Peter Gallagher

As reported by Sandy Szwarc, the inde­fati­ga­ble Kather­ine Fle­gal has pro­duced a stun­ning rebuff to the excesses of the diet zealots who are spin­ning up an ‘obe­sity epi­demic’ (abet­ted by the World Health Orga­ni­za­tion).

Fle­gal and her col­league Barry Graubard have exam­ined data from the huge HANES III sur­vey of the U.S. pop­u­la­tion (1988 — 1994) which included a bunch of anthro­po­met­ric mea­sure­ments of a large sam­ple of the pop­u­la­tion, in addi­tion to body-weight. They then exam­ined death cer­tifi­cates from the U.S. death reg­istry that linked to the HANES III sur­vey par­tic­i­pants through the year 2000 to see who died when, and of what.

What they found is sum­ma­rized in the PubMed abstract

“For the level equiv­a­lent to BMI <18.5, esti­mates of excess deaths ranged from 0.3% for waist-hip ratio to 2.4% for per­cent­age body fat. All except waist cir­cum­fer­ence, waist-hip ratio, and waist-stature ratio were sig­nif­i­cantly greater than zero (P < 0.05). For the level equiv­a­lent to BMI 25 to <30, the per­cent­age of excess deaths was 0.1% for per­cent­age body fat and neg­a­tive for all other vari­ables; esti­mates were sig­nif­i­cantly below zero only for cir­cum­fer­ences and waist-stature ratio. For the level equiv­a­lent to BMI > or = 30, esti­mates ranged from –1.7% for waist cir­cum­fer­ence to 1.5% for per­cent­age of fat; none were sig­nif­i­cantly dif­fer­ent from zero. Esti­mates for all-cause mor­tal­ity, obesity-related causes of death, and other causes of death showed no sta­tis­ti­cally sig­nif­i­cant or sys­tem­atic dif­fer­ences between BMI and other variables.”

In other words, no mat­ter how you mea­sure body shape—by ‘BMI’, hip-to-waist, waist cir­cum­fer­ence etc—the risks fat peo­ple face of dying from any cause is no greater than the risk faced by peo­ple of ‘nor­mal’ weight. Like pre­vi­ous results, this study shows that skinny peo­ple may even have a slight dis­ad­van­tage.

Sandy Szwarc reports that the authors went on to examine

“[D]eaths from 21 dif­fer­ent “obesity-related” dis­eases, includ­ing: car­dio­vas­cu­lar dis­ease, can­cers (colon can­cer, breast can­cer, esophageal can­cer, uter­ine can­cer, ovar­ian can­cer, kid­ney can­cer, or pan­cre­atic can­cer) and dia­betes or kid­ney dis­ease. For the low risk level of mea­sure­ments, the asso­ci­a­tions with deaths “were all quite small and close to zero,” they wrote. There were no sta­tis­ti­cal asso­ci­a­tions for the inter­me­di­ate lev­els. Finally, “for the obese level, none of the 21 dif­fer­ences would have been sig­nif­i­cant even at a P value of 0.05,” they reported.”

The HIV study: Esther Duflo looks at the evi­dence from two stud­ies of dif­fer­ent approaches to HIV edu­ca­tion in rural Kenya. The first looked at preg­nancy rates among unmar­ried teenage girls as a proxy for unpro­tected sex­ual activ­ity. It found that in schools where teach­ers were trained in the offi­cially sanc­tioned ABCD cur­ricu­lum (“Abstain, Be faith­ful, use a Con­dom… or you Die”), the rates of preg­nancy were about the same as they were in schools where the teach­ers had not been trained to deliver this pre­ven­tion message

Duflo com­pares this result with the report of a pro­gram to deliver young women a much sim­pler mes­sage on the dis­tri­b­u­tion of AIDs in Kenya by gen­der and age. The data clearly show that hav­ing unpro­tected sex with older men (‘sugar-daddies’) would be dan­ger­ous because the preva­lence of AIDs in this group is much higher than among young men of sim­i­lar age to the girls. This risk-reduction strat­egy is more effective.

“Teenage girls who ben­e­fited from this pro­gram reported hav­ing sex less fre­quently with older men and more often with boys their own age (with whom they use con­doms more eas­ily). After a year, the preg­nancy rates were 5.5% in schools that had not received the pro­gram and 3.7% in schools that had received it. This reduc­tion is mainly due to a 67% reduc­tion in preg­nan­cies with an older partner.”

Duflo sur­mises that the sterner risk-prevention strat­egy which cov­ers basic med­ical facts about AIDS, HIV trans­mis­sion, pre­ven­tion (exclud­ing con­doms), absti­nence before mar­riage and care for peo­ple liv­ing with AIDS attempts to con­vey a lot of infor­ma­tion with­out actu­ally impart­ing any “prac­ti­cal wisdom”.