Since the Fourth of July, we’ve been on an ice cream making tear, a fact which probably hasn’t escaped the notice of regular readers of this blog. I’ve heard from some of you who have tried your hand at making low-carb ice cream from the blog recipe and most of you reported that the taste and texture of these low carb creams are pretty darned good. And I agree. Their one deficiency is in what’s called ‘mouth feel.’

Mouth feel, a sort of ineffible quality of a food, has to do with how it sits in your mouth and how it dissolves and changes. For instance, does it break apart smoothly? Does it seem gritty? Is it gone too fast or does it linger just long enough? Does it leave an oily sensation or a clean one on the tongue when it’s gone?

It’s in the mouthfeel department that low-carb homemade ice creams, delicious though they are, slightly miss perfection. And the problem stems chiefly from their lack of sugar, which in a traditional ice cream recipe serves functions beyond sweetening the product, an important one of which is to modify the behavior of the fat content.

Something happens to fat when it’s blended with sugar; it loses its oily mouth feel. To easily illustrate this effect, think of what it would feel like to chomp down on big spoon of pure butter, to wallow it around in you mouth, and swallow it. What do you get? An oily mouthfeel.

Now, cream that butter with powdered sugar and what do you get? Frosting…and, like magic, no oily mouthfeel. The same thing happens when you chill and churn cream (or half and half or coconut milk) sweetened with sugar to make that frozen concoction that helps us hang on (apologies to Jimmy Buffet.)

But in a low carb version that uses Splenda or Stevia or some other artificial (non-sugar) sweetener, there’s no magic transformation and the mouth-waterin’ mouthful leaves a decidedly oilier residue lingering on the tongue than we might ideally desire. Not quite like the straight butter example, but not the rich, but clean finish it would have had were sugar in the mix.

That slight imperfection led us to do a little sleuthing, reading, and experimenting, which led us to a handbook on sweetener alternatives and the properties they impart, which led us to try adding some polydextrose powder to the mixture. Afterall, it’s what the big boys use, so why not us home folk?

Polydextrose is a long chain of dextrose molecules (call me mistress of the obvious) and dextrose is nothing more than double glucose–ie, two glucose molecules hooked together. Glucose, of course, is blood sugar. That’s why dextrose is the sugar used in IV solutions. It’s the ‘D” that’s being called for when the doctor shouts out orders to hang a liter of D5W (5% dextrose water) or D5Normal (5% dextrose in salt water) or to push some D50 (50% dextrose solution) on ER or Gray’s Anatomy.

While the human GI tract can easily snap apart a single molecule of dextrose to harvest and absorb the two glucoses, once polymerized into the long polydextrose chains, the tin snips of the digestive process can no longer break the links and so polydextrose behaves as a fiber. Like most soluble fibers it passes intact and unabsorbed through the small intestine, then on down stream to the waiting friendly bacteria (in the colon, mainly) who do have the tools to break it apart and use it or ferment it.

Even though once in the body it doesn’t behave like sugar in raising insulin or blood glucose levels, polydextrose still retains many of sugar’s food tech properties, imparting a slight sweetness (about 1/10 the sweetness of sugar) and taking sugar’s place in recipes by adding bulk to and improving the mouthfeel of the finished product. Like fiber, however, its indigestible nature can cause a little bit of GI rumbling, bloating, and gas if you overdo it. Commercially available polydextrose powder also contains a bit of citric acid and about 10% sorbitol, which probably also contributes to the mild degree of GI agitation that can occur.

You may be able to find polydextrose at your local natural foods grocery store, but if not, you can find it online in reasonable-sized quantities. For a source, click here*.

By educated guessing and some trial and error, I arrived at an amount of polydextrose to add to a batch of ice cream that would achieve the desired mouthfeel without causing much in the way of GI side effects.

And the answer is: 1/3 to 1/2 cup per 1 quart batch.

By quart batch, I refer to the Coconut Milk Ice Cream recipe in the previous blog. Obviously, to make the low carb version, opt for the Splenda instead of the dextrose powder and use your choice of cream and/or half and half for the dairy version or the coconut milk plus rice milk for the non-dairy version. Then, modify the instructions there as follows: add about 1/3 cup polydextrose to the 6 beaten egg yolks (as you would normally have added sugar in a traditional recipe) and continue to beat until thick and pale. Then temper them with the sweetened hot cream (or coconut milk) and proceed with the recipe just as written. If you still perceive a slightly oily mouthfeel, next time up the amount of polydextrose to 1/2 cup. (You’re walking a fine line here between good mouthfeel and too much of the fiber effect, so see what works best for you.)

Last night, the time came to test the hypothesis. Since we’d had my fresh fruit flavor choice (peach) the other day,it was Mike’s turn to choose. For him, the go-to fruit choice for almost anything from savory compotes to cobblers to pastries and most particularly for ice cream is cherry. So bowing to spousal pressures, I cranked up a batch of Cherry Vanilla.

We’d bought a bag of fresh dark sweet cherries at the farmers’ market, which I thanklessly halved and pitted, turning my fingernails maroon in the process. After a course chop, I added about 1 cup of the cherries to the cooled ice cream base, then chilled the whole lot properly before churning. It was, indeed, yummy with just the right mouthfeel when it came freshly churned from the machine.

Ben and Jerry got nothin on me! I think I’ll call it Poly Woly Cherry.

