OPEN LETTER TO DAVID GILLESPIE – AUTHOR OF SWEET POISON
Dear David,
You have requested identification of the 'errors and dubious claims' mentioned in the Nut-Net FAQ about Sweet Poison.
By
way of introduction, ‘Nut-Net’ is the abbreviation of Nutritionists
Network, a professional nutrition email discussion list with more than
800 nutritionists and dietitians as subscribers. As a service to the
general public, Nut-Net develops nutrition related FAQs which are made
available through the Nutrition Australia website. I coordinate the
writing of the Nut-Net FAQs, which represent a consensus, or if one
cannot be reached, a majority viewpoint of those Nut-Net subscribers who
contributed to the development of each FAQ. (I should also point out
that no subscriber expressed a dissenting view on any aspect of the
final wording of the FAQ on Sweet Poison.)
A similar procedure was used to develop this response, so it constitutes the Nut-Net position.
The following is not a complete list of the errors detected in Sweet Poison,
but it should give you an indication of the problems Nut-Net has with
the quality of information in your book and Ockham’s Razor program. The
identified problems range from relatively minor mistakes (or
unsubstantiated claims) to major misinterpretations of the scientific
literature.
(i) On page 8 of Sweet Poison
an attempt is made to draw an association between a presumed increase
in sugar consumption and the onset of obesity and several chronic
diseases in western nations. An association does not prove cause and
effect—it can only lead to hypotheses that then need to be rigorously
tested. The state of the science with respect to the effects of fructose
on human health is immature, as demonstrated in later points in this
response. It is at least premature, and possibly wholly inappropriate to
conclude, as you do, that fructose is the sole cause of the epidemics
of obesity, type 2 diabetes and heart disease.
Further, in his book In Defense of Food (p. 111) Michael Pollan1
points out that a reduction in tax on sugar in 1874 led to a massive
increase in sugar consumption in England, to the extent that it
comprised one-sixth of total energy intake by the end of the 19th
century. This is similar to the current levels of intake in western
nations, yet according to Pollan: '… rates of diseases like cancer and type 2 diabetes are considerably higher today than they were in 1900'.
We
are also unaware of reports of major epidemics of obesity or heart
disease in England in the late 19th/early 20thcenturies. If Pollan is
correct, it appears that intakes of fructose similar to those in modern
western nations do not inevitably lead to these chronic health problems.
Finally, there is doubt about the strength of the association (or even
if an associationexists) between recent trends in fructose consumption
and the obesity epidemic in developed nations.
For
example, Alicia Sim and Alan Barclay presented a paper at the 2010
conference of the Dietitians Association of Australia in which they
concluded that 'total fructose consumption is inversely associated with overweight/obesity in Australia, the UK and Japan since the 1970s'. You have previously been made aware of this paper.
(ii) On page 14 (and earlier) the claim is made that 'all food is essentially glucose' (i.e.
that apart from the protein that is used for growth and repair, and
also for the production of hormones and enzymes, food is converted to
glucose as an energy source. This is not correct—even for the
macronutrients—and it is entirely inappropriate with respect to the wide
range of micronutrients and also dietary fibre. In terms of the
macronutrients, fat cannot be converted to glucose (although the reverse
is possible), and protein consists of two types of amino acids—the
glucogenic (which can be converted to glucose) and ketogenic (those that
can be converted to ketones but not to glucose).
(iii) On page 58 the following claim is made about a study conducted by Reiser et al. [Am. J. Clin. Nutr. (1985); 42 (2): 242]:
'The
study had to be terminated when four of the men in the fructose group
developed cardiac problems ranging from severe tachycardia (his heart
rate tripled) to mild heart attacks, within the first 11 weeks.'
This is not correct—the authors of this paper point out that there was a '... lack of relationship between the onset of the abnormalities and the type of dietary carbohydrate'.
