How does your body-fat distribution affect your risk of death?

General and Abdominal Adiposity and Risk of Death in Europe
Original source: New England Journal of Medicine Volume 359:2105-2120 November 13, 2008 Number 20
Authors: T. Pischon, H. Boeing, K. Hoffmann, M. Bergmann, M.B. Schulze, K. Overvad, Y.T. van der Schouw, E. Spencer, K.G.M. Moons, A. Tjønneland, J. Halkjaer, M.K. Jensen, J. Stegger, F. Clavel-Chapelon, M.-C. Boutron-Ruault, V. Chajes, J. Linseisen, R. Kaaks, A. Trichopoulou, D. Trichopoulos, C. Bamia, Ph.D., S. Sieri, D. Palli, R. Tumino, P. Vineis, S. Panico, P.H.M. Peeters, A.M. May, H.B. Bueno-de-Mesquita, F.J.B. van Duijnhoven, G. Hallmans, L. Weinehall, J. Manjer, B. Hedblad, E. Lund, A. Agudo, L. Arriola, A. Barricarte, C. Navarro, C. Martinez, J.R. Quirós, T. Key, D.Phil., S. Bingham, K.T. Khaw, P. Boffetta, M. Jenab, P. Ferrari, and E. Riboli
Publication date: November 13, 2008
DOI (digital document ID): (unavailable)
Requires subscription? No
Publicly-available abstract: Yes

Synopsis:
This article addresses the question of whether the risks of obesity are related to the distribution of body fat, rather than the existence of fat excess body weight alone.

The investigators examined Body Mass Index (BMI), waist circumference, and waist-to-hip ratio with the risk of death among 359,387 participants from nine countries over a period of almost ten years. The results were analyzed based on age at recruitment into the study, location at which they were entered into the study, height, educational level, whether or not the subject smoked, how much alcohol the person drank, and how physically active the person was.

The 1% of respondents with the highest and lowest energy intakes relative to estimated energy requirements were excluded from the study, as were respondents with a history of cancer, heart disease, or stroke.

"Normal" weight was considered to range from BMI 18.5 to BMI 25.0

The lowest risks of death based on BMI were at 25.3 for men and 24.3 for women. (This corresponds to the borderline between "normal weight" and "overweight".) After adjusting for this, absolute waist size and large waist-to-hip ratio were associated with greater risk of death. These last two correlations were stronger for people with lower BMI than for those with higher BMI, and stronger for men than for women.

One possible reason for the gender-related difference may be that overweight men tend to carry more fat in their midsections (visceral-fat) than equally-overweight women. This is important because visceral-fat tissue secretes a number of chemicals (lipokines) that can cause or exacerbate a number of chronic diseases.

Interestingly, it was found that the reasons for increased risk of death changed depending on waist measurement: those people who died and who were simply overweight or obese had more deaths from circulatory issues than from other causes, whereas those who had more abdominal obesity tended to die from respiratory issues rather than from other causes. (This may be a confounding issue, related to the observation that smokers tend to have more abdominal fat than non-smokers of similar BMI, and that smokers have higher risk of dying from respiratory system issues. The authors note that there is a known association between abdominal adiposity and chronic bronchitis, but do not state if this is a smoking-related association or not.)

The researchers conclude that these data suggest that both high body weight and abdominal fat are associated with higher risk of death, and support the use of waist circumference and or waist-to-hip ratio in addition to BMI in assessing the risk of death.

Reliability:
This study followed a large number of people for a relatively long period of time, making the study population relatively reliable; however, high BMI is not always related to adiposity. Body-builders, as a noted example, can have BMI in the obese range, but body fat measurement by either bioelectrical measurement or water immersion will indicate extremely low body fat percentages. On the other hand, these people generally have waist measurements and waist-to-hip ratios that are significantly smaller than people of the same body mass index, but with greater percentages of body fat. They also tend to have higher metabolic rates, making them more likely to place in the highest caloric intake per kilogram of body mass; some people who fit in this category were probably excluded from the study for that reason.

Another consideration, noted briefly by the investigators, is that there are a number of chronic diseases that were not excluded from the study. Some of these (such as Type 2 diabetes) affect, or are affected by, body weight and weight distribution. These could also affect the relative risks of death of the population observed.

Practical Applications:
The use of waist measurements and waist-to-hip ratios, in conjunction with BMI, can assist doctors and public health professionals in assessing individual and community health risks.

