Main Menu              Health Library         

GLYCEMIC LOAD, HEART DISEASE and CANCER

Abstract:  The Glycemic Load or GL measures how dietary foods increase blood glucose or blood sugar.  Research now confirms that diets high in GL can increase not only the risk of diabetes but can independently increase the risk of heart disease and some types of cancer by 50%.  These are increases in risks beyond those now recognized for dietary content of saturated and transfats and other nutrient factors.  Life Ahead provides a convenient method for estimating the GL of any entered diet from the included values for the 340 foods included in its food library.  It also identifies probable changes in risks of major disease and Well-Days of life for changes in GL. Free dietary sugar as that in soft or other drinks has a larger effect on risk of diabetes and heart disease than is estimated from the GL.  Life Ahead thus provides a method for health-valuing this added effect for certain forms of sugar.

Carbohydrates can produce increases in levels of blood sugar or glucose, and in the body's response of insulin development  to control these levels.  But different types of carbohydrates produce differing blood sugar and insulin responses. The Glycemic Index directly measures the effect of different types of carbohydrates in foods in producing this blood glucose response.  Foods with a high glycemic index tend to increase blood sugar and the accompanying response in insulin that attempts to control this.  Over time this can produce a resistance to or decreased effect of insulin in controlling blood sugar. An increase in insulin resistance can lead to development of type 2 diabetes that substantially increases risk of cardiovascular diseases. The Glycemic Index concept dates only from about the 1970's, and should provide a more accurate basis for assessing behavior and risk than do previous  classifications such as via simple and complex carbohydrates.

The Glycemic Load of a food multiplies the Glycemic Index by the amount of carbohydrate present or by the amount of carbohydrate present less the amount that is fiber.  It seems logical that a food containing a tiny amount of carbohydrate will not produce the blood sugar response of one with a large amount of carbohydrate having the same glycemic index. Thus the Glycemic Load provides for this needed adjustment. But much earlier discussion about the glycemic concept was limited to use only of the Glycemic Index.  Glycemic Load values usually are the more significant correlate with disease because Glycemic Index values measure only a part of the total glycemic effect.  Thus although Life Ahead identifies both the Glycemic Index and Load of overall diets, it now identifies values of each food and risk of disease based on the Glycemic Load.  But a Glycemic Index value of any included food can be obtained by dividing the optionally displayed Glycemic Load value in the program's diet nutrient display by its grams of carbohydrates minus its grams of fiber, and multiplying by this by 100.

Table A following provides the actual results of all useful research found to date relating Glycemic Index and Load values to risks of cardiovascular diseases and cancer.  Although this research has usually been considered only indicative, the global analysis following suggests that glycemic load probably is a quite significant risk factor that deserves recognition in identifying the relative health values of various foods.

Glycemic Load, Heart Disease and Diabetes:  Three studies, #1 to 4 in Table A following are found relating glycemic index and load to Heart Disease.  Study #1 from the Nurses study of 75,000 women is the major reference, Study #4 is a more detailed later result on this same population.   A dose related and significant heart disease risk of 1.57 was found for a difference of 117 to 206 in glycemic load.  This risk for minimum adjustment was obtained in spite of an obvious inverse inter-relationship of the load value with dietary fats.  A risk was 1.98 was computed after adjustment for this.  The result of study #2 only from glycemic index is statistically consistent with that of study #1 within its wide error range of 0.67 to 1.87 but does not have the resolving accuracy to be statistically useful. Study #3 included a difference in glycemic load of only 12 units vs the difference of 89 units in study #1.  The results are consistent with the risk of only about 1.10 that would be expected for this small difference in load, but again are not confirmed with adequate statistical significance. The results of the three (or four) studies point toward a possible risk of at least 1.5 for a difference glycemic load of 89 units.  The latest paper #4 on the largest population found a risk of 1.9 times for a difference of 72 in glycemic load.

