Liquid-only diet: good or bad?

[FrederickSebastian]

Ok... so I have a kind-of weird question...

 

Has any research been done concerning a liquid only diet (like putting fruits and vegetables in a blender then drinking the juice)... I feel like it may allow you to live longer because your organs won't get worn out as much processing solid food... But then again, maybe by not processing solid food, the organs would grow weaker because they're not used as much... What do you think?

[Dean Pomerleau]

I used to eat most of my (strict vegan) meals in smoothie form like you suggest, for convenience. But I switched back to eating meals in whole form for aesthetic reasons (more enjoyable), and to slow digestion and therefore reduce glucose spikes after meals.

 

The human digestive system is amazingly efficient, and I don't know of evidence showing that it can get "worn out" over time eating whole foods if properly cared for.

 

Dean

[Saul]

I agree totally with Dean.  When you eat a "juiced" vegetable, you're not getting the benefit of the fiber -- this slows the passage of food through your digestive tract.  Also, a "juiced" vegetable allows easy absorption of starch in your vegetable -- the starch might have been in the form of resistant starch, some of which will not be digested -- instead, you'll be absorbing all of the calories in the starch, with a high glycemic load.  Finally, your teeth are kept healthy, by crunching on your vegetables; they get no benefit when you drink a juice.

 

For example, a raw carrot is a crunchy, CR-friendly food with high fiber, as well as the many vitamins and carotenoids in the vegetable, with a low glycemic index.  Carrot juice is equivalent  to sugar water, fortified with vitamins.

 

  -- Saul

[Dean Pomerleau]

Saul is right that actual juicing removes the fiber and makes a liquid diet even worse for your health, although perhaps still better than a standard American diet. But that wasn't what FrederickSebastian was asking about, which was turning fruits & veggies into a smoothie in the blender.

 

In short, when it comes to health, whole foods > blending > juicing except perhaps for people with compromised digestive systems. Plus smoothies and juices are not as satiating, making it more difficult to practice CR.

 

--Dean

[Dean Pomerleau]

Dr. Greger, everyone's favorite nutrition expert :-), just finished a four-part video series on the healthiness of smoothies:

 

http://nutritionfacts.org/topics/smoothies/

 

The scientific evidence appears to show that fiber-preserving smoothies (as opposed to juices) allow for better phytonutrient absorption and do not negatively impact glucose metabolism or insulin levels when compared with eating the fruits / vegetables in whole form, particularly if the smoothie is consumed slowly.

 

As a result of these insights, I've started adding a good fraction of my daily leafy greens to my homemade blended salad dressing, rather than eating them all whole.

 

Dean

[Michael R]

I don't have much patience for watching online videos (and haven't had a TV in my house that wasn't packed away in a box in >10 y, or sat through an entire film in > 2 y, and the last one before that was in 2006), and Dr. Greger is decidedly not my favorite nutritionist: I see on the page you link that he has a video subtitled "Ranking foods by antioxidants per dollar spent," a mere minute of which confirms that he's dividing test-tube-nonsense Trolox-equivalents by dollars per pound, which is nutritional voodoo (similar to a video of his I did watch years ago, in which he ranked vegetables as anti-cancer foods based on in vitro effects on cancer cells in culture). And then there's his incredibly misleading representation of scientific studies on olive oil ...
 
Anyway, I hope you'll forgive me if I don't sit through a four-part series on the subject that Greger has assembled (and then double-check his sources).
 
My own review of the evidence leads me to the conclusion that turning fruits and vegetables into blended smoothies does worsen glycemic response, albeit to a lesser degree than juicing per se. The mechanical disruption of the food matrix both sunders the fiber molecules from the glycemic carbohydrates, speeding absorption, and also (surprisingly) breaks down the large, resistant carbohydrate molecules into smaller, more readily-digestible molecules.
 