The authors' Figure 2, representing relative risk based on waist circumference and waist-to-hip ratio, shows a standardized risk of 1.0 for men at approximately 93 cm (about 36"), 0.94 waist-to-hip ratio (36" waist corresponding to about 39" hip), and for women at approximately 77 cm (about 30") waist size, 0.79 waist-to-hip ratio (30" waist corresponding to about 38" the hip). These measurements are lower than the standards currently used for determination of risk (102 cm/40" for men, 88cm/35" for women). Interestingly, these measurements correspond to the largest "standard" reference clothing sizes (men's pants to jacket size 42, women to U.S. Misses size 16/European size 42) as expressed in the home-sewing pattern industry, or "true-to-size" in the ready-to-wear industry.

This suggests that an average person should be able to determine his or her risk based on off-the-rack clothing size.

For further reading:
Home-sewing pattern measurements:

• Simplicity Patterns (United States):
  
  • US measurements
  • European measurements


• Burda Patterns (Germany):
  
  • US and European measurements

This synopsis was based on the complete article.

Diabetes and Risk of death from coronary events

Diabetes Mellitus and Mortality after Acute Coronary Syndrome as a First or Recurrent Cardiovascular Event
Original source: PLoS ONE Vol. 3, No. 10
Authors: Richard M. Cubbon, Afroze Abbas, Stephen B. Wheatcroft, Niamh Kilcullen, Raj Das, Christine Morrell, Julian H. Barth, Mark T. Kearney (corresponding author), Alistair S. Hall for the EMMACE-2 investigators
Publication date: 22 October 2008
DOI (digital document ID): 10.1371/journal.pone.0003483
Requires subscription? No
Publicly-available abstract: Yes

Synopsis:
Diabetes has been associated with problems recovering from, and with higher risk of death from, acute coronary syndrome. The investigators were looking to find whether this increased risk was true for all persons with diabetes, or only for those persons with diabetes who already had cardiovascular disease, compared to people without diabetes (with, or without, diagnosed cardiovascular disease). The investigators wondered whether cardiovascular disease is diagnosed earlier in people with diabetes than in the general population, and if that early diagnosis, plus aggressive preventative treatment, reduced the risk of dying from acute coronary syndrome.

For the purposes of the study, "Acute Coronary Syndrome" covered unstable angina or myocardial infarction (heart attack). "Cardiovascular Disease" was devined as myocardial infarction, angina, stroke, peripheral vascular disease, or coronary revascularization (either the heart's circulatory system rerouting itself, or the need for stents, bypasses, or other surgical techniques to open or replace blocked blood vessels). Note that by these definitions, "acute coronary syndrome" is a subset of "cardiovascular disease".

The researchers looked at the data collected in the EMMACE-II observational study in 2003, which tracked just under 2500 patients. The investigators used the patients' previous medical records to split the patients into four groups: those without diabetes or previously-diagnosed cardiovascular disease; those without diabetes, but who had existing cardiovascular disease; those with diabetes but without a history of cardiovascular disease; and those with both diabetes and a history of cardiovascular disease. The data from these four groups were analyzed separately and compared.

The results showed that while patients with diabetes had an overall increased risk of death from acute coronary syndrome, the rates of death of patients with no history of cardiovascular disease were approximately the same, regardless of whether or not they had diabetes.

After adjusting for age and gender, there was still an increased risk of death for patients with diabetes and a history of cardiovascular disease over patients from the other three groups; however, when the results were adjusted for levels of Troponin I -- which was used to indicate how much of the heart muscle had died due to cardiovascular disease, the correlation between mortality and diabetes disappeared. This lack of correlation remained even when the investigators broadened or narrowed the definition of cardiovascular disease used in their analysis.

These findings suggest that the risk of dying from acute coronary syndrome may be related to the degree to which the heart's blood vessels have been blocked, and that the reduced degree of blockage in patients with diabetes, but no history of cardiovascular disease, may be related to aggressive preventative treatment that is not routinely offered to people without diabetes.

Reliability:
The authors are from accredited medical institutions and the research is published in a peer-reviewed journal. The size of the study is large enough to show potential trends and suggest directions for further research. The authors note limitations of the study include the absence of several key risk indicators (such as albuminuria -- protein in the urine -- and obesity) and details of any preventative therapy the patients may have received. They also note that patients with a history of previous cardiovascular disease are less likely to receive prescriptions for preventative drugs (such as statins and ACE inhibitors) than patients being discharged after their first acute coronary event, and because this is counterintuitive to standard practice, understanding the reasons for this difference would also help to quantify the differences in risk.

Practical Applications:
Understanding the true origins of the risks of death from acute coronary syndrome can help our doctors better treat us and help us reduce those risks to the greatest degree possible.

For further reading:
The British Cardiac Society Working Group definition of myocardial infarction: implications for practice -- abstract describing the EMMACE-II study.

This synopsis was based on the entire article.