(iv) On page 59 it is stated that '...
in 1984 the American Diabetic Association (ADA) recommended that
diabetic patients be given foods that used fructose as a sweetener
rather than sugarI'. Page 60 includes the claim that 'by 2002 the ADA had completely reversed its recommendation on fructose, saying that fructose should be completely avoided'. The actual wording of the ADA position statement [Diabetes Care (2008) 31:1 S61-S78] is:
'In
individuals with diabetes, fructose produces a lower postprandial
glucose response when it replaces sucrose or starch in the diet;
however, this benefit is tempered by concern that fructose may adversely
affect plasma lipids. Therefore, the use of added fructose as a
sweetening agent in the diabetic diet is not recommended.'
That
is, the ADA simply no longer recommends that people with diabetes use
added fructose as a sweetener. This does not equate to 'fructose should
be completely avoided'.
Further, the ADA2
makes crystal clear its position that sugar (which is the major source
of added fructose in the Australian diet) is not the cause of type 2
diabetes:
'The
myth that sugar causes diabetes is commonly accepted by many people.
Research has shown that it isn't true ... The biggest dietary risk
factor for developing type 2 diabetes is simply eating too much and
being overweight - your body doesn’t care if the extra food comes from
cookies or beef, it is gaining weight that is the culprit.'
This
position statement is diametrically opposed to your claim that
consumption of added fructose is the sole cause of type 2 diabetes.
(v) On page 78 the claim is made that 'every gram of fructose we eat is directly converted to fat'.
This is not correct—fructose may go down the gluconeogenic pathway
(leading to its conversion to glucose and subsequent storage in the
liver as glycogen) or the fructolytic pathway (resulting in the
production of fat). The available evidence suggests that fructose may be
preferentially converted to glycogen until liver glycogen is
replenished. It appears that only then will the fructolytic pathway
predominate [Biochem. J. (1988) 251 (3): 795–802].
(vi)
Your Ockham’s Razor program includes the claim that as a result of the
previously mentioned 1985 study by Reiser et al. no further fructose
feeding studies were conducted on humans. This is incorrect—fructose
feeding studies involving human participants continued (and are still
being conducted).
(vii)
Your Ockham’s Razor program also states—without supporting
evidence—that there has been an increase of 30% in food intake by
Australians in the past 30 years. Dr Rosemary Stanton has provided
science-based evidence against this claim (and you have been made aware
of this evidence).
(viii)
Your Ockham’s Razor program also includes the claim—again without
supporting evidence—that fructose provides nearly 20% of energy intake
in Australia, while also stating that Australian intake is about
two-thirds that of the United States. The scientific literature [see for
example Medscape J. Med. (2008); 10(7): 160] indicates that fructose
provides ~10% of energy intake in the US. Assuming that you are correct
about Australian intake of fructose being substantially less than in the
United States, average intake in Australia would be somewhat less than
10% of total energy, not 'nearly 20%'.
(ix)
Sweet Poison includes the claim that added fructose is a poison at any
dose, and your Ockham’s Razor program calls on regulatory authorities to
'… immediately ban added fructose as a food'. There is no
evidence in the scientific literature to suggest that added fructose is
toxic at any dose. National and international health related
organisations such as the World Health Organisation, the Australian
National Health and Medical Research Council and the American Heart
Association all recommend maximum intakes of sugars; none of these (or
any other reputable organisation) recommends complete avoidance of added
fructose.
(x)
You also claim that fructose is undoubtedly addictive to humans (and
therefore that sucrose and high-fructose corn syrup are addictive). The
evidence on this is equivocal, to say the least. For example, a recent
meta-analysis [Clin. Nutr. (2010) 29(3): 288-303] concludes that 'There is no support from the human literature for the hypothesis that sucrose may be physically addictive ...'
(xi) On page 172 of Sweet Poison the claim is made that 'all processed foods contain some sugar'.
This is not correct. The following is an incomplete list of processed
foods that generally do not contain added sugars: Canned and dried
soups; canned vegetables; frozen vegetables; dried vegetables; frozen
fruit; dried fruit; most frozen ‘TV dinners’; cured meats; Vegemite;
potato crisps; frozen potato chips; unflavoured corn chips; pasteurised
milk; cheese; natural yogurt; unsweetened fruit-flavoured yogurt; cream;
butter; margarine; cocoa powder; instant coffee; some breakfast cereals
(e.g. All-Bran, Weetbix, rolled oats); and many savoury
crackers/biscuits.
