Is High Fructose Corn Syrup Worse Than Regular Sugar?


A recent study out of Princeton University has the high-fructose corn syrup alarmists out in full force.  This study compared the effects of high fructose corn syrup (HFCS) to regular table sugar (sucrose), looking at their effects on body weight, body fat, and triglycerides (fats that float around in your blood).  The study found that the rats fed HFCS gained more weight and abdominal fat than the rats fed sucrose.  This study has strengthened the belief of some people that HFCS is contributing to obesity in our society, and that it is worse than regular sugar.  But is it really?

To answer this question, we need to take a close look at this study.  The researchers performed 2 experiments.  In the first experiment, male rats were divided into 4 groups.  Group 1 (the control group) was fed a regular diet.  Group 2 was fed the same diet, with the addition of 24-hour access to water sweetened with HFCS.  Group 3 had the regular diet with 12-hour access to the HFCS-sweetened water.  Group 4 had the regular diet, with 12-hour access to sucrose-sweetened water.  The rats were tracked for 8 weeks; weight was measured, along with food, sucrose, and HFCS intake.

You can see the results for experiment 1 in the following chart:

The rats who got HFCS for 12 hours gained significantly more weight than the other 3 groups.  At first glance, this would make you believe that HFCS makes you gain more weight than sucrose, even if you are eating the same number of calories.  However, there is a problem with these results.  Take a look again at the chart above.  If the rats fed HFCS for 12 hours gained more weight, why didn’t the rats fed HFCS for 24 hours also gain more weight?  They got HFCS for a full 12 hours more, yet didn’t gain more weight.  This is a glaring inconsistency in the results…an inconsistency that the researchers never tried to explain.

Rather than some unique effect of HFCS, a more likely explanation is one of chance.  Put on your math hat, because we need to talk about some statistics.  Researchers use statistics to get an idea of the probability that their results are due to chance.  When the scientists run their stats, they get what is known as a P value.  The P value tells you the probability that the results are not due to chance.  Usually, if the P value is less than 0.05, a scientist will call the results “significant.”  In other words, if you did the experiment 100 times, you would only see these results less than 5 times if there wasn’t a true effect.

The above only holds true if you’re doing a single comparison.  If you start comparing a bunch of groups all to each other, the probability of a fluke result dramatically increases.   The Princeton study is a perfect example.  There are 4 groups all being compared to each other.  That makes for 6 total comparisons (group 1 to group 2, 1 to 3, 1 to 4, 2 to 3, 2 to 4, and 3 to 4).  Each one of these comparisons is being tested against that 5% level.  To calculate the probability of a fluke result in this case, we calculate 1 – (0.95×0.95×0.95×0.95×0.95×0.95) = 26%.  In other words, there is a 1 in 4 chance that the greater weight gain in the HFCS-fed rats is a fluke.  I don’t know about you, but I wouldn’t put too much faith in results that have a 1 in 4 chance of being wrong.  There are ways that scientists can adjust for this, but the Princeton researchers didn’t appear to make those adjustments.  Thus, it is not surprising that there was a significant result observed in 1 out of the 4 groups…you would expect this to happen based on random chance alone.

In Experiment 2, the researchers divided male rats into 3 groups:  12-hour HFCS, 24-hour HFCS, and control.  They tracked the rats for 6 months.  Both HFCS-fed groups gained more weight and fat than the control, and also had higher triglycerides.  However, the researchers didn’t compare HFCS to sucrose in this group, so this experiment doesn’t’ say anything about whether HFCS is any worse than sucrose.  The researchers also didn’t say anything about food intake and whether the HFCS-fed rats ate more than the control rats. 

