Junkfood Science: Japanese researchers find no support for healthy weight guidelines

August 06, 2008

Japanese researchers find no support for healthy weight guidelines

The BMIs that have been deemed by the World Health Organization as ‘healthy’ among Asians actually put people at higher risks for premature death, found Japanese researchers. WHO recommends that Asians maintain lower BMIs than Western populations, less than 23 kg/m2. But national health statistics, from the NIPPON DATA80 Study of Japanese men and women followed for 19 years, found "no support that this is a healthy size for Asians."

The results of this Japanese Ministry of Health and Welfare-sponsored study were published in the July issue of the journal Obesity, but most consumers never heard about it. In the ongoing interest of bringing you important research the media missed, and round out the nonstop marketing of a crisis of obesity, here is a quick look.

This study, led by Dr. Atsushi Hozawa with the Department of Health Science at Shiga University of Medical Science in Otsu, Japan, examined whether it is appropriate to apply Western weight, waist and body fat guidelines defining obesity to Asian populations. With the recent focus on metabolic syndrome and the Japanese government mandating flab checks among workers, ensuring medical guidelines are evidence-based is taking on increasing importance.

As the authors noted in their introduction, several studies conducted in Japan and other Asian countries have suggested that underweight have the highest all-cause mortality rates:

Several explanations have been proposed to explain how being underweight is associated with [greater] all-cause mortality, such as the effect of heavy smoking on weight loss; the effect of age (older participants are relatively leaner than those who are middle aged); total cholesterol is lower in leaner individuals and low total cholesterol effects hemorrhagic stroke, cancer and liver diseases; and weight loss is often associated with subclinical diseases.

Their concerns were also well-founded, given that health statistics in the United States and other Western countries have not seen lower risks of premature death among thinner BMIs in their populations; but, in fact, higher mortality risks are most associated with those whose BMIs fall within classifications of ‘normal’ and ‘underweight.’

U.S. citizens — the U.S. Centers for Disease Control and Prevention National Center for Health Statistics reports that age is the most significant risk factor for death; NHANES mortality data since 1971 shows thinness has the highest risks (controlling for confounding factors to eliminate reverse causation); and risks associated with obesity have been greatly exaggerated, with the most recent NHANES data finding that obesity is associated with lower all non-cancer and non-cardiovascular deaths and infection-related deaths, only 5% of heart-related deaths associated with obesity, and that obesity was unrelated to overall cancer risks

men — the Aerobics Center Longitudinal study of more than 25,000 men followed for more than ten years found the highest BMIs associated with lowest mortality rates, especially as they aged; Dr. Walter Willett at Harvard's Brigham and Women's Hospital found among men 65 years and older, when most deaths occur, “no significant relations between BMI and overall, cardiovascular disease, or cancer mortality risk”

women — the Study of Osteoporotic Fractures, clinically followed more than 8,000 mature women for eight years found that the highest risks of mortality were among the lower BMIs, starting BMI under 22.8, concurring with the National Health and Nutrition Examination Survey; the Cardiovascular Health Study of 5,200 older men and women found BMI was inversely related to mortality

elderly — the EPESE (Established Populations for Epidemiologic Studies of the Elderly) found no tenable link between BMI and disability and the lowest mortalities more associated with those who were obese or overweight compared with those in recommended BMIs [in green]; the National Long Term Care Survey, a representative survey of U.S. seniors followed for more than nine years, found the lowest mortality rates among the obese and that being low weight (after adjusting for every possible confounder) was still associated with the highest risks of dying prematurely

So, these Japanese researchers examined the association between BMI and deaths from all causes (the most important endpoint), as well as cardiovascular disease-related deaths, among all adult Japanese, 30 years of age and older, without pre-existing heart disease. They used the most reliable population available. The 8,924 study participants (3,969 men and 4,955 women) were from the Japanese National Cardiovascular Survey of 1980 conducted with the annual National Nutrition Survey. In order to reduce the effects of confounding factors and pre-existing disease on mortality, they also analyzed the data on just healthy, nonsmoking Japanese under age 70, and after removing deaths that occurred during the first five years.

During the 19 years of follow-up, there were 1,718 deaths, 607 due to heart disease. The researchers found that while many risk factors (such as blood pressure, total cholesterol, diabetes) were progressively lower among those in lower BMIs, under 25 kg/m2, they were inversely related to actual mortality rates. This was seen in the entire population sample, as well as just among the healthy participants. [Image is the risks for all-cause deaths among the entire population and the healthy.]

Before people at either extreme worry, though, note that none of the relative risks, despite the chartsmanship, were strongly tenable for an epidemiological study. The correlations most strongly show that a wide range of body sizes share very similar risks for all-cause mortality. Population-wide, risks were nearly identical between BMIs of 18.5 to 27+ (RR=1.17 - 1.19); and increasing to RR=1.39 for BMIs under 18.5. But, as these researchers were most concerned with, there is no evidence that recommended lower weight cut-offs for Asian populations reduce risks.

The authors also concluded that the higher mortality among lean people could not be fully explained by those popularly suggested reasons (smoking, weight loss, pre-existing illness, low cholesterol, or age). The risks associated with lower weights were slightly greater among the healthiest segments of the population. “The relationship between BMI and cardiovascular disease mortality was also similar to associations with all-cause mortality,” they reported. Their findings, they added, were consistent with a recent study of Korean men and women without cardiovascular disease, cancer, liver disease, diabetes, or respiratory diseases. As they concluded:

Our study does not support the notion that a lower BMI cutoff value in Asian vs. Western populations is warranted because all-cause mortality risk does not appear to differ between BMI ranges from 18.5 to 22.9 and from 23.0 to 24.9 kg/m2. Our findings further suggest that risk of death may be highest in individuals with a BMI <18.5 kg/m2, even for those who fell in our restricted healthy sample.

Curious, that while Japanese workers have been frantically trying to get down to goverment-recommended slenderness, these 19-year national findings revealing no evidence for the recommendations, were not widely reported.

© 2008 Sandy Szwarc

[Click on images to enlarge.]

Bookmark and Share