Although taken by themselves, results of studies 1-4 would not meet the usual need for including a factor in Life Ahead, the results of studies #10 and 11 relating Glycemic Load and Index to Diabetes add appreciably to the strength of the total valuation.  Study #10 found a significant risk for glycemic load in developing a women's risk of 1.47 times for diabetes. Study #11 found a significant increase in risk of 1.37 times for men's diabetes related to glycemic index. In addition study #11 found the very high risk of 2.17 for glycemic load when also including low fiber. 

Diabetes increases the risk of heart disease mortality by four to ten times for women, and  2 to 3 times for men. (see diabetes).  Thus these confirmed and significant increases in risk of diabetes for higher glycemic load translate directly into a further and significant increase in risk of heart disease.  A combination of the study risks in studies #10-11 with those from #1-3 confirms with reasonably high probability that an increase in glycemic load will increase risk of heart disease significantly.

A heart disease risk value of 1.5 for a difference in glycemic load of 89 (206 minus 117) is selected as a conservative valuation basis to be used in Life Ahead.  This computes to a risk in heart disease and assumed heart disease mortality of 1.00457 per unit increase in glycemic load.  Diabetes is well known to increase risk of stroke by 2 to 4 times. Thus it seems highly likely that increases in glycemic load will produce an increased risk of stroke.  Thus although specific research about glycemic load and stroke is not available, Life Ahead assumes that the risk established for heart disease probably will extend to and be similar for stroke.

Glycemic Values and Colorectal Cancer:   There is a substantial body of evidence showing that insulin should enhance the growth of factors involved in colon cancer.  This was shown in an extensive analysis by Giovannucci (J Nutr 2001, 131:3109S).  Unfortunately, the available research directly relating glycemic index to colon and other colorectal cancer is confusing and statistically inconsistent.  Large case control studies #20 and #21 each found significant risk ratios averaging about 1.7 for both glycemic index and glycemic load.  Yet the Canadian study #22 found no effect.  And the study #23 of adenomas that can be  precursors of cancer also found no effect.

A problem with Study #22 is that the increase in glycemic load was mostly due to differences in energy intake of the respective populations.  Calorie intake of the high glycemic load group was nearly twice that of the low glycemic load group. Thus the difference in glycemic index involved was small. Also, the only risk values provided were those extensively adjusted statistically for 12 different factors that produced in total large error margins in result of 0.73-1.53.  Such extensive adjustment risks corruption via inter-correlation of variables.  Other research on cancer factors experienced the inconsistency of findings on adenomas with those for cancer found for study #23. 

Giovannucci cited a number of studies showing that colorectal cancer was related to type 2 or non-insulin dependent diabetes. This leads support to the hypothesis that increased glycemic index that is related to diabetes probably also will be associated with an increased risk of cancer. And a number of other studies showed relationships between the insulin that develops with higher glycemic values and cancer.

Thus despite the confusion of the above four direct studies, a recognition of some probable effect of glycemic load on colorectal cancer is included in Life Ahead now as a compromise to caution the health interested person. A risk value of 1.4 for a difference in 100 in glycemic index or 1.00337 per glycemic unit is included as nearly statistically consistent with results of studies #20-22.  This assumption usually will develop only a small effect on overall risks of major disease and Well-Days of life.

Glycemic Load and Breast Cancer Risk:  The three available reports on breast cancer risk show little if any effect of either glycemic index or load on risk.  Large study 30 found a significant risk of 1.3 to 1.4, and the earlier report on the Nurses study found a 1.45 risk of those have a BMI above 25. But the later report of the large Nurses study #31 in the table following was decidedly negative on a useful effect of GI on breast cancer risk. .

