Question

Abstract 2. Mann JR and McDermott S. Are maternal genitourinary infection and pre-eclampsia associated with ADHD in school-aged children? J Atten Disord 2011;15(8):667-73.

OBJECTIVE: To investigate the hypothesis that maternal genitourinary infection (GU) infection is associated with increased risk of ADHD. METHOD: The authors obtained linked Medicaid billing data for pregnant women and their children in South Carolina, with births from 1996 through 2002 and follow-up data through 2008. Maternal GU infections and pre-eclampsia were identified on the basis of diagnoses made during pregnancy, and cases of ADHD were identified on the basis of diagnoses made in the child's Medicaid file. RESULTS: There were 84,721 children in the data set used for analyses. Maternal genitourinary infection was associated with significantly increased odds of ADHD (OR = 1.29, 95% CI = 1.23-1.35). Pre-eclampsia was also associated with increased risk (OR = 1.19, 95% CI = 1.07-1.32). Children whose mothers had both GU infection and pre-eclampsia were 53% more likely to have ADHD, compared to those with neither exposure. When we examined specific infection diagnoses, chlamydia/nongonococcal urethritis, trichomoniasis, urinary tract infection, and candidiasis were associated with increased risk of ADHD, whereas gonorrhea was not. DISCUSSION: Maternal GU infection appeared to be associated with increased risk of ADHD, and based on the findings it was concluded that further research is needed to describe the mechanism(s) underlying the association.

The authors report “pre-eclampsia was also associated with increased risk (OR=1.19, 95% CI=1.07-1.32” of ADHD. Based only on this statement, which of the following is most likely NOT responsible for this observed result?

a	Confounding by gestational age
b	Systematic error

c	Random error

Answer 1

In order to ascertain the impact of errors on the results/conclusions ,we will need to find out the standard deviation in both cases.

We will take help of data from confidence intervals(CI) to ascertain it.

The CI is given by

where SE is the standard error and Z is the test statistic

For the pre sclamsia cases given that

OR=1.19, 95% CI=1.07-1.32

or,

Keeping values

1.32-1.19=z.SE

or z.SE=0.13

At 95 % Confidence the z statistic is given by 1.96

Keeping it we get,

SE =0.13/1.96

=0.066

Similarly the SE (GU infection) =0.030

Now, observe the value of the standard errors form the two samples..

Clearly both are significant at 95 % confidence.

Thus we are left with only one option of confounding by gestational age which appears to be the least likely reson for the observation of the result since we are not provided with any data to infer anything about its relation with the ADHD probability.