See eg. this classic study of apples vs. isocaloric apple juice and purée (which last, of course, retains all the fiber in the original fruit, unlike the juice), finding that the purée increased the insulin spike and led to a faster trough in blood glucose:
 

(From (1); Figure reproduced in (2). Larger version here). They even forced subjects to slow down their eating of the purée so as to control for the faster entry of food into the stomach.
 
Similarly, though less dramatically, (3) compared the glycemic and appetitive responses to intact peaches consumed with 750 mL water (solid lines) to the same amount of peaches blended into 500 mL water with 250 mL water on the side (again totaling 750 mL — dashed line):
 
(Grr... can't post it ... full figure here). Lots of other studies have come to the same conclusion on puréed/blended fruits and veg in stimulating more hunger at a subsequent meal,(4-6) even when in (6) the carrots and carrot purée were given as part of a more complete meal. And most but not all studies find that simply adding reasonable doses of fiber to liquid foods has little blunting effect on glycemia; hence David Ludwig's quip that OJ plus Metamucil is not glycemically equivalent to an orange with equivalent fiber content.
 
References
1: Haber GB, Heaton KW, Murphy D, Burroughs LF. Depletion and disruption of dietary fibre. Effects on satiety, plasma-glucose, and serum-insulin. Lancet. 1977 Oct 1;2(8040):679-82. PubMed PMID: 71495.
 
2: Stubbs RJ, Mazlan N, Whybrow S. Carbohydrates, appetite and feeding behavior in humans. J Nutr. 2001 Oct;131(10):2775S-2781S. Review. PubMed PMID: 11584105.
 
3: Martens MJ, Lemmens SG, Born JM, Westerterp-Plantenga MS. Satiating capacity and post-prandial relationships between appetite parameters and gut-peptide concentrations with solid and liquefied carbohydrate. PLoS One. 2012;7(7):e42110. doi: 10.1371/journal.pone.0042110. Epub 2012 Jul 26. PubMed PMID: 22844543; PubMed Central PMCID: PMC3406044.
 
4: Flood-Obbagy JE, Rolls BJ. The effect of fruit in different forms on energy intake and satiety at a meal. Appetite. 2009 Apr;52(2):416-22. doi: 10.1016/j.appet.2008.12.001. Epub 2008 Dec 6. PubMed PMID: 19110020; PubMed Central PMCID: PMC2664987.
 
5: Mattes RD, Campbell WW. Effects of food form and timing of ingestion on appetite and energy intake in lean young adults and in young adults with obesity. J Am Diet Assoc. 2009 Mar;109(3):430-7. doi: 10.1016/j.jada.2008.11.031. PubMed PMID: 19248858; PubMed Central PMCID: PMC2680008.
 
6: Anne Moorhead S, Welch RW, Barbara M, Livingstone E, McCourt M, Burns AA, Dunne A. The effects of the fibre content and physical structure of carrots on satiety and subsequent intakes when eaten as part of a mixed meal. Br J Nutr. 2006 Sep;96(3):587-95. PubMed PMID: 16925866.

[Dean Pomerleau]

Thanks Michael,

 

Dr. Greger spends much of the short video (only 4 min - are you really that busy? There is also a "view transcript" button next to the video if you're rather read it than watch it) on the apple study you mention, and concludes the same thing you did - eating whole apples is better for glycemic control than a pureed apple smoothie, even when drunk slowly. He goes on to suggest adding berries to smoothies based on evidence that phytochemicals in the pureed berries improve the glycemic response.

 

Speaking of double checking sources :), regarding the study of pureed vs. whole peaches [3] you reference:

 

 

Similarly, though less dramatically, (3) compared the glycemic and appetitive responses to intact peaches consumed with 750 mL water (solid lines) to the same amount of peaches blended into 500 mL water with 250 mL water on the side (again totaling 750 mL — dashed line):
 
(Grr... can't post it ... full figure here). 