A few studies have related carbohydrates and sugars to risk of breast cancer with mixed results. Glycemic Load is substantially related to overall carbohydrate.  A recent study (Cancer Epidemiol Biomarkers Prev 2004:1283) found a very large risk of 2.2 for population quintiles of carbohydrate.  Unfortunately this very high risk was for such  small change in carbohydrate of only from about 50 to about 65% that a casual relationship appears dubious.  Such an effect would suggest improbably that a near zero risk of the cancer would result for carbohydrates at levels below 30% based on any type of extrapolation. In contrast, the larger above Nurses study found no effect or possibly an inverse effect of carbohydrate for a larger difference in carbohydrate. The implication here is that the true casual factor involved in breast cancer risk has not been identified.  No effect of glycemic factors on breast cancer risk can be accepted at this time for use in Life Ahead.

Glycemic Load and Other Causes of Cancer:  Studies 40, 50, 60,70, and 80 show consistent and mostly significant effects of glycemic load and index on five different types of cancer.  Although the usual three studies needed for each cause are not available now and probably will not be in the near future, the consistency of positive effects for these five quite separate major disease groups gives confidence that the effect of glycemic load in producing cancer is extensive. The effect of glycemic load of 2.7 on endometrial and 1.7 on ovary cancer is quite substantial.  Life Ahead does not yet include an effect of glycemic load on all other causes of cancer and thus overall the effect of this factor on all causes of cancer is probably quite conservative. The effect of glycemic load is a potentially important factor for the risk of low fat meat, because meat intake can lower the glycemic load of a diet significantly.  This effect could partly or even mostly offset the harmful effect of saturated fats.  This has been a claim of the Atkin's diets.    

Values of the Glycemic Load:  Life Ahead computes glycemic load via the procedure recommended in the "Revised International Tables of Values" that subtracts fiber from the grams of carbohydrate included in a food. Most research studies indicated that total carbohydrate was used in the computation of load. This means that the adjusted values of the glycemic load used in the Tables and in Life Ahead may be somewhat lower than those cited in the research, and may be somewhat lower than some published elsewhere. Life Ahead portions and those usual in the US also averaged smaller in size than those used in the "International Tables" and values used herein were adjusted to actual portion sizes cited. This also accounts in part for the lower values computed by the program than those sometimes quoted for glycemic load. The data base on glycemic load is not nearly as complete as that for other nutrients.  But most of the 320 values included in Life Ahead for different foods should be at least approximately valid.  

Despite much discussion of the glycemic concept, and many recommendations to "Reduce eating of high Glycemic Index foods" no documented guide about desirable healthful daily dietary levels of the key measure of Glycemic Load (GL) was found. Various guides and some computer software do provide Glycemic Index values of foods that can be helpful to diabetics. But no answer to the key question "What level of GL is a desirable healthful target for most health-interested persons to achieve?" was found. No answer was found to the further key question of "What benefit or detriment to overall health will accrue from making a given change in our dietary GL?"   Life Ahead now may provide some useful answers to these questions.

Before addressing these questions, it is pertinent to observe that the Glycemic Load is an absolute value expressed as grams of carbohydrates adjusted by a factor. An appropriate value will be lower for a average women than for an average man, and should vary for individuals of differing height in a way similar to that for optimum weight and calorie intake varies for sex and height. Second, a value will be expected be lower when a person is dieting than when on a weight maintenance diet. Thus a target value must be related to that appropriate for an average healthy weight maintaining diet. This is now estimated in Life Ahead at 2140 calories and 1340 calories per day respectively for a typical healthy US 50 year old man and woman of healthful weight. And third, it will depend on a person's choice of a normal or low carbohydrate diet. A low carbohydrate diet may involve a tradeoff between eating less carbo that can reduce Glycemic Load and consuming more fat that may have a larger health negative. And Glycemic Load target may be much lower for one with type 2 diabetes than for a disease free person. Thus it is unlikely that any definitive value for Glycemic Load can be proposed that is appropriate for all individuals.

Glycemic Risk and Sugar:  Some large studies have found that soft drinks that include substantial amounts of dissolved sugar have a significantly larger effect on the risk of diabetes than would be forecast from their glycemic load.  Life Ahead thus includes this added risk for sugar that is directly consumed or included in soft drinks.  See the Life Ahead paper on diabetes for more on this.

