Consider The Evidence: Med/Peds Journal Roundup

November 20, 2006

Thiazides increase Diabetes, ALLHAT says dont worry

Filed under: archives of internal medicine, diabetes, hypertension — medblog @ 11:08 pm

Archives IM 11/13- ALLHAT data first made news with the conclusion that older, thiazide type diuretics were just as effective at reducing hypertension related disease as newer types of anti-hypertensives. One result of the study that is less well known is that thiazides appeared to increase the risk of incident diabetes. In the last issue of Archives, the ALLHAT investigators addressed this issue more directly.[Article]

The investigators performed post hoc analyses on the subgroups assigned to Ca-channel blockers, ACE-Is, and thiazides. Patients with known diabetes prior to start of therapy were excluded from analysis, as were those without baseline fasting blood glucose(FG) levels.

The data shows that all subjects had an increase in FG, but the thiazide group had a significantly greater increase in fasting glucose levels than the other groups. In addition, subjects in the thiazide group had significantly greater odds of developing diabetes.

Is this a problem? The investigators state that the increase in DM did not result in a greater rate of adverse events compared to the other groups. Are they trying to sweep the issue under the rug? Or are they postulating that the diabetic state induced by thiazides is less malignant than naturally occuring DM? Adverse events data was only 2 years out – not long enough to show effects from new onset DM. The increased risk may have also been partially washed out by the increase in DM among all groups, and the lack of comparison with placebo. I’m not sure I would dismiss the risk as easily as they suggest – but I guess I dont have to worry about grant money…

With both beta-blockers and thiazides being linked to decreased glucose tolerance, the choice of anti-hypertensives is becoming much more complicated.

November 9, 2006

SIDS breakthrough? Mittens for stroke rehab

Filed under: JAMA, SIDS, stroke — medblog @ 11:26 am

JAMA (11/1)- An article in last weeks issue by Paterson, et al. is showing up in many news reports as a major advance in the understanding of sudden infant death syndrome (SIDS). The investigators explored the role of seritonergic(5-HT) neurons in the pathogenesis of the syndrome. Frankly, alot of it was basic science gobledygook to me but I got the idea that 1.) seretonin neurons in the medulla oblongata are involved with autonomic and respiratory function, 2.) postmortem exams of the brains of SIDS victims show significant differences in the number and location of 5-HT neurons, as well as lower relative density of seritonin transport protein. With 35% of the SIDS samples coming from premature babies, and the number of controls that were premature not being reported – I was left wondering if this was all confounded by normal developmental differences?

In a more immediately applicable study, Wolf et al. took about 200 stroke patients with hemiparesis and and had half of them put a restraining mitt on their good hand. Its called constraint-induced movement therapy (CIMT). The idea was to reinforce the connections to the affected limb by restricting use of the other limb. The subjects were 3 to 9 months out from the stroke. The CIMT group were supposed to wear the mitt during 90% of waking hours for 2 weeks. In addition they had daily task practice sessions with the paretic limb. The control group had usual care stroke rehab. In the end, the CIMT group had significantly better motor function that persisted at 1 year after therapy.

Blog at