 

Here is the figure you were trying to include (click to enlarge):

 

 

As you (and the caption) say, the solid lines in the graphs represent the 'intact peach + water' condition and the dashed lines represent the 'pureed peaches + water' condition. It appears to me that the dashed-line peach smoothie results in significantly greater post-meal satiety (upper left graph), lower glucose spike (although not significant, upper right graph), and significantly smaller insulin spike (middle right graph) than the whole peaches.

 

Don't these results this contradict your claim that eating fruit in smoothie form is worse than eating the whole fruit in terms of satiety and peak glucose/insulin response due to fiber disruption? Am I missing something?

 

But the real reason that one might want to blend fruits and vegetables (beyond convenience / palatability) is to improve bioavailability of micronutrients, as addressed in this video in Dr. Greger's series. In it, he references [1], which found folate bioavailability from spinach was significantly higher after mincing or blending than when subjects consumed spinach in whole leaf form.  He also discusses [2], which found the bioavailability of a range of carotenoids in carrots was enhanced by both cooking and blending, and the effects of the two food preparation methods were additive. He also discusses [4], which found breaking down cell walls via pureeing improved carotenoid availability for both carrots and tomato products.

 

To be clear, I'm not recommending consuming all of one's fruits and vegetables in smoothie form. I continue to eat most of mine in whole, raw form. But I have started adding some of my greens into my blended salad dressing rather than eating all my greens whole to improve phytonutrient bioavailability. The other salad dressing ingredients are tomatoes, oranges, cucumbers, mixed berries and spices, blended in my Vitamix for 30 second until pretty smooth.

 

--Dean

 

----------------

 

[1] Ann Nutr Metab. 2000;44(4):163-9.

Bioavailability of folate from processed spinach in humans. Effect of food matrix
and interaction with carotenoids.

Castenmiller JJ(1), van de Poll CJ, West CE, Brouwer IA, Thomas CM, van
Dusseldorp M.

Author information:
(1)Division of Human Nutrition and Epidemiology, Wageningen University,
Wageningen, The Netherlands. jacqueline.castenmiller@staff.nutepi.wau.nl

The effect of the food matrix and dietary fibre on the bioavailability of folate
is not known. In a controlled, 3-week dietary intervention study, 28 men and 42
women were divided into six groups to receive either a control diet (n = 10), or
the control diet plus 20 g/MJ per day (n = 12 per group) of whole-leaf spinach,
minced spinach, liquefied spinach, or liquefied spinach to which dietary fibre
(10 g/kg wet weight) was added. The sixth group received the control diet plus a
synthetic carotenoid supplement with similar amounts of beta-carotene and lutein
as found in spinach. A significantly higher plasma folate response was found for
the pooled spinach groups than for the control group. Among the spinach groups no
significant differences were detected. However, the plasma folate response of the
pooled minced and liquefied spinach groups was greater than that of the
whole-leaf spinach group (p = 0.03). Re-addition of dietary fibre to the
liquefied spinach to compensate for the fibre broken down during liquefaction did
not reduce the plasma folate response. The consumption of the carotenoid
supplement did not have an effect on plasma folate concentrations compared with
the control group. The food matrix in which the folate is entrapped plays a role
in folate bioavailability.

PMID: 11111131

 

-------------------

[2] J Agric Food Chem. 2010 Dec 22;58(24):12769-76. doi: 10.1021/jf102554h. Epub 2010

Dec 1.

Particle size reduction leading to cell wall rupture is more important for the
β-carotene bioaccessibility of raw compared to thermally processed carrots.

Lemmens L(1), Van Buggenhout S, Van Loey AM, Hendrickx ME.

Author information:
(1)Laboratory of Food Technology and Leuven Food Science and Nutrition Research
Centre (LFoRCe), Department of Microbial and Molecular Systems (M2S), Katholieke
Universiteit Leuven, Kasteelpark Arenberg 22, 3001 Leuven, Belgium.