The Potential Health Benefits for Reducing Dietary Glycemic Load:  The Life Ahead Demo diets having the highest glycemic index when adjusted to average weight maintenance calories were the US typical weekly diet #1 at 137 GL and the vegetarian daily diet #6 at 134 GL.  The two low fat diets similarly adjusted had GL values of 103, and the two DASH diets at an average GL of 107.  The key foods producing GL in weekly diet #1 were bread, 25 GL; cranberry juice 14 GL; jam 13 GL, and coca cola 10 GL.  The total of 62 GL for these four items probably could be reduced in half by not-difficult substitutions.  Daily diet #6 included 28 GL in bread and potatoes potential for reduction. A general review of these diets suggested that a daily GL of about 110 usually should be potential for this average man from most diets by eliminating sugars and reducing highest GL foods.  For an average woman with a usual 67% of the calorie intake of a man a goal would be about 75 GL from a normal-carbo diet. Atkins weight loss and maintenance diets at men's calorie level were 22 and 65 GL respectively and thus those using low carbo diets will usually obtain low values the GL.

Life Ahead provides for sorting the foods in any entered diet - and also those in the entire food library - in order of their contribution to GL.  Thus those contributing most are easily identifiable.  Thus a first approach is the simple substitution of other foods for those of highest GL.  A reduction in overall GL of 20 units that often can be obtained via such substitutions would on the above bases reduce risk of heart disease and several causes of cancer by about 10% and women's genital cancer by 15%.  Although these are modest benefits compared to those from many other potential life style and diet changes the improvement also would reduce risk of diabetes.

But much larger negatives to health probably can be caused by eating very high GL foods.  Some examples follow of the increase in daily GL and accompanying risk of disease for the continuing daily consumption of one portion of the following  foods having high GL levels:                       

Various combinations of popular Fast Foods as hamburgers, fries and milk shakes and pies can increase a daily food intake GL by up to 60 units or more.  Consumed regularly over long periods of time this could increase risk of both heart disease and cancer considerably.  Risks of GL probably  accrue from mechanisms independent of those involved for saturated and transfats.  Thus these risks for GL probably will multiply further the high risks for these foods due to their other included nutrients.  Thus Life Ahead now includes the effect of the GL of each food in computing overall risks of both individual foods and diets on major disease and on likely Well-Days of average person life.

The above risks identify the specific computed effects of glycemic load per se. The high risks are due in part to the fact that these specific Fast foods also are included as quite large serving sizes, and risks are computed per serving.  A Life Ahead actual computation of risks of these foods as added to a diet will include not only the effect of glycemic load, but also effects of the other 22 nutrients in the foods valued by the program that differ from those in a base diet. Because this will include effects of added vitamins, minerals, saturated and other fats, the program will not reproduce the above values directly. 

 But Life Ahead shows that above average GL foods added regularly to a diet can have a significant effect on long range risk of major disease and health. This is shown further in the next following table. This shows Life Ahead computed changes in risk of heart disease and cancer for adding one daily portion of the listed food to the typical diet of a 50 year old US man from this age forward.  Well-Days are the differences in estimated future major disease free years of life.  Risks in the center columns are those computed without including the effect of glycemic load.  Values at right are the increase in risks for adding these foods when including the effect of glycemic load. Note that values in the right columns show changes in risk, not absolute values of risks.  For example the net heart disease risk for the Bran Muffin with glycemic index included is -15.7%, still a favorable value. 

The bran muffin benefited risk vs the average risk of the diet mostly from its high omega-3 and low saturated fat content.  The candy detracted from the average risk mostly due to the reverse of low omega-3 and high saturated fat.  The other foods had risks only moderately different from the average diet. But when recognizing of the probable effect of glycemic load the risk of heart disease from addition of these foods increased further by 7-10% and risk of cancer increased further from about 2-4%.  The lower value for sugar is for only about  a tenth of usual amount considered in grams.  Taken in similar amounts it would have a similar negative effect.  But note that these foods are those of higher than usual GL or GI.  The increased health risk for most foods due to their GL will be modest.