(xii) On page 174 you state that 'non-caloric sweeteners ... contain the same number of calories per gram as any other carbohydrate'. This is not correct—“non-caloric” implies that no energy is provided by these artificial sweeteners.
(xiii) You provide the following definitions for units relevant to human metabolism:
a. Joule – 'The metric equivalent of the calorie is a joule, and calculated using Einstein’s famous equation E=mc2 …”'(p. 203); and
b. Watt – 'Measures electrical energy consumed per hour ... a 60-watt light bulb uses 60 watts of energy every hour' (p. 137).
Both
these definitions are incorrect—the joule is the basic SI unit of
energy (and also of work), and is defined as the energy transformed (or
work conducted) when a mass of one kilogram is accelerated at one
metre-per-second-squared over a distance of one metre. The reference to
‘E=mc2’ is wholly inappropriate; we wonder if you have confused
Einstein’s equation with the equation ‘W=f.d’ (work equals force times
distance) which is appropriate in this context.
Similarly,
the watt does not measure 'electrical energy consumed per hour'—it is a
measure of power output, regardless of the form of energy, and measures
the quantity of work conducted (or energy transformed) per second, not
'per hour'. One watt of power output occurs when one joule of work is
done per second. (And the claim that 'a 60-watt light bulb uses 60 watts
of energy every hour' is meaningless.)
In addition to containing these errors (among others), Sweet Poison
selectively refers to the scientific literature and, as explained
above, occasionally grossly misinterprets the literature to develop an
exaggerated case that fructose is the sole cause of obesity, type 2
diabetes and heart disease.
The
actual state of the science is that the hypothesis has been advanced
that high (note the word high) intakes of fructose may be implicated in
these chronic health problems, but there is no consensus that fructose
is the sole, or even major cause. For example, Johnson et al. [Endocrine
Rev. (2009); 30: 96–116] hypothesise that '... excessive fructose intake (>50 g/day) may be one of the etiologies of the metabolic syndrome and type 2 diabetes'.
That is, it is plausible that high (but not low or moderate) intakes of
fructose are contributing to adverse health outcomes.
This hypothesis
now needs to be subjected to rigorous testing. This is quite different
to your conclusion that dietary fructose is unquestionably the sole
cause of the major chronic health problems afflicting developed nations.
Another
illustration that the debate is far from over is the exchange between
the authors of a meta-analysis [Am. J. Clin, Nutr. (2008); 88: 1419-37]
and the authors of an accompanying editorial [Am. J. Clin, Nutr. (2008);
88:
1189-90]. The meta-analysis concludes that fructose is safe at
relatively high intakes and may even be beneficial at moderate intakes,
while the editorial questions these conclusions as follows: 'Whereas
some ... (adverse) effects (of fructose) have been reported only in
animals, these findings raise important questions about the safety of
high doses of fructose in humans'. Even if the non-peer-reviewed
editorial is given greater weight than the peer-reviewed metaanalysis
(and there is no reason why it should), its contents do not equate to
your claim that fructose is indisputably a poison at any dose.
Another major problem is that Sweet Poison
ignores (in fact denigrates) the Dietary Guidelines for Australians
published by the National Health and Medical Research Council. These are
science-based, and constitute the most appropriate diet-related advice
currently available to Australians. Although your book includes
reference to the value of dietary fibre, and mentions the problems
associated with consumption of trans fats, it does not address the major
health problems associated with excessive consumption of salt and
alcohol. It also (inappropriately) includes the claim that physical
activity is of no value for people whose aim is to lose weight, and it
condones (even promotes) high fat consumption (other than trans fats).
In summary, Sweet Poison
does not reflect the current state of scientific knowledge with respect
to the health effects of dietary fructose. As a result, the Nut-Net FAQ
will continue to recommend against this book.
Sincerely,
Chris Forbes-Ewan
Coordinator Nut-Net FAQs
22 Oct. 10
1
Michael Pollan has written extensively on food and nutrition. His
background research for In Defense of Food included consulting eminent
nutritionists such as Walter Willett, Joan Gussow and Marion Nestle.