The amount of nutrients that can be released from food products (i.e., nutrient
in vitro bioaccessibility) is often studied as it is a starting point for
investigating nutrient bioavailability, an indicator for the nutritional value of
food products. However, the importance of mastication as a particle size
reduction technique is poorly understood and is often neglected during in vitro
procedures determining bioaccessibility. Therefore, the aim of the present work
was to study the effect of mechanical breakdown on the β-carotene
bioaccessibility of carrot samples, having different textural/structural
characteristics (as a result of thermal processing). In the first part of this
study, the all-E-β-carotene bioaccessibility of carrot particles of different
sizes (ranging from cell fragments up to large cell clusters), generated from raw
as well as from gently and intensely cooked carrot samples, was determined. In
the second part of the study, the effect of human mastication on the particle
size reduction of raw as well as of gently and intensely cooked carrot samples
was investigated in order to allow identification and validation of a technique
that could mimic mastication during in vitro procedures. Results showed a strong
dependency of the all-E-β-carotene bioaccessibility on the particle size for raw
and gently cooked carrots. After intense cooking, on the other hand, a
considerable amount of all-E-β-carotene could be released from cell fragments
(smaller than a cell) as well as from small and large cell clusters. Hence, the
importance of mechanical breakdown, and thus also of (in vitro) mastication, is
dependent on the carrot sample that is considered (i.e., the extent to which the
carrot sample has been thermally processed prior to the particle size reduction).
Structural changes occurring during mechanical and thermal processing are hereby
key factors determining the all-E-β-carotene bioaccessibility. The average
particle size distribution curves of raw and cooked carrots, which were chewed by
15 persons, could be mimicked by mixing 50 g of carrots using a Grindomix
(Retsch) at 2500 rpm during 5 s. Based on this scientific knowledge, the
identified in vitro mastication technique was successfully integrated in the in
vitro digestion procedure determining the all-E-β-carotene bioaccessibility of
carrot samples.

PMID: 21121612

 

-------------------------

[3] 3: Martens MJ, Lemmens SG, Born JM, Westerterp-Plantenga MS. Satiating capacity and post-prandial relationships between appetite parameters and gut-peptide concentrations with solid and liquefied carbohydrate. PLoS One. 2012;7(7):e42110. doi: 10.1371/journal.pone.0042110. Epub 2012 Jul 26. PubMed PMID: 22844543; PubMed Central PMCID: PMC3406044.

 

----------------------

[4] J Agric Food Chem. 2012 Dec 5;60(48):11995-2003. doi: 10.1021/jf303502h. Epub

2012 Nov 27.

Relation between particle size and carotenoid bioaccessibility in carrot- and
tomato-derived suspensions.

Moelants KR(1), Lemmens L, Vandebroeck M, Van Buggenhout S, Van Loey AM,
Hendrickx ME.

To study the effect of particle size on the relative all-E-β-carotene and
all-E-lycopene bioaccessibility in carrot- and tomato-derived suspensions,
respectively, an in vitro digestion approach including oil was used. Adding olive
oil (2%) during digestion, especially as an oil-in-water emulsion, resulted in a
substantial increase in carotenoid uptake in the micellar phase. Carotenoid
bioaccessibility decreased with average particle size. Only particles smaller
than an individual cell resulted in high bioaccessibility values, pointing out
the importance of the cell wall as the main barrier for carotenoid uptake. The
relation obtained between particle size and bioaccessibility was used to predict
the carotenoid bioaccessibility in carrot- and tomato-derived purées. These
predictions indicated that carotenoid bioaccessibility in plant-based food
suspensions is not only determined by the cell wall integrity (related with
particle size) but is also affected by interactions between the structural
compounds of the complex food matrix.

PMID: 23157717

[Michael R]

Dean, and all:
 

Dr. Greger spends much of the short video (only 4 min - are you really that busy?

