY-2 that we followed up in our unit the only similarity was fasting glucose >5.5 mmol/l (100 mg/dl) (D. Iafusco, A. Galderisi, A. Cocca and F. Prisco, unpublished results). This is why we fully agree with Ryan’s proposal of tighter diagnostic criteria that would reduce the number of individuals diagnosed, thereby targeting the diagnosis more carefully. However, his suggestion has an additional unexpected benefit: being more rigorous in the identification of ‘gestational diabetes and of its forms’ represents a milestone in the prevention not only of LGA births and complicated deliveries , but also of small-for-gestational-age (SGA) births.
Diagnosis according to the HAPO criteria increases the risk of treating women with insulin when they do not need it, because some women will be affected by an unrecognised form of monogenic diabetes such as MODY-2. MODY-2 occurs in 2–5% of the population with diabetes , or possibly even more frequently . Treating a mother affected by MODY-2 avoids the birth of an LGA child caused by maternal hyperglycaemia: as happens for other forms of gestational diabetes, high maternal glucose levels may be a trigger for insulin secretion in the fetus that results in an increase in birthweight and in increased likelihood of trauma during delivery. Insulin treatment is essential for this form of diabetes. The situation is different if both mother and child carry a genetic mutation predisposing to MODY-2: in this case the fetus has poor insulin secretion that, if not increased by exposure to high maternal glycaemic levels, may result in inadequate intrauterine growth and the birth of an SGA baby. The SGA baby has greater postnatal risks than the LGA baby—low birthweight for gestational age represents an independent risk factor for neonatal mortality .
Considering this risk, it seems safer not only to avoid an unhelpful increase in the number of pregnant women treated for poorly defined ‘gestational diabetes’ but also, and more importantly, to avoid treatment being given to mothers and babies who do not need it, and in whom treatment can be life-threatening.
This opens up once more the great debate of modern medicine in the choice between a tendency towards generalisation and systematisation of diseases in order to make clear and effective diagnostic and therapeutic flowcharts, and the increasing tendency to distinguish diseases not by their phenotypes but by their pathogenesis, thus providing ‘tailored therapy’.
At the moment, with the exception of obesity (patients with MODY-2 are not obese), we cannot clinically distinguish women with MODY-2 from those with other causes of gestational diabetes. We have to investigate the family history of the woman, use ‘tighter’ diagnostic criteria (identical to the ones used for the non-pregnant population) and, if MODY-2 is suspected we have to search for the genetic diagnosis of the mother and the prenatal diagnosis of the child.
Choosing to provide intensive treatment can really make a difference in helping an affected mother to produce a healthy child. Where the gynaecologist and the diabetologist agree on less rigorous metabolic control, however, this may prevent an SGA birth.
The next step in the management of gestational diabetes has to be, as happened for paediatric hyperglycaemia, the identification of pathogenetic mechanisms in order to provide better care for the right patients.
We thank F. Barbetti (Department of Molecular Genetics of Monogenic Diabetes, University of Tor Vergata, Rome, Italy), L. Sacchetti (Department of Biochemistry and Medical Biotechnologies, University Federico II of Naples, Naples, Italy), and M. Pinelli and F. Acquaviva (Department of Cellular and Molecular Pathology, University Federico II of Naples, Naples, Italy) for their useful suggestions.
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.