2 www.diabetes.org/food-and-fitness/food/what-can-i-eat/sweeteners-and-desserts.html
Dear David,
You have requested identification of the 'errors and dubious claims' mentioned in the Nut-Net FAQ about Sweet Poison.
By
way of introduction, ‘Nut-Net’ is the abbreviation of Nutritionists
Network, a professional nutrition email discussion list with more than
800 nutritionists and dietitians as subscribers. As a service to the
general public, Nut-Net develops nutrition related FAQs which are made
available through the Nutrition Australia website. I coordinate the
writing of the Nut-Net FAQs, which represent a consensus, or if one
cannot be reached, a majority viewpoint of those Nut-Net subscribers who
contributed to the development of each FAQ. (I should also point out
that no subscriber expressed a dissenting view on any aspect of the
final wording of the FAQ on Sweet Poison.)
A similar procedure was used to develop this response, so it constitutes the Nut-Net position.
The following is not a complete list of the errors detected in Sweet Poison,
but it should give you an indication of the problems Nut-Net has with
the quality of information in your book and Ockham’s Razor program. The
identified problems range from relatively minor mistakes (or
unsubstantiated claims) to major misinterpretations of the scientific
literature.
(i) On page 8 of Sweet Poison
an attempt is made to draw an association between a presumed increase
in sugar consumption and the onset of obesity and several chronic
diseases in western nations. An association does not prove cause and
effect—it can only lead to hypotheses that then need to be rigorously
tested. The state of the science with respect to the effects of fructose
on human health is immature, as demonstrated in later points in this
response. It is at least premature, and possibly wholly inappropriate to
conclude, as you do, that fructose is the sole cause of the epidemics
of obesity, type 2 diabetes and heart disease.
Further, in his book In Defense of Food (p. 111) Michael Pollan1
points out that a reduction in tax on sugar in 1874 led to a massive
increase in sugar consumption in England, to the extent that it
comprised one-sixth of total energy intake by the end of the 19th
century. This is similar to the current levels of intake in western
nations, yet according to Pollan: '… rates of diseases like cancer and type 2 diabetes are considerably higher today than they were in 1900'.
We
are also unaware of reports of major epidemics of obesity or heart
disease in England in the late 19th/early 20thcenturies. If Pollan is
correct, it appears that intakes of fructose similar to those in modern
western nations do not inevitably lead to these chronic health problems.
Finally, there is doubt about the strength of the association (or even
if an associationexists) between recent trends in fructose consumption
and the obesity epidemic in developed nations.
For
example, Alicia Sim and Alan Barclay presented a paper at the 2010
conference of the Dietitians Association of Australia in which they
concluded that 'total fructose consumption is inversely associated with overweight/obesity in Australia, the UK and Japan since the 1970s'. You have previously been made aware of this paper.
(ii) On page 14 (and earlier) the claim is made that 'all food is essentially glucose' (i.e.
that apart from the protein that is used for growth and repair, and
also for the production of hormones and enzymes, food is converted to
glucose as an energy source. This is not correct—even for the
macronutrients—and it is entirely inappropriate with respect to the wide
range of micronutrients and also dietary fibre. In terms of the
macronutrients, fat cannot be converted to glucose (although the reverse
is possible), and protein consists of two types of amino acids—the
glucogenic (which can be converted to glucose) and ketogenic (those that
can be converted to ketones but not to glucose).
(iii) On page 58 the following claim is made about a study conducted by Reiser et al. [Am. J. Clin. Nutr. (1985); 42 (2): 242]:
'The
study had to be terminated when four of the men in the fructose group
developed cardiac problems ranging from severe tachycardia (his heart
rate tripled) to mild heart attacks, within the first 11 weeks.'
This is not correct—the authors of this paper point out that there was a '... lack of relationship between the onset of the abnormalities and the type of dietary carbohydrate'.
(iv) On page 59 it is stated that '...
in 1984 the American Diabetic Association (ADA) recommended that
diabetic patients be given foods that used fructose as a sweetener
rather than sugarI'. Page 60 includes the claim that 'by 2002 the ADA had completely reversed its recommendation on fructose, saying that fructose should be completely avoided'. The actual wording of the ADA position statement [Diabetes Care (2008) 31:1 S61-S78] is:
'In
individuals with diabetes, fructose produces a lower postprandial
glucose response when it replaces sucrose or starch in the diet;
however, this benefit is tempered by concern that fructose may adversely
affect plasma lipids. Therefore, the use of added fructose as a
sweetening agent in the diabetic diet is not recommended.'