You originally directed us to a "four-part video series," not a single 4-minute video, and the link you used was actually to a page full of a dozen or so Greger videos; I wasn't going to sit through that!
 

on the apple study you mention, and concludes the same thing you did - eating whole apples is better for glycemic control than a pureed apple smoothie, even when drunk slowly. He goes on to suggest adding berries to smoothies based on evidence that phytochemicals in the pureed berries improve the glycemic response.

If you look at the studies that he cites (in the sources given in the relevant video to which you helpfully point), they used an awful lot of berries to get these modest effects: in his (7), they used 150 g of them that's 95 Cal of blackcurrants, and 238 Cal of lingonberries! And in his (8), they weren't actually even using berries, but a berry extract that "equated to an amount of bioactives in ≈2 cups of fresh whole blueberries"!

And in those studies they were testing its effects against pure sucrose (7) or a hyperinsulinemic-euglycemic clamp (8): in my (9) below, from the same authors and with the same berries (and others) included, these same berries in the same amounts failed to blunt the glycemic response to whole-grain bread (although strawberries or a mixture were effective). Similarly, in (10), adding berries to yogurt counteracted the glycemic rise from the added sugar in a yogurt, but failed to impact the inherently lower glycemic response to a yogurt with an isocaloric addition of fat instead of sugar.

The obvious thing, tho', is to eat the intact food, and add on berries as an added bonus (or skip the extra Calories).
 

Speaking of double checking sources :), regarding the study of pureed vs. whole peaches [3] you reference [MR's (3):] ... It appears to me that the dashed-line peach smoothie results in significantly greater post-meal satiety (upper left graph), lower glucose spike (although not significant, upper right graph), and significantly smaller insulin spike (middle right graph) than the whole peaches.

Ah. Well, I have to shame-facedly admit to having unconsciously mistaken the effects of the one for the other — perhaps even under the influence of my own expectations from other studies. That is surprising, and counter to other findings. I'm not sure what explains that (the fact that peaches, if they're ripe, are much softer (if ripe), have a higher GI to begin with (56 vs 40), and/or the very large volume of extra water (750 mL/≈3 C)?), but I'd say it's an exception rather than the rule — and either way, not nearly so large as the effect on apples (without extra water) in my (1).
 

Don't these results this contradict your claim that eating fruit in smoothie form is worse than eating the whole fruit in terms of satiety and peak glucose/insulin response due to fiber disruption? Am I missing something?

Well, you've not taking on board the apple study, and the other references I gave: the peach study is an exception, and a minor one granted the relative magnitide of effect (or the non-effect vs. an effect in the other studies in the case of satiety).
 

But the real reason that one might want to blend fruits and vegetables (beyond convenience / palatability) is to improve bioavailability of micronutrients, as addressed in this video in Dr. Greger's series.

See, now you're already asking me to watch more Greger videos ;) . Broadly, I would point out that while the effect on these micronutrients (carotenoids and folate) might be relevant to people with relatively low-vegetable and/or very low-fat diets, it's hardly a concern to people on typical CR diets, which are utterly loaded with these micronutrients and in most cases contain a reasonable amount of fat. Indeed, your/his (3) higlights that "Adding olive oil (2%) during digestion, especially as an oil-in-water emulsion, resulted in a substantial increase in carotenoid uptake in the micellar phase." I would also point out that your/his (2,3) are studying in vitro release under simulated GI tract chemical conditions, not actual bioavailability when humans consume these foods under alternating conditions.

