Absolute and relative risk

From the BBC:

Fathers-to-be should stop smoking to protect their unborn child from the risk of stillbirth or birth defects, scientists say.

University of Nottingham researchers found that pregnant women exposed to smoke at work or home increased their risk of stillbirth by 23% and of having a baby with defects by 13%. They looked at 19 previous studies from around the world. A UK expert said it was "vital" women knew the risks of second-hand smoke.

Ho hum.

The figures are probably complete rubbish, but let's take them at face value for now. According to Tommy's there are around 4,000 still births a year in the UK against around 800,000 live births, and let's assume that a quarter of mothers smoke or are exposed to 'second-hand smoke'.

While I accept that a still birth is heart breaking and distressing for the parents involved, let's do our bit to inform women of the absolute risks of second-hand smoke:

Not exposed - 600,000 pregnancies x 0.473% = 2,837 still births
Exposed - 200,000 pregnancies x [0.47% x 1.23=0.58%] = 1,163 still births
Total = 4,000 still births.

So in absolute terms, 'second-hand smoke' increases the risk by 0.11% (from 0.47% to 0.58%). What's even more heart breaking is of course that 81% of still births where 'second-hand smoke' was in play would have happened anyway, so this will cause additional guilt and suffering on the part of the parents.

Finally, let me do a bit of Victimhood Poker for once and point out that a far greater risk factor is how many grandparents the parents of the child have in common.

Today's facts and logic fail

From The Times:

Between 20 and 50 per cent of all Muslim marriages are said to be consanguineous [the PC term for inbreeding or even incest] and 8.5 per cent of all births are to parents related by blood [so that's an epic fail within the same sentence, unless the vast majority of consanguineous Muslim marriages are childless, whatever].

Religions vary as to what they allow. The Koran prohibits uncle-niece marriages, even though these are permitted by Jews and Hindus. Yet uncle-niece marrages involve the same amount of inbreeding as marriages between cousins — with 12.5 per cent of genes being identical.

Nope. We can argue the toss about the 12.5% figure, but an uncle-niece baby is far more inbred than one born to first cousins. The following diagram shows why:
To explain, in the top half, A and B have child AB, C and D have two children CD and CD; E and F have child EF. AB marries one of the CDs and the other CD marries EF, and so on. The two first cousins ABCD and CDEF have a baby ABDCCDEF. So that baby's genes are 25% from A or B; 50% from C or D and 25% from E or F.

In the bottom half, Uncle CD and Niece ABCD have a baby. Their baby's genes are 25% from A or B and 75% from C or D. That is a considerably smaller gene pool than a baby born to first cousins.

Separated at birth?

Shamelessly pinched from Capitalists@Work.

Ann Cryer MP (Lab, Keighley)

Prompted by Verity over at DK's, I have uncovered some fascinating statistics.

Another Labour MP making evidence-based-policy! Whatever next?

She doesn't always get it right though - she called for immigrants to be forced to learn English a few years ago, which is bollocks of course. Once people are here they are here, you can't deport them for failing a language test. The cheaper and better solution would be to grant residence permits only to people who can already speak reasonably good English when they apply.

* Update - a subsequent post at Vindico's has set me thinking - would such a ban be in the slightest enforceable? Hmmm...