That
is, the ADA simply no longer recommends that people with diabetes use
added fructose as a sweetener. This does not equate to 'fructose should
be completely avoided'.
Further, the ADA2
makes crystal clear its position that sugar (which is the major source
of added fructose in the Australian diet) is not the cause of type 2
diabetes:
'The
myth that sugar causes diabetes is commonly accepted by many people.
Research has shown that it isn't true ... The biggest dietary risk
factor for developing type 2 diabetes is simply eating too much and
being overweight - your body doesn’t care if the extra food comes from
cookies or beef, it is gaining weight that is the culprit.'
This
position statement is diametrically opposed to your claim that
consumption of added fructose is the sole cause of type 2 diabetes.
(v) On page 78 the claim is made that 'every gram of fructose we eat is directly converted to fat'.
This is not correct—fructose may go down the gluconeogenic pathway
(leading to its conversion to glucose and subsequent storage in the
liver as glycogen) or the fructolytic pathway (resulting in the
production of fat). The available evidence suggests that fructose may be
preferentially converted to glycogen until liver glycogen is
replenished. It appears that only then will the fructolytic pathway
predominate [Biochem. J. (1988) 251 (3): 795–802].
(vi)
Your Ockham’s Razor program includes the claim that as a result of the
previously mentioned 1985 study by Reiser et al. no further fructose
feeding studies were conducted on humans. This is incorrect—fructose
feeding studies involving human participants continued (and are still
being conducted).
(vii)
Your Ockham’s Razor program also states—without supporting
evidence—that there has been an increase of 30% in food intake by
Australians in the past 30 years. Dr Rosemary Stanton has provided
science-based evidence against this claim (and you have been made aware
of this evidence).
(viii)
Your Ockham’s Razor program also includes the claim—again without
supporting evidence—that fructose provides nearly 20% of energy intake
in Australia, while also stating that Australian intake is about
two-thirds that of the United States. The scientific literature [see for
example Medscape J. Med. (2008); 10(7): 160] indicates that fructose
provides ~10% of energy intake in the US. Assuming that you are correct
about Australian intake of fructose being substantially less than in the
United States, average intake in Australia would be somewhat less than
10% of total energy, not 'nearly 20%'.
(ix)
Sweet Poison includes the claim that added fructose is a poison at any
dose, and your Ockham’s Razor program calls on regulatory authorities to
'… immediately ban added fructose as a food'. There is no
evidence in the scientific literature to suggest that added fructose is
toxic at any dose. National and international health related
organisations such as the World Health Organisation, the Australian
National Health and Medical Research Council and the American Heart
Association all recommend maximum intakes of sugars; none of these (or
any other reputable organisation) recommends complete avoidance of added
fructose.
(x)
You also claim that fructose is undoubtedly addictive to humans (and
therefore that sucrose and high-fructose corn syrup are addictive). The
evidence on this is equivocal, to say the least. For example, a recent
meta-analysis [Clin. Nutr. (2010) 29(3): 288-303] concludes that 'There is no support from the human literature for the hypothesis that sucrose may be physically addictive ...'
(xi) On page 172 of Sweet Poison the claim is made that 'all processed foods contain some sugar'.
This is not correct. The following is an incomplete list of processed
foods that generally do not contain added sugars: Canned and dried
soups; canned vegetables; frozen vegetables; dried vegetables; frozen
fruit; dried fruit; most frozen ‘TV dinners’; cured meats; Vegemite;
potato crisps; frozen potato chips; unflavoured corn chips; pasteurised
milk; cheese; natural yogurt; unsweetened fruit-flavoured yogurt; cream;
butter; margarine; cocoa powder; instant coffee; some breakfast cereals
(e.g. All-Bran, Weetbix, rolled oats); and many savoury
crackers/biscuits.
