Question

Infertility treatment and Asthma: The millennium cohort study was set up as a nationally representative study of 18818 infants born from 2000-2002 in the UK who are followed up over time for different health outcomes. At the time of enrolment mothers were asked about conception history which was categorised into the following….

1. Unplanned (unplanned, unhappy about pregnancy)
2. Mistimed (unplanned, happy about pregnancy)
3. Planned (planned and time to conception<12 months)
4. Untreated-subfertile (planned and time to conception >12 months)
5. Ovulation induction (planned – used ovulation inducing drugs)
6. Infertility treatment used (planned- used assisted reproduction technologies)

When the children reached the ages of 5 mothers were asked about asthma in the child with the following results…

Mother’s Group	Total children	Total children with Asthma *	Cumulative Incidence rate (risk)
Unplanned	2029	372
Mistimed	3650	570
Planned	6480	885
Untreated-subfertile	505	93
Ovulation induction	173	20
Infertility Treatment	104	18
TOTAL	12941	1958

*Asthma diagnosed at any age up until the age of 5.

Q9: Calculate the cumulative incidence rates (risk) of Asthma for the groups (and overall) and fill them in on the above table

Q10: Calculate the cumulative incidence ratio (relative risk) of asthma in children of those with an unplanned (group i) pregnancy compared to those with a planned pregnancy with time to conception <12 months (group iii).

Q11: The relative risk of asthma in children of mothers on infertility treatment compared to those with a planned pregnancy is 1.27 with a 95% confidence interval around this of 0.84 to 1.96. How would you interpret this result (consider the relative risk and the 95% CI)?

Q12: In the paper the authors adjust for the sex of the child in the analysis and this gives an estimate for the relative risk in Q11 of 2.10 with a 95% confidence interval of 1.16 to 3.81.

1. How would you now interpret the RR (and 95% CI)?

2. Can you give a possible reason why the RR might increase when adjusting for the child’s sex?

3. If the study size was larger how (if at all) would you expect the 95% CI change?

Q13: The analysis was repeated in 2013 looking at asthma diagnosed by age 10. Between age 5 and 10 a total of 2444 children had been lost to follow-up, but a date was known for the last follow up time for each child (e.g. for a child the survey was completed at age 8 but not after that).

1. What proportion of children had complete follow up between age 5 and 10?
2. How could you analyse the data taking into account this loss to follow-up?

Q14: What type of cohort study is this?

Answer 1

9) Cumulative incidence rate is:

Total children with asthama/ Total children

For the mother's group:

Unplanned CI = 0.1833

Mistimed CI = 0.1562

Planned CI = 0.1366

Untreated - subfertile CI = 0.1842

Ovulation induction = 0.1156

Infertility treatment CI = 0.1731

Total CI = 0.1513

10) The cumulative incidence ratio of asthma in children of those with an unplanned (group i) pregnancy compared to those with a planned pregnancy with time to conception <12 months (group iii) is

0.1833/0.1366 = 1.34

11) The relative risk of asthma in children of mothers on infertility treatment compared to those with a planned pregnancy is 1.27 which means that the risk of asthma in children with infertility treatment will be 1.27 times the risk in children born thorugh planned pregnancy.

The 95% confidence interval describes that in all the samples taken for the relative risk of asthamatic children for infertile treament pregnanvies and planned pregnancy the true value of the relative risk has a 95% chance to lie in the range (0.84, 1.96) .

12) After adjusting for gender, The relative risk increases from 1.27 to 2.1. Which means that now the risk of asthma in children with infertility treatment will be 2.1 times the risk in children born thorugh planend pregnancy.

The 95% confidence interval is (1.16, 3.81) which means that in a sample the true value of the relative risk lies in the range (1.16, 3.81) with a confidence level of 95% i.e. there is a 95% chance that the true relative risk vaue will lie in the this interval.

(We can only answer four parts at a time. For rest of the parts please post them as separate question.)