To reassure women that their babies are healthy, medicine has come up with a battery of tests that look for fetal abnormalities. Some, like amniocentesis, are invasive tests that carry risks for the baby. Others are non-invasive screening tests that use a single blood sample to look for chemicals, known as ‘markers’, in the blood to help calculate a woman’s potential risk of having a baby with a neural tube defect, Down’s syndrome or other genetic abnormalities.
One such marker is alpha-fetoprotein (AFP), a substance produced by the fetal liver. If spina bifida or anencephaly (no brain tissue) is present, larger amounts of AFP leak into the amniotic fluid and the mother’s blood, so levels are higher than normal. Conversely, low levels of AFP are associated with a Down’s baby (Am J Obstet Gynecol, 1984; 148: 886-94).
The AFP blood test is usually performed at 16-20 weeks and is most useful for detecting neural tube defects. But, beyond that, like other blood tests, it is not very accurate and its results are affected by a vast number of variables. Less than 10 per cent of mothers diagnosed with raised AFP levels in their blood, and who subsequently undergo other tests such as amniocentesis, have affected babies.
In reality, there is a wide range of ‘normal’ levels of AFP. Levels can vary from day to day and, in the second half of pregnancy, double every four weeks. Indeed, inaccurate dates and multiple pregnancies account for a quarter of all raised AFP levels (Obstet Gynecol, 1989; 74: 17-20). If you had an early threatened miscarriage, if you smoke, have viral hepatitis or are black, you are also more likely to show ‘high’ AFP levels. Women carrying boys show slightly higher levels of AFP. If your pregnancy is less advanced than estimated, or if you are very overweight or insulin-dependent, this can push AFP levels down (Am Fam Physician, 2002; 65: 915-20, 922).
In theory, the more markers the test looks for, the greater its accuracy (BMJ, 1988; 297: 883-7). This has led to the ‘double screen’, which looks for high levels of AFP and low levels of unconjugated oestriol, and the ‘triple screen’, which looks for, in addition to these, raised levels of total human chorionic gonadotrophin (a fetal hormone). The latter is now the ‘gold standard’ of blood-screening tests, though new tests looking for even more markers are being developed all the time.
Because it is non-invasive with no risk of miscarriage, many women feel they have nothing to lose by having the AFP/ double/triple tests. However, the high false-positive rate of the test means that a large number of women spend a significant proportion of their pregnancies worrying about whether they are carrying ‘damaged goods’ (Prenat Diagn, 1991; 11: 381-5), only to find out that there was nothing to worry about after all. In most hospitals, a retest is done routinely for women with abnormal results – and often gives a completely different result.