(xii) On page 174 you state that 'non-caloric sweeteners ... contain the same number of calories per gram as any other carbohydrate'. This is not correct—“non-caloric” implies that no energy is provided by these artificial sweeteners.
(xiii) You provide the following definitions for units relevant to human metabolism:
a. Joule – 'The metric equivalent of the calorie is a joule, and calculated using Einstein’s famous equation E=mc2 …”'(p. 203); and
b. Watt – 'Measures electrical energy consumed per hour ... a 60-watt light bulb uses 60 watts of energy every hour' (p. 137).
Both
these definitions are incorrect—the joule is the basic SI unit of
energy (and also of work), and is defined as the energy transformed (or
work conducted) when a mass of one kilogram is accelerated at one
metre-per-second-squared over a distance of one metre. The reference to
‘E=mc2’ is wholly inappropriate; we wonder if you have confused
Einstein’s equation with the equation ‘W=f.d’ (work equals force times
distance) which is appropriate in this context.
Similarly,
the watt does not measure 'electrical energy consumed per hour'—it is a
measure of power output, regardless of the form of energy, and measures
the quantity of work conducted (or energy transformed) per second, not
'per hour'. One watt of power output occurs when one joule of work is
done per second. (And the claim that 'a 60-watt light bulb uses 60 watts
of energy every hour' is meaningless.)
In addition to containing these errors (among others), Sweet Poison
selectively refers to the scientific literature and, as explained
above, occasionally grossly misinterprets the literature to develop an
exaggerated case that fructose is the sole cause of obesity, type 2
diabetes and heart disease.
The
actual state of the science is that the hypothesis has been advanced
that high (note the word high) intakes of fructose may be implicated in
these chronic health problems, but there is no consensus that fructose
is the sole, or even major cause. For example, Johnson et al. [Endocrine
Rev. (2009); 30: 96–116] hypothesise that '... excessive fructose intake (>50 g/day) may be one of the etiologies of the metabolic syndrome and type 2 diabetes'.
That is, it is plausible that high (but not low or moderate) intakes of
fructose are contributing to adverse health outcomes.
This hypothesis
now needs to be subjected to rigorous testing. This is quite different
to your conclusion that dietary fructose is unquestionably the sole
cause of the major chronic health problems afflicting developed nations.
Another
illustration that the debate is far from over is the exchange between
the authors of a meta-analysis [Am. J. Clin, Nutr. (2008); 88: 1419-37]
and the authors of an accompanying editorial [Am. J. Clin, Nutr. (2008);
88:
1189-90]. The meta-analysis concludes that fructose is safe at
relatively high intakes and may even be beneficial at moderate intakes,
while the editorial questions these conclusions as follows: 'Whereas
some ... (adverse) effects (of fructose) have been reported only in
animals, these findings raise important questions about the safety of
high doses of fructose in humans'. Even if the non-peer-reviewed
editorial is given greater weight than the peer-reviewed metaanalysis
(and there is no reason why it should), its contents do not equate to
your claim that fructose is indisputably a poison at any dose.
Another major problem is that Sweet Poison
ignores (in fact denigrates) the Dietary Guidelines for Australians
published by the National Health and Medical Research Council. These are
science-based, and constitute the most appropriate diet-related advice
currently available to Australians. Although your book includes
reference to the value of dietary fibre, and mentions the problems
associated with consumption of trans fats, it does not address the major
health problems associated with excessive consumption of salt and
alcohol. It also (inappropriately) includes the claim that physical
activity is of no value for people whose aim is to lose weight, and it
condones (even promotes) high fat consumption (other than trans fats).
In summary, Sweet Poison
does not reflect the current state of scientific knowledge with respect
to the health effects of dietary fructose. As a result, the Nut-Net FAQ
will continue to recommend against this book.
Sincerely,
Chris Forbes-Ewan
Coordinator Nut-Net FAQs
22 Oct. 10
1
Michael Pollan has written extensively on food and nutrition. His
background research for In Defense of Food included consulting eminent
nutritionists such as Walter Willett, Joan Gussow and Marion Nestle.
2 www.diabetes.org/food-and-fitness/food/what-can-i-eat/sweeteners-and-desserts.html