Experiment 2 also featured female rats on one of the 4 diets used in Experiment 1.  These rats were tracked for 7 months.  The following chart shows the results of the experiment:

The female rats fed HFCS for 24 hours a day gained significantly more weight than the other groups.  Now compare these results to the chart for Experiment 1 earlier.  Do you see the disparity?  In Experiment 1, the rats fed HFCS for 12 hours per day gained the most weight.  However, in Experiment 2, the rats fed HFCS for 24 hours per day gained the most weight, and the female rats fed HFCS for 12 hours didn’t gain any more weight than the other groups.  Why did the 12-hour group gain the most weight in one experiment, but the 24-hour group gain the most weight in a nearly identical experiment?  This is a glaring contradiction in the results, and a problem which the researchers did not discuss.  We also have the same statistical problem that we did with Experiment 1.  Since there are 6 comparisons, there is a 1 in 4 chance that the results are wrong (and ironically, we have 1 out of the 4 groups showing a significant result).  In fact, when we take both experiments combined, we have at least a 50% chance that the results of one of the experiments are wrong.  Out of all the comparisons being made, we would expect to see a couple groups show a significant result based on random chance…and that’s exactly what happened in this study.

The bottom line is that there is no valid reason for HFCS to be any different than sucrose in the way that it affects your body.  They are both nearly identical in their composition, containing roughly half fructose and half glucose.  They are both nearly identical in the way they are metabolized by your body.  There is no practical difference between the two as far as your body is concerned.  Now, I’m not saying that you should go out and consume all the HFCS that you want.  The point is that there is nothing uniquely “bad” about HFCS compared to regular sugar.  HFCS is not uniquely responsible for weight gain as some people would have you believe.

If you see a product with HFCS and a similar product with natural table sugar, don’t assume the product with natural sugar is any better.  Rather than worrying about whether something contains HFCS, you should strive to reduce your intake of all types of added sugar and refined carbohydrates in your diet.  It is much more important to look at the big picture;  keep your physical activity high, manage your overall food intake, make sure most of your food is from minimally refined sources, and keep your protein intake high.  This is what will help you lose weight and keep it off, rather than singling out HFCS in your diet.  Don’t let the fructose fear-mongerers fool you.

REFERENCE:  Bocarsly, M.E., et al.  High-fructose corn syrup causes characteristics of obesity in rats:  Increased body weight, body fat and triglyceride levels.  Pharmacology, Biochemistry and Behavior (published ahead of print; available online February 26, 2010)

  38 Responses to “Is High Fructose Corn Syrup Worse Than Regular Sugar?”

  1. WOW this was very interesting. Nice that someone finally was honest about how research is swayed to the highest bidder. The best part about the article is that it was very easy for the average person to read. Thanks and keep them coming.


  2. James,

    You might also want to review and comment on Dr. Lustig’s commentary about HFCS. His position is contrary to what you state. Since over 400,000 have viewed this video – – it might be worth a critical review on your part.

    On my end, I try to avoid all processed sugars.

    Looking forward to reading your blogs.

    Ken Leebow


    • Hi, Ken,

      Thanks for your comment. Alan Aragon has written a very good critique of Dr. Lustig’s lecture.



      • You label Lustig as a High Fructose alarmist by linking to Aragon’s critique in your lede. The fact is that Lustig is very clear that HFCS and sucrose are metabolically equal. He says so a number of times in his famous lecture and it should be underlined by the fact that he references the work of John Yudkin’s published in 1972 as the precursor to his own work. Yudkin’s laid out the problems with sugar before the industrial adoption of HFCS.

        Because HFCS brought down the cost of sweetening, the fact that it’s adoption tracks with the obesity epidemic and is a sneaky ingredient in so much industrial “food” Lustig talks about HFCS but his critique is of fructose. Unfortunately many observers of Lustig’s lecture seem to have difficulty avoiding conflating HFCS and fructose.

        You do this as well in your final statement:
        “. . . rather than singling out HFCS in your diet. Don’t let the fructose fear-mongerers fool you.”

        You counsel to avoid sucrose and HFCS equally, yet sucrose is half fructose. A little cognitive dissonance is creeping in.


        • HFCS is also nearly half glucose and half fructose. There can be more variation in HFCS than sucrose but its within 2 – 4% So its nearly equivalent to sucrose in proportion.


          • The most common formulation of HFCS is 55% fructose while sucrose is 50%. It seems to me the difference computes to 10%. Besides that some studies indicate that the ratio is often varied by beverage companies so that a higher ratio in favor of fructose over glucose is used. This is a very significant difference, especially over a long period of time.


