BODY WEIGHT and DEATHS and DISEASE
Abstract: Serious body overweight as above 210 lbs for men and 185 lbs for women of average height will increase risk of death substantially from cardiovascular disease, from cancer and from other causes. Obesity can result in losses of 5 to 15 years of Well-Days of life. Yet there is a quite substantial range of weight levels over which the variation of health risk is quite small. But for most those who are not much overweight, weight may be a health factor of lesser importance than is the exercise that produces cardiofitness, smoking, or of a maintained healthful diet.
The Important Studies: Vast publicity has been given the problem of our population trend toward overweight recently. Thus the health debits of overweight are valued in Life Ahead by a comprehensive statistical analysis of the available research by cause. Weight control should be an important part of any health program. There probably are more than a hundred different published research studies that cite that effect on disease and death of body weight measured by the Body Mass Index called the BMI. Yet the most comprehensive and accurate information appears to be provided by two major studies developed by the American Cancer Society. The early one on 750,000 men and women was by Lew and Garfinkel, J Chron Dis 1979, 32:563); the later ones on more than 900,000 US adults were by Calle, EE. et al (N Engl J Med 1999, 341:1097 and 2003, 348:1625). These key studies provided the effect of a wide range of body weights on risks of all causes of death and risks by type of disease for both men and women of various ages. Despite the fact that these two studies considered populations from earlier and later decades during which overall population health risks had changed substantially, their measured risks of death from disease for body weight were quite similar. There has been some statistical confusion among some authors' conclusions from other studies about the effect of very low body weights and some other weight-related factors on risk. Yet the results of most of these other studies really appear statistically consistent within their scope with the results of these two largest ones. A recent study Flegal, JAMA 2005, 293:No 15 claimed that the effect of obesity may have declined over time. But the risk ratios cited in this study had error margins far too high to justify such a conclusion. Two other important studies included in the Life Ahead analysis of weight and risk of cancer were those on the Nurses, (Manson, JE N Engl J Med 1995 333:677) , and that of Rodriquez, C, Cancer Epidemiol Biomarkers Prev 2001, 10:345). The Life Ahead analysis that follows is based on the combined results of these major studies. Most other studies do not have the resolving power to contribute to this type of an analysis.
The Lew study used a weight index for height that was based on a ratio of weight adjusted for height to the average weight vs. height of this population. Although this probably is the most accurate index relating weight to height, most researcher have based results on the more approximate BMI. Thus the results of Lew were converted to values of the BMI for consistency. The BMI is weight in kilograms divided by the square of height in meters. Or BMI is weight in lbs times 703 divided by the square of height in inches.
The Average Population Weights for Height: Life Ahead uses deaths of an average US population as a starting basis for risks of both disease and death. Thus all risks are computed as departures in risk from this average population that includes a usual variation of weight with age and sex. For reference, the average values of population weight used in key research studies are:
Age 20 30 40 50 60 70 80
For Men 155 167 171 171 169 165 157
BMI for 68" height 23.6 25.4 26.0 26.0 25.7 25.1 23.9
For Women 125 135 141 143 143 142 138
BMI for 64" height 21.4 23.2 24.2 24.5 24.5 24.4 23.7
These average population values have changed over the years and are higher in the above Life Ahead profile - especially for women - than body weights that existed during the earlier Lew and Garfinkel Study. And they are lower than average body weights today because average US body weights have moved up about 10 pounds during the past two decades and 20 pounds since the 1960's. But these changes in population average weight should not affect the validity of a formula for changes in risk of disease and death with changes in weight. This simply means that weight index values and actual risks average higher today than they were then because people are heavier now and thus will usually have higher risks for this factor. The effect on health risk of body weight has not changed appreciably over the years studied.
Risk of Death vs. BMI and Body Weight. The risks of death vs. body weight in BMI for various disease groups used in the Life Ahead from analysis of the above major research studies follow for non-smoking men (M) and women (W). The bases of this analysis and equations developed is detailed later for those interested. Allow about 6-7 more or less pounds of weight for each inch of higher or lower height to the weight values listed. (Pounds are kilograms times 2.204).