The above analysis illustrates once again the fallacy than can result from attempting to value the health of a food  based on the amount of only a single nutrient. The bran muffin has the highest GL   But it still computes as a net health-beneficial food overall.  Even after recognition of the GL debit, it's continued daily use in this diet could contribute a 15.7% and 2.4% lowering in risks of heart disease and cancer respectively, and a useful net health benefit of 395 Well-Days. A challenge to the food industry is to develop foods such as muffins that are lower in GL that still retain their other benefits. 

Available research did not consider the effect of GL duration on risk. The Nurses study followed women for 10 years, and included 3 dietary assessments during this period.  Thus a duration of diet GL somewhat beyond 10 years probably was measured.  The case control studies included only a single diet assessment and this probably identified a GL duration of fewer years than this as people can change their usual diets.  It is highly likely that any effect of GL on risk of heart disease will be duration related over long periods of time.  This is shown strikingly in the effect of diabetes on heart disease on this Life Ahead web site that confirms a continued effect of duration extending for more than 30 years. Control of blood glucose has been shown to be related to the gradual development of the atherosclerosis that is involved in heart disease.  Thus although lifetime risks for GL may be higher than those now estimated, short term consumption of high GL foods probably will not produce much change in long term risk.  Cancer risks usually develop only from long periods of exposure, and thus cancer risks for GL probably will be similarly related to duration of diet. 

A duration effect for GL is not now included in Life Ahead computations of risk.  But because benefits probably are duration related it is unlikely that an occasional consumption of high GL and other high risk factor foods will produce a measurable health effect. Rather, a concern is noted for the regular consumption of such foods for long periods of time.  Also, because health targets for GL can vary widely for different people Life Ahead does not include the benefits of improving GL levels to some GL target in computing benefits for "Good" health habits. Thus benefits for improving GL to targets can be additional to those for all other health habits now computed. 

Life Ahead includes the effect GL as one of its nutrient factors that affects risks of major disease. The program shows GL levels of each diet and compares these with the usually desirable maximum levels of 110 for average men and 75 for average women for normal-carbo diets.  Users should sort their diets on the Glycemic Load option to view those foods that contribute  most to this GL, and consider substituting lower GL foods for these where appropriate.  The GL of various food alternatives can easily be viewed by sorting on the Food Library values. But foods should be tested for the overall health benefits before removing them just for their GL.

Life Ahead provides for valuing the health benefit of any change in a diet from a combination of all nutrient factors now included, and this now includes a change in Glycemic Load (GL).  First, download and install the free program from  Download LifeAhead.  To view how the program works, click the Demo program, highlight either man or woman, and access the 2nd listed option Enter or Modify a Diet Entry.  Ask 'View Results of Demo Diets' and enter 1 for weekly #1, the typical US diet.  This will bring up the diet entry screen with diet entered for this particular diet shown at right.  Ask to 'Continue Demo", and on next screen click 'Compute Outlook with this Diet'.  This will immediately display the result screen for this person age and diet.  (The modern computer actually processes ten of thousands of computations in less than one second to obtain these results)  To view diet deficiencies, click 'Actions for More Well-Days' and then 'Diet Analysis'  This will display the amounts of all nutrients in this diet vs targets for best health.  The value for Glycemic Load now is the last listed item in this group, but often will not show a target because of the various factors discussed above.  If you are a diabetic you have a particular need to keep you intake of Glycemic Load low.

To view a result from your own diet you should first enter your name, age and other factors asked from the first listed option, and enter all questions asked on successive screens.  (You do not have to enter everything asked because the program will assume average population values for items not entered).  But you must complete the exercise entries. After saving this information, restart the program and ask to enter your own usual diet as described, either for a day or for a usual full week.  Once this is done you can explore how your diet develops all nutrient factors, how these compare with those needed for best health.  Your Glycemic Index will be computed for each diet or diet modification entered.