  3. Great article, James.

    I will echo Julie’s post – I love the way your break the math and science down into something easily digestible for us layfolk. You don’t dumb it down – you make it accessible.

    I have been reading your stuff on the BS Detective for some time now, and will happily follow along with you here.


  4. I’m sorry, but much of what you state about p-values here is incorrect – for example, “The P value tells you the probability that the results are not due to chance” is wrong – and you’re using a p of 0.05 when the study only says that p < 0.05. Their p value could easily have been 0.00001.


    • Hi, Kevin,

      “The P value tells you the probability that the results are not due to chance” is wrong

      The P value is the probability of obtaining a test statistic at least as extreme as the one that was observed, assuming the null hypothesis is true. It tells us the chance that a experiment of a “null effect” gives us such a result. In this case, it tells us the chance that we would see a greater weight gain in the HFCS-fed rats, if HFCS has no such effect. In other words, if I were to do 100 identical experiments, a P value of 0.04 would mean that I would expect to see such a result only 4 times, if in fact there is no HFCS effect.

      My statement was nothing more than putting the above paragraph into laymen’s terms. With a biological experiment such as this, we are trying to determine how probable it is that we would see these results, if there is no biological effect. A layman would consider that a “chance” effect. Technically, I understand what you are saying, and my statement was over-simplified on purely technical grounds, but remember that these articles are meant to be understood by someone with no background in statistics. And in laymen’s terms, statistics are all about trying to get an idea of how likely an experimental result is random.

      and you’re using a p of 0.05 when the study only says that p < 0.05. Their p value could easily have been 0.00001.

      When we’re talking about the “decision point” to consider something significant, and we’re dealing with multiple comparisons, then this decision point needs to be adjusted downward according to how many comparisons are being made. Otherwise, the probability of a false positive increases. With the number of comparisons performed in this experiment, the decision point is no longer at the 0.05 level. The researchers did not control for the family-wise error rate. Even if the P value was 0.00001 (which it wasn’t….more on this later), the point is that it’s a P value that hasn’t been adjusted for the fact that multiple comparisons were made. So statements such as “P < 0.05" are meaningless because those P values have not been adjusted to control for the family error rate.

      Second, the P value isn't 0.00001, because throughout the paper, the researchers made statements of "P < 0.01" for P values smaller than 0.01. So it's obvious that the "P < 0.05" values were between 0.01 and 0.05.

      Third, my article was to explain how much easier it is to see a significant result when doing unadjusted multiple comparisons. The family error rate is the probability of making one or more false discoveries or Type I errors. With this HFCS experiment, we know that the Type I error rate is no longer is much greater than that, which means, in laymen's terms, we are much more likely to take a chance result and consider it real.


  • Good article. For further reasons why this study shows crap, read Alan Aragon’s april research review for an article about rats and how their carbohydrate metabolism differs from humans.


    • Thanks for the comment, Josh. Yes, there are distinct differences between rodents and humans in regards to fat & carbohydrate metabolism, which means they aren’t always the best models for looking at these things. Glad to hear that Alan covered this in his research review.


  • “Since there are 6 comparisons, there is a 1 in 4 chance that the results are wrong (and ironically, we have 1 out of the 4 groups showing a significant result)”

    You should learn the meaning of the word “irony” before using it. Right now, even Alanis Morissette could teach you a thing or two about the word “irony.” There can be no irony when the outcome EXACTLY meets your expectations, as is the case with this study. I’ll give you an example of proper usage:

    “It is ironic that the biggest defenders of HFCS (James Krieger and Alan Aragon) also acknowledge that a healthy diet should contain spare amounts of this sweetener.”

    But then again, in light of how easy it is for bloggers to become addicted to attention-seeking, your defending HFCS might not be so ironic after all. So here’s two more examples:

    “It is ironic that people like Krieger are incredibly reluctant to accept any claim that is not backed up by by a study but are quick to jump to assumptions about what undisclosed p-values might have been based on ludicrous behavioral inference on what researchers ‘probably’ did.” (See reply to Kevin)

    “It is ironic that Krieger will nitpick on statistical significance when the studies he holds in high regard are mostly done on statistically insignificant, non-random samples that do not even encompass all human ethnicities.”