BMI 18 20.5 24 28 32 36 40 44 48
Weight for 68" Man,lb 118 135 158 184 210 237 263 289 316
Weight for 64" Woman,lb 105 119 140 163 186 210 233 256 280
Risk Ratio Values for:
All causes of death M 1.25 1.10 1.03 1.12 1.35 1.72 2.20 2.71
W 1.25 1.07 1.01 1.09 1.31 1.61 1.87 1.93
Cardiovascular Death
avg M&W 1.25 1.10 1.02 1.08 1.22 1.43 1.62 1.8
All Cancer M 0.95 0.96 0.97 1.01 1.08 1.19 1.34 1.56 1.87
W 1.00 1.01 1.03 1.07 1.16 1.29 1.49 1.78 2.20
Colorectal Cancer M 0.87 0.96 1.09 1.26 1.48 1.69 1.95 2.26 2.62
W 0.92 0.96 1.03 1.11 1.20 1.30 1.41 1.52 1.64
Breast Cancer W 0.94 1.02 1.14 1.30 1.47 1.67 1.90 2.16 2.46
Prostate Cancer M 0.97 0.97 0.99 1.06 1.19 1.41 1.75 2.29 3.13
Endometrial Cancer W 0.72 0.89 1.20 1.69 2.39 3.37 4.75 6.71 9.45
Ovary Cancer W 1.05 1.05 1.06 1.12 1.23 1.40 1.66 2.04 2.66
Cervical Cancer W 1.00 1.03 1.12 1.35 1.73 2.32 3.25 4.72 7.09
Avg Genital Cancer W 0.90 0.95 1.14 1.46 1.90 2.53 3.43 4.71 6.56
Effect on Yrs of
Well-Days, for Men +0.3 +0.3 -0.8 -3.1 -5.8 -8.3 -10.7 -13.0 -14.7
Avg US Diet Women -0.95 +0.1 +0.31 -0.6 -2.4 -4.9 -7.0 -9.0 -10.9
Life Ahead
Overall risks of both cardiovascular diseases and death reach a minimum level at a BMI of about 23 for men and 24 for women. Values of weight in BMI at both lower and especially at higher levels than this result in increased risks. Risks of cancer, however, appear continuously lower at lower levels of body weight. Results on overall risk of death suggest that body weight contributes to risk of death from causes other cardiovascular and cancer. Thus a risk for weight on these other causes of death also is recognized in Life Ahead.
The results show that substantial overweight can produce a major increase in risk of earlier death. Very high body weights in the 220 to 300 lb range can result in 5 to 15 years loss in Well-Days of life. But the important risks are due to major overweight, as above 200 lbs for an average man, or above 180 lbs for an average woman. Women have higher risks than men for some causes, but lower risks than men for other causes. But women tend to suffer higher risks for overweight than do men. The risk of cancer in the genital area can become very high for seriously overweight women. An overweight woman with a family risk of cancer develops an extremely high risk of genital cancer and a quite high risk of breast cancer. Overall genital cancer risks shown in Life Ahead assume 58% risk from endometrial (uterus) cancer; 32% from ovary, and 10% from cervical cancer.
Risk values in this table are for individuals of an average age of about 45-50. Risk ratios for cardiovascular disease become significantly higher for those younger and a bit lower for those older. This is noted further in the description of the data analysis that follows.
Weight over the quite wide BMI range from 20 to 28 has a surprisingly small effect on health risk for men. This range includes weights from 135 to 184 pounds for men of average height. The higher cardiovascular and overall risk for those underweight occurs only for those who are quite seriously underweight. This risk may be in part due to illness or genetic disease problems that result in underweight rather than the effect of weight per se.
