Fri, Jul-15-16, 10:07
|
|
|
|
Plan: P:E/DDF
Stats: 225/150/169
BF:45%/28%/25%
Progress: 134%
Location: NC
|
|
Study: Higher BMI Signals Earlier Death
We have had many other studies and threads in this forum both supporting AND denying this new study on the validity of the BMI, but none seemed right to add, so starting a new thread....
Quote:
Also finds no evidence for 'obesity paradox'
by Parker Brown, MedPage
A massive meta-analysis pooling data from millions of people in several countries reaffirmed that body mass index (BMI) has a J-shaped relationship with mortality, with the lowest death rates among those in the traditional "normal" range of 20-25.
The study of nearly four million people revealed that those in every BMI category above and below the normal range had significantly higher mortality rates.
The elevation in risk applied to even mildly overweight people, and was highest for those with overt obesity, according to researchers with the Global BMI Mortality Collaboration, a part of the University of Cambridge.
The study was published in The Lancet on Wednesday.
The hazard ratios for all-cause mortality went up as BMI went up from the normal range in the analysis (relative to BMI 20 to <25):
1.07, 95% CI 1.07-1.08, for BMI of 25.0 to <27.5
1.20, 95% CI 1.18-1.22, for BMI 27.5 to <30
1.45, 95% CI 1.41-1.48, for BMI of 30.0 to <35.0
1.94, 95% CI 1.87-2.01, for BMI of 35.0 to <40
2.76, 95% CI 2.60-2.92, for BMI of 40 to <60
"Our results challenge recent suggestions that overweight and moderate obesity are not associated with higher mortality, bypassing speculation about hypothetical protective metabolic effects of increased body fat in apparently healthy individuals," wrote the authors. Their data showed the J-shaped relationship maintained for every age group, albeit attenuated somewhat among those in the 70-89 age range.
The risk was also increased in the lowest BMI strata: the hazard ratio was 1.13 (95% CI 1.09-1.17) for BMI of 18.5 to <20 and 1.51 (95% CI 1.43-1.59) for BMI of 15.0 to <18.5.
For a BMI of more than 25, all-cause mortality increased log-linearly, and the hazard ratio per every 5 BMI points was 1.39 (95% CI 1.34–1.43) in Europe, 1.29 (1.26-1.32) in North America, 1.39 (1.34-1.44) in East Asia, and 1.31 (1.27-1.35) in Australia and New Zealand.
The authors added that, generally, that relationship was more pronounced in younger people than in older people -- the hazard ratio per 5 BMI points above 25.0 was 1.52 (95% CI 1.47–1.56) for BMI measured when the participant was age 35–49 versus 1.21 (1.17–1.25) for 70–89 years (P<0.0001).
The relationship was also stronger in men than in women (1.51, 95% CI 1.46–1.56, versus 1.30, 1.26–1.33; P<0·0001), though that wasn't true for studies that relied on self-reported BMI.
The study raises two major issues, according to David Berrigan, PhD, MPH, at the Division of Cancer Control and Population Sciences at the NIH and two colleagues in an accompanying editorial. The first is that it's still unclear whether the relationship between BMI and mortality can be generalized , they wrote. The second issue is what kinds of public health measures would come from a study like this, which pooled data from several countries.
"Substantial research and conceptual questions remain for each of these issues," they wrote. There are questions of who to exclude and whether that would bias the results. Excluding too many people, like ever-smokers, would limit the generalizability of the findings.
There has been a lot of debate about studies that examine the relationship between body mass index and mortality, according to Berrigan and co-authors, since BMI is a limited measure. A number of studies have identified an "obesity paradox" where some outcomes are better for heavier than skinnier people, particularly in the elderly.
But the authors said that their findings challenge assertions that overweight and grade 1 obesity (BMI of 30 to <35) are not associated with a higher risk of mortality.
This analysis pooled 239 studies across 32 countries and included about 1.6 million deaths. Only data from large, prospective studies was included, and only studies performed after 1970 were included. The hazard ratios of all-cause mortality were generally similar across different continents, though the risk was higher for underweight and for the highest class of obesity in Europe than in East Asia.
The authors also examined death by coronary heart disease, stroke, other cardiovascular disease, cancer, and respiratory disease. Above a BMI of 25, BMI was correlated with coronary heart disease, stroke, and respiratory disease; it was also moderately correlated with cancer.
Only never-smokers were included in the study to avoid residual confounding, since smoking has a relationship to both mortality and to BMI. In addition, those known to have pre-existing chronic diseases were excluded, and the initial 5 years of follow-up were excluded because diseases at baseline that cause death within 5 years might lead to reverse causation.
Limitations of the study include that it pooled observational studies only, so it is not designed to determine causality. In addition, the only measure of adiposity was BMI, which can be more inaccurate than other measures of fat.
"Despite the limitations of observational studies for causal inference of obesity and mortality, many crucial questions about BMI will continue to rely on observational data," wrote Berrigan and colleagues. "To date, few sufficiently sized randomized trials have been done to address whether weight-loss interventions reduce mortality or morbidity."
|
http://www.medpagetoday.com/Endocri...&eun=g977378d0r (subscription)
A thread on opposite viewpoint, Nov 2015: http://forum.lowcarber.org/showthread.php?t=470745
Last edited by JEY100 : Fri, Jul-15-16 at 14:41.
|
|