    You and Aragon are two of my favorite resources for learning (the other two being lyle and Berkhan). But I think you guys get way too bogged down in being e-celebs and miss the big picture: HFCS is a SHIT food and telling people “hey, we have no reason to believe it’s actually bad but btw dont eat it” sends a moronic message, fails to accomplish your objective of appearing unbiased, and does a disservice to everyone trying to learn about nutrition. Spend less energy trying to topple other e-celebs and more energy posting awesome shit about insulin, please.


    • @vic

      No, I’d say the big picture is being logical. The results in this paper clearly do not indicate HFCS as being worse than sucrose. You’ve provided nothing of substance against that claim, only commenting on irrelevant diction.

      I haven’t reviewed the literature well on this topic…James, aren’t there any papers with well controlled methods that show some downside of HFCS vs sucrose?


    • The way I have interpreted his article is that he is saying that it is no worse for you than sugar as it is pretty much identical in the way it is processed and it’s molecular structure.

      He is also emphasising that one shouldn’t eat it in abundance, but then again, I don’t think anyone would advise you to eat sugar in abundance, so in this respect he is treating them as being equal.

      I’ve no idea if Mr Krieger and Mr Aragon have some ulterior motive for publicly branding HFCS as being of equal value to sucrose, be it e-celebrity status as you point out. However they may just be getting annoyed that it is being used as a media scapegoat for the current health issues across the globe. Thus solely focusing on this as one of the main reasons we are in this state discredits the rest of the work they do on what they believe are more pressing issues, for example calories in > calories out or our ever increasing sedentary lifestyles.


  • its clear so damn clear that that chemicals not only destroy the brain and body but cause cancer and so many other diseases high fructose corn syrup is a chemical people, not natural the ingredient they use to make it has a (skull and bone sign) watch king corn and do some research on it


  • @Daniel … I watched “King Corn,” it was an interesting documentary. I didn’t get any evidence from it that HFCS was bad or unhealthy, simply that it is in a lot more foods than we think. It is a valid point that people wanting to watch their intake of sugars will have to read labels very carefully, but that’s always been the case. “Chemicals destroy the brain and body but cause cancer” … you’d have to state which chemicals specifically, and how they destroy what. Note there are chemicals that we use to treat diseases and help people with brain dysfunctions as well … or do you consider aspirin a killer? It’s a chemical, you know.

    The POINT, people is that we can’t blame any one thing for obesity or ill health in our country. I’m not sure why everyone is demonizing HFCS, but then again, people used to demonize sugar (“natural”) as well, and fat. It’s just the latest scapegoat. None of this helps. It’s your entire diet and exercise lifestyle and that contributes to health, as well as your genetics … not one single “evil” substance.


  • Good old Fisher developed the use of P values for statistical inference, mainly for agricultural research. It is said that he himself supported the the notion of adopting a value of P=2.5% as providing an inference of “not due to chance”.

    In my own personal decision making I pay no attention to results which do not reach a level of P=1% – but then I also ignore results have Risk Ratios less than 500% or Numbers to Treat less than 15.

    I have a view that I should be persuaded that my current practice should change – so I prefer to focus on (1-P).

    Thanks, James, for the presentation.

    The questions that occur to me as a result of your post are:

    - other than “lacing the water with sweeteners” were the animals supplied with same chow and did the eat “ad libitum”

    - did the controls in both studies show “rat healthy” levels of fat for their age, 8 weeks and 26 weeks.

    However, taking both results together, one gets an impression that rats, too, can get fat by providing sweetened water!

    So what’s new, Princeton?


  • Hi James,

    The study below indicates a possible reason why HFCS might be more fattening – some of it may have much higher levels of carbohydrate (thus calories) than labels claim.

    Dr. Ray Peat sent this to me.

    What do you think?