Numerous other studies have compared risks of disease with the BMI. Such results are usually consistent with and do not add much to the results of the above very large studies. As example, a very large study (J Nat Cancer Inst 2006:98:920) on 368,000 men and women in various European Countries found a risk for colorectal cancer similar to the above for men. Various studies including this have studied the effect of overweight by measures of waist and hip size, and some researchers claim a better relationship with body size measurements than with the BMI. This may be true, but advantage for using body measurements is marginal.
Reducing Weight by Diet and Exercise: Nearly every program for weight control admonishes use of exercise along with dieting. This leads to a paradox. It takes a lot of exercise to burn many calories, and reducing calories via diet is the key need to reduce the weight. But from a standpoint of health benefits, the effects of exercise and diet reverse. A maintained weight reduction of say 20 pounds might reduce risk of death by perhaps 20% for those having a BMI above 28. An accompanying cardiofitness program that is maintained can reduce risk of death by 50% or more. This emphasizes the importance of including cardio type exercise as a part of any program to reduce weight in order to realize an important health benefit.
Valuing the Benefits of Weight Change with Life Ahead: Life Ahead can display the potential benefits or debits of a change in weight. Download the free computer program from Download Life Ahead. The Demo program is not useful for the valuation because it assumes base values for an average population. You should enter your actual age, weight, and all other entries on the ten or eleven wellness entry screens to value weight and weight change for a specific person. You do not have to complete all entries asked as the program will assume average population values for items not entered. (But you do have to enter the exercise entries.) Then go back to the first entry screen and access the 4th option, 'Compute Results from Data now Entered'. This will compute results and display the result screen that is self-explanatory. Then access the 'Change Weight' option to revise weight at present age, and compute results again. This will display the reduction in risks of major disease and change in Well-Days of future life from maintaining the entered weight henceforth.
A somewhat more accurate result can be obtained after you enter a usual diet, compute results for this diet, and access the change weight entry.
Does Dieting Really Help Reduce Risk?: Much research on body weight from smaller studies is inconsistent. Some found little or no benefit for those that reduced weight. Some studies including an important one by Hammond on a million people suggested that people that reduced and then regained weight more than once had increased risk. This type of repeated but unsuccessful or 'Yo-Yo' dieting is recognized as another risk factor in Life Ahead. There were inferences that health depended more on long term weight than on recent weight. Reasons for this are not quantitatively explained.
But there are two problems with valuing the benefits of changes in weight. First, most people that diet do not keep their weight off over the long term. Health risks usually are measured over periods of at least 5 to 15 years. And second, some of those that lost weight lost it because of health problems. One study of "Intentional" vs. "Unintentional" weight loss found that only the unintentional loss presumably from illness or health caused increased risk. Thus although research does show that weight variability increases coronary disease risk, this can be misleading. Those that gain weight should have increased risk, and many of those that lost weight lost it because of health problems that were the real cause of the increased risk. The overall relationship between body weight via the BMI is very well confirmed. It seems highly likely that a person that reduces weight substantially by diet and keeps that weight off over the long term will achieve the benefit expected from this change. This benefit should be assured if added exercise is included and maintained with the weight loss program.
Weight, Health Risk and Mechanisms: Much research exists on the relationship between weight and risk of death, but little was found identifying WHY weight produces these effects. There is a clear relationship between overweight and blood pressure, but this explains only a little of the overall risks of body weight on cardiovascular death and none of its risk for cancer. The effect of weight on blood pressure is removed in Life Ahead's valuation the weight factor to avoid a duplication of these effects. There is a tendency for those overweight to exercise less. Yet an offsetting factor here is that heavy people can get more heart and other muscle exercise from a given physical activity than do those of lower weight. Total cholesterol usually falls somewhat when people lose weight by diet. But this is expected from difference in food intake. Heavier people usually do have somewhat higher serum cholesterol but this relationship is in part due to those dieting and does not explain the effect of weight on different diseases.
