Chatu Yapa, Masters in Applied Epidemiology Trainee, Health Protection NSW
In the 1960s, researchers were keenly investigating the potential causes of Down’s syndrome. The aetiology of Down’s syndrome was largely unknown and there were different hypotheses as to the cause of this condition. Some of the hypotheses related to the geographical distribution of cases – higher rates being seen in urban areas compared to rural areas, other theories looked at the influence of intrauterine infections such as cytomegalovirus, rubella and toxoplasma, and yet other considerations included factors such as birth order and maternal/paternal age.
A study done by Stark and Mantel in 1966 aimed to examine in more depth, the apparent association between maternal age and birth order with the risk of Down’s syndrome.
The graphs below show some of their results (reproduced from the original study).
Graph 1 shows the trend between the prevalence of Down’s syndrome and birth order.
There appears to be an association between increasing birth order and the risk of Down’s Syndrome – a 5th- born child appears to have almost a 4-fold increase in the risk of being born with Down’s syndrome, compared to a 1st-born child.
However, consider also that the order in which a woman's children are born is also linked to her age at the time of child birth. When Stark and Mantel examined the relationship between maternal age at birth and risk of the child having Down’s syndrome, they observed the relationship shown in the graph below.
This graph above shows a striking relationship between maternal age at birth and the child’s risk of being born with Down’s syndrome.
It would be reasonable to assume that a woman giving birth to her fifth child would be, on average, older than a woman giving birth to her first child. It appears that the two factors – birth order of children and maternal age when a child is born – are inter-related.
- How might Stark and Mantel further evaluate the true association between the two variables and the risk of Down’s syndrome?
- What is this mixing of effects called in epidemiology?