    * FASEB Journal 2010 PN Wahjudi

    Carbohydrate Analysis of High Fructose Corn Syrup (HFCS) Containing Commercial Beverages
    Paulin Nadi Wahjudi1, Emmelyn Hsieh1, Mary E Patterson2, Catherine S Mao2 and WN Paul Lee1,2 1 Los Angeles Biomedical Research Institute, Torrance, CA 2 Pediatric, Harbor-UCLA Medical Center, Torrance, CA

    The carbohydrate analysis of HFCS is based on methods which first hydrolyze the syrup into simple sugars before quantitative analysis. We have examined whether HFCS can be hydrolyzed under the same conditions suitable for hydrolyzing sucrose. A new GC/MS method for the quantitation of fructose and glucose as their methoxyamine derivatives and 13C labeled recovery standards was used to determine the carbohydrate content of HFCS in 10 commercial beverages. Samples were analyzed before and after acid hydrolysis. The carbohydrate contents in commercial beverages determined without acid hydrolysis were in agreement with the carbohydrate contents provided on the food labels.

    However, the carbohydrate contents of beverages determined after acid hydrolysis were substantially (4–5 fold) higher than the listed values of carbohydrates. As fructose and glucose in HFCS may exist as monosaccharides, disaccharides and/or oligosaccharides, analysis of the carbohydrate content of HFCS containing samples may yield widely different results depending on the degree of hydrolysis of the oligosaccharides. With inclusion of mild acid hydrolysis, all samples showed significantly higher fructose and glucose content than the listed values of carbohydrates on the nutrition labels. The underestimation of carbohydrate content in beverages may be a contributing factor in the development of obesity in children.


    • Lillea,

      You could be on to something here. Thanks for bringing this to my attention. I’ll try to review this article in a future issue of Weightology Weekly.


      • James and Lillea,

        The full study is not yet available, so making any definitive comment on its validity is not possible. However, I think it is important to recognize that this is simply an abstract at this point, has not been published in a peer –review journal, and therefore has not passed the scrutiny of trained scientists, with no vested interest in the subject, whose views would be needed for an objective assessment of the adequacy of the experimental design and the validity of the conclusions.

        Manufacturers of HFCS do not use acid hydrolysis and gas chromatography-mass spectrometry (GC/MS in the analysis of sweeteners. That procedure is too harsh for delicate carbohydrate molecules and has a very good chance of hydrolyzing the sugars into smaller, non-sugar pieces and creating artifacts that would be inappropriately tallied as sugars. Instead, the gentler HPLC method has been successfully tested, validated by AOAC, and used for several decades. The abstract does not indicate whether the method the authors used was properly validated by AOAC.

        Therese, Corn Refiners Association


  • I appreciate your being the Devil’s advocate regarding Dr.Lutig’s “Sugar: The Bitter Truth” above. I’ve always been an advocate of Mark Twain’s “Lies, Damned Lies and Statistics”… Your final comments are key to your rebuttal, i.e. ” It is much more important to look at the big picture; keep your physical activity high, manage your overall food intake, make sure most of your food is from minimally refined sources…” How do we convince people to do that?


    • Thanks, rkeinc. I’m not sure how we convince people to do that. I think it’s human nature to want to point our fingers at one thing thing that is causing obesity. People want a scapegoat, and in this case it’s HFCS. It’s tough to fight human nature and I think we’re always going to deal with the problem of people losing the forest for the trees.


  • I think everyone has forgotten that fructose is MUCH sweeter than glucose or sucrose. So any mixture with free – and an excess – of fructose is more likely to be addictive (sugar sure seems to act in the same centres as cocaine and nicotine doesn’t it? Is that glucose or fructose?) in well formulated food and drink. When HFCS was developed, the big food companies did not do an equivalent sugar substitution based on sweetness did they? They INCREASED the dose (because it was cheaper) because it increases addiction – and sales. Sweetness is perceived in the mouth. So if fructose is present as it will be in the HYDROLYSED mixtures then it hits the mouth and we get pleasure plus it hides a lot of other disgusting things. Sugar has a natural sweetness, but hydrolyses in the stomach. No sensors there for sweetness huh?