The effect of higher body weight in increasing risk of cancer may simply be due to the fact that more body mass provides a larger potential site area for cancer development. The present Life Ahead analysis of weight is statistical only and not the more basic Global type analysis preferred for health factors. Much more study of why weight produces its debits is needed.
The Life Ahead Formulas for Risk vs. Body Weight: Polynomial equations of 1st to 3rd powers were fitted to the above noted combined research results for risk of death from all causes, coronary heart disease, and death from causes other than cardiovascular disease to BMI values from 18 to 40. Results for risks of colorectal, breast, prostate, endometrial, ovary, cervical cancer and cancer from all causes were related to differing functions of BMI. These formulas are detailed following the text for those interested. The correlations were designed to achieve highest feasible accuracy, most correlation coefficients were above 0.90, and most standard deviations of actual from correlated values were small. Because morbid obesity certainly will result in higher health risks than the highest BMI measured in the above research, risk ratios were extrapolated conservatively to higher than these formula levels for BMI above 40 to 44.
The Lew and especially the Calle study noted that risk ratios for smokers differed somewhat from those of non-smokers. The Life Ahead base population is a population of non-smokers, and thus the present valuation best values risks for non-smokers. The Life Ahead method automatically recognizes that risk ratios at the higher risk level of smokers will be lower percentage wise, but will start and maintain at a much higher absolute levels than those for non-smokers.
Both the Lew and Calle studies noted that cardiovascular risks (but not cancer risks) for body weight were higher than average for those younger, and lower than average for those older. This effect is recognized in Life Ahead by application of an exponent to the average weight related risk derived from the research data. This exponent is about 1.6 at age 55 declining to 0.97 at age 70, and to 0.82 at and above age 85. Formulas for this age-related expression of this risk are applied to results from the appended formulas at each computed year of age to provide a more accurate but smoothed representation of the actual research results
The high cardiovascular risk associated with overweight for younger people may have been poorly recognized. Life Ahead can forecast that a 35 year old much overweight man that also has two or more other high risks from causes such as smoking, poor fitness, family history, or cholesterol can have a 50+% likely heart attack and will probably die before age 50.
Formulas for BMI and Body Weight: Valid for 18-40 BMI only. r =correlation coefficient; sd = std deviation in %
For Cardiovascular Disease and Death from All Causes:
Risk Ratio = exp (a + b * BMI +c * BMI^2 + d * BMI^3)
Death from All Causes: M a= 4.585 b= -0.4553 c= 0.0142 d= -0.000129 r=0.95 sd=9.6%
W a= 5.739 b= -0.5918 c= 0.01937 d= -0.0001942 r=0.97 sd=6.6%
Cardiovascular Death M&W a= 4.50 b= -0.4524 c= 0.0145 d= -0.000142 r=0.93 sd=7.1%
For Cancer: Useful mainly for BMI from 20.5 to 40 only.
Risk Ratio = exp(a + b * (BMI -20.5) ^ c
Cancer all Causes M a= -0.040 b= 0.000953 c= 2 r= 0.99 sd= 2.4%
W a= +0.0229 b= 0.001025 c= 2 r= 0.98 sd= 3.0%
Colorectal Cancer M a= -0.044 b= 0.0366 c= 1 r= 0.98 sd= 5.2%
W a= -0.040 b= 0.0195 c= 1 r= 0.91 sd= 7.1%
Prostate Cancer M a= -0.032 b= 0.00156 c= 2 r= 0.92 sd= 7.8%
Breast Cancer W a= +0.0183 b= 0.0183 c= 1 r= 0.98 sd= 5.2%
Endometrial Cancer W a= -0.120 b= 0.0861 c= 1 r= 0.94 sd= 12.6%
Ovary Cancer W a= +0.0438 b= 0.00121 c= 2 r= 0.79 sd= 13.2%
Cervical Cancer W a= +0.0263 b= 0.0134 c= 1.5 r= 0.93 sd= 17.4%
All Genital Cancer W a= -0.0121 b= 0.0706 c= 1 r= 0.99 sd= 5.0%