By contrast, (11) was an in vivo study and found no difference in chronic serum carotenoid levels between carrot juice and raw carrots:
 

"38 healthy, non-smoking female volunteers, aged 20–53 years ... were instructed to consume low-carotenoid diets for 10 days before starting the [...] study. ... The participants were assigned into three groups with similar mean serum β-carotene concentrations. Each group was supplemented for 6 weeks with approximately 12 mg of β-carotene/day, either as raw carrots (120 g/day), carrot juice (1 dl/day) or β-carotene capsules (2 capsules/day). ... [A]fter 3 and 6 weeks of supplementation[,] The mean serum β-carotene concentration increased significantly in all test groups ... No difference in the serum response between carrot and carrot juice was detected."(11)

If there were some amazing benefit to smoothie-drinking, it might be worthwhile to take extra steps to counteract the glycemic downside; as it is, this seems a bit like diabetics "covering" the extra carb in a meal with more insulin.

The semi-exception is lycopene, and even there the effects of cooking and added fat are reported to be more pronounced than those of mechanical disruption absent same, and  it's pretty clear from the epidemiology that the specific food sources most associated with disease prevention are cooked, fat-containing tomato products, leading to the surprising protective effects of a nutrient primarily sourced from foods that are generally linked to less-healthy diets: "80% of dietary lycopene intake in the U.S. is derived from processed tomato products such as ketchup, tomato juice, spaghetti sauce, and pizza sauce (Clinton 1998)." Hence the headlines that "A Slice of Pizza Cuts Prostate Cancer Risk" and "Reduce the risk of prostate cancer by eating pizza, ketchup,  and baked beans." I quite intentionally consume cooked and mechanically-disrupted tomato products for this reason, and with added fat to boot — but they're consumed as part of a much larger meal of otherwise-intact foods (aside from the fact that I do "of course" ;) chop most of my vegetables).

References
7: Törrönen R, Kolehmainen M, Sarkkinen E, Mykkänen H, Niskanen L. Postprandial glucose, insulin, and free fatty acid responses to sucrose consumed with blackcurrants and lingonberries in healthy women. Am J Clin Nutr. 2012 Sep;96(3):527-33. doi: 10.3945/ajcn.112.042184. Epub 2012 Aug 1. PubMed PMID: 22854401.

8: Stull AJ, Cash KC, Johnson WD, Champagne CM, Cefalu WT. Bioactives in blueberries improve insulin sensitivity in obese, insulin-resistant men and women. J Nutr. 2010 Oct;140(10):1764-8. doi: 10.3945/jn.110.125336. Epub 2010 Aug 19. PubMed PMID: 20724487; PubMed Central PMCID: PMC3139238.

9: Törrönen R, Kolehmainen M, Sarkkinen E, Poutanen K, Mykkänen H, Niskanen L. Berries reduce postprandial insulin responses to wheat and rye breads in healthy women. J Nutr. 2013 Apr;143(4):430-6. doi: 10.3945/jn.112.169771. Epub 2013 Jan 30. PubMed PMID: 23365108.

10: Linderborg KM, Järvinen R, Lehtonen HM, Viitanen M, Kallio HP. The fiber and/or polyphenols present in lingonberries null the glycemic effect of the sugars present in the berries when consumed together with added glucose in healthy human volunteers. Nutr Res. 2012 Jul;32(7):471-8. doi: 10.1016/j.nutres.2012.06.004. Epub 2012 Jul 18. PubMed PMID: 22901554.

11: Törrönen, R., Lehmusaho, M., Häkkinen, S., Hänninen, O. & Mykkänen. Serum β-carotene response to supplementation with raw carrots, carrot juice of purified β-carotene in healthy non-smoking women. Nutr. Res. 1996 Apr;16(4):565-575.

[Dean Pomerleau]

Thanks Michael,

 

 

I quite intentionally consume cooked and mechanically-disrupted tomato products for this reason, and with added fat to boot — but they're consumed as part of a much larger meal of otherwise-intact foods (aside from the fact that I do "of course" ;) chop most of my vegetables).

 

As do I. All I was trying to do was provide evidence that there are some processing methods that are better than others (i.e. pureeing > juicing) for people who are already inclined for one reason or another to prefer drinking their fruits and veggies rather than consume them in whole form.

 

--Dean