    So we go on this upward competitive spiral with the low-fat myth (careful here but John Yudkin DID predict all this in 1972! in his book “Pure, White and Deadly – go read it) where food companies use these cheap ingredients with artificial flavours to produce (very well formulated!) cheap and addictive foods. Even rats can get addicted to sweetness surely? Another factor in food formulation is “mouth feel” and the total combination of sugar, salt, fat/low fat with “mouth feel and texture” is what makes this crap so nice to eat – and then (probably) kills us. Now I have been fat in my adult life and I have tried counting calories and exercise (and I ate back ALL the calories from exercise). It worked well and within 9 months I had lost 25 kilos. Pretty good but it came roaring back on when I stopped counting. Why? Because in counting calories I included all my favourite orange and apple juice drinks and I became ADDICTED to them. In the end my calorie total were these and really fast carbs and the protein-fat content went really low. My cyclic appetite drive was enormous (insulin cycle driven by sugar?) and as soon as I tried to readjust I rocketed back up. Devastated. So two years on I heard Lustig and simply cut out sugar from my food and ALL foods with added sugar (however it is described). Result? I lost it all again AND I HAVE STAYED THERE. NO HUNGER PANGS AT ALL. This is not Atkins (that failed miserably too), though it looked a bit Harcombe to start with. It is just no sugar. At all. No processed food with added sugar. I cook everything myself from scratch.

    My liver is also doing quite well as is no longer very fatty. Is this because I no longer get exposed to much fructose and the fat is getting metabolised at last (fructose IS lipogenic as far as the liver is concerned isn’t it?). I take all your points on the statistics and experimental traps that we can fall into adn they are good ones. Lustig has to be careful in his arguments here or he could lose credibility. But for me the core thing is that – if we live in ignorance – then the intelligent thing to do is the obvious one. I have worked in the flavour industry, so I know to what extent manufacturer’s go to to maximise profit and develop addictive taste (especially sweeteners) over nutrition. So in the absence of the true Randomised Control Test I will happily go on belief and experience. I gave up sugar and got thin – dramatically. So I will readily BELIEVE fructose is a chronic long term poison. We’ll see anyway because my proportion of fat and protein has gone up significantly though I balance it out with a good range of vegetables for slow carbs. We all die of something and maybe I have already done too much damage to my system and the oxidised LDLs will see me off. Death awaits us all and we will become statistical points in an epidemiological study. I wonder if this argument will still be going on then?

    Good discussion though – let’s keep it up. Control that/those independent variable(s)!


  • James,

    I am not familiar with the statistics terms, so this comment is based solely on the information in your article. based on my understanding from your article, the pair wise-calculated p-value of ~0.25 in dictates that one of the four sample groups could have happened by chance. is it possible that the 24-hour group in Experiment 1 was low by chance and that the 12-hour group was correct? If so, then wouldn’t the results be inconclusive of the presence or absence of effect from HFCS?


    • Hi, Bob,

      When doing statistics such as these, you are always testing the null hypothesis. In this case, the null hypothesis is that there is no difference between any of the groups. The alternative hypothesis is that a difference exists. You need sufficient evidence to reject the null hypothesis and accept the alternative. A significant P value means you have sufficient evidence to reject the null hypothesis. However, if that P value is significant because you didn’t adjust for multiple comparisons, then there is a good chance that you falsely rejected the null hypothesis. In other words, it’s a false positive. Since it’s the 12-hour group being reported to be significantly higher than the others, it is the 12-hour group that would be considered the false positive.


  • James,
    you said:
    ” If the rats fed HFCS for 12 hours gained more weight, why didn’t the rats fed HFCS for 24 hours also gain more weight? They got HFCS for a full 12 hours more, yet didn’t gain more weight.”

    this is how they explain it in the full text:

    “We selected these schedules to allow comparison of intermittent and continuous access, as our previous publications show limited (12-h) access to sucrose precipitates binge-eating behavior”

    However it does not make sense anyway, as their own long-term study shows the opposite


  • I tend to agree that there is no single culprit — and by the same token, no single cure — for obesity. Two generations ago children grew up on sweetened cereal, but were able to stay slimmer than their counterparts today.

    I’ve lost 50 pounds and maintained the loss on a high-carb diet (with most of my carbs in the form of whole grains), while people all around me are demonizing carbs. Carbs make me feel uniquely satisfied, both physically and psychologically, so I see no reason to give them up if what I’m doing is working.


  •  Leave a Reply



    You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>