*As Amazon Associates, we earn a small commission on qualified purchases.


  1. Wow, this is a very timely post as I’m getting ready to make a batch of the coconut milk ice cream this weekend and was wondering if I should add some poly-d to it. My standard LC ice cream recipe that I use calls for it and indeed, it does take away from that oily mouth feel. The other benefit is that the ice cream stays scoopable when stored in the freezer rather than turning into a rock-hard lump. My only question was one of how much and you have just provided me with the answer! Yum, I can’t wait to taste it!

    Like Mike, I like all things cherry but much prefer the sour ones over the sweet. I wonder if there can be such a thing as sour cherry ice cream? Just might have to give it a try…

    COMMENT from MD EADES: I’m sure Mike would love sour cherry ice cream, too. Give it a whirl and let us know how it goes.

    As to the scoopability, if your freezer really freezes cold cold cold (as our does) this amount of polydextrose doesn’t seem to really keep it soft and scoopable. I still end up sticking it in the microwave for 15 or 20 seconds. Course I do that with commercial ice creams, too. I fear, though, if you add much more there would be a significant up-tick in unpleasant GI rumbling.

  2. I’ll have to try the polydextrose. We made some ice cream last night using erythritol. Have you tried (or have any opinions on} this sugar alcohol? The claim is that while absorbed (thus no GI discomfort) it has minimal effect on blood sugar, and is mostly excreted intact.

    COMMENT from MD EADES: Although I think the jury is still not fully in on it from a metabolic standpoint, it may turn out that erythritol is wonderul. For me, personally, it leaves a faintly metallic aftertaste that I can detect and don’t relish. But that’s just a personal issue for me.

  3. Thank you so much for the great healthy ice cream recipe. We just bought a new ice cream maker today and I can’t wait to try this out! One question though…when using the 1/3 – 1/2 cup Poly-D, would that be instead of Splenda, or in addition to Splenda? I read above that Poly-D is only 1/10 as sweet as sugar, so I’m wondering if you still have to add sweetener on top of it. Thanks in advance!

    COMMENT from MD EADES: Yes, in addition to; it wouldn’t be very sweet at all with just the polydextrose. You might drop the Splenda by a couple of two or three packets since the polydextrose will offer a little sweetness, but I just used the whole amount of Splenda specified in the original recipe.

  4. Polydextrose is definitely a wonderful addition to ice cream.

    Regarding the sorbitol content… technically, polydextrose is made with 10% sorbitol, but most of this sorbitol is converted during the manufacturing process. Here is the final nutritional breakdown for the brand of polydextrose that the majority of home cooks use:

    Polydextrose (8.1 net carbs/100g)

    More than 90% fiber
    8.1% net carbs, including:
    4% glucose
    0-2% sorbitol
    0-4% levoglucosan

    It still contains some sorbitol, which may very well lead to GI issues, but it’s definitely not 10%.

    Regarding the ‘faintly metallic aftertaste’ of erythritol… Erythritol is a very special sweetener in that it has an entirely different flavor profile in it’s granular/crystallized state than it does when it’s dissolved. Crystallized it has a strong cooling effect (that some would describe as a metallic flavor). Literally, when it hits your tongue, it causes a decrease in temperature. Dissolved, there is zero cooling effect- it’s a dead ringer for slightly less sweet sugar (70% as sweet). The tricky part is that erythritol hates being dissolved. If you’ve ever tried boiling a sugar syrup into candy, you’ll know how easily sugar recrystallizes. Erythritol is that times 10. This is where polyd comes in. Polyd is a powerful crystallization inhibitor. The same way corn syrup is added to candy recipes to prevent sugar from crystallizing, polyd can be used in conjuntion with erythritol for the same purpose. If you use enough polyd, the erythritol stays dissolved. How much? Well, it depends on the other ingredients in the recipe, but generally speaking, 4 parts polyd to 1 part erythritol is a safe ratio. Erythritol is especially useful in ice cream because it’s one of the most powerful freezing point depressors in the lc sweetener arsenal. The more erythritol you can use, the better the scoopability in the final product.

    Polyd has assimilation issues. When it hits liquid it has a tendency to form hard, pebble like clumps. The heat in this recipe will probably help to dissolve any clumping, but it might be a safer bet to find another way of incorporating it. Microwaving it with the rice milk first to create a syrup might give you more consistent results.

    Lastly, when it comes to mouthfeel, the emulsifying effects of polyd (and the egg yolks) help a great deal, but soluble fiber gums (xanthan, guar, thickenthin, etc.) are invaluable for creating a smaller ice crystal size. The smaller the ice crystals, the smoother the mouthfeel.

    Btw, the ‘tin snips of the digestive process’ is a wonderful metaphor.

    COMMENT from MDEADES: Thanks for all the great info about polydextrose and erythritol. It could just be the fiber bolus that causes most of the GI effects, if most of the sorbitol is converted to something else. What is it converted to, do you know? Just out of curiositiy.

    I will give the erythritol another try in ice cream; I grant that the metallic hint may very well be the cooling effect and if dissolving it does the trick I’m more than willing to try it in the next batch of ice cream. I always make a cooked custard base to dissolve all the polydextrose and Splenda anyway, so I could just try to dissolve the erythritol in the milk as it warms. I usually take it to about 140-160 degrees or so before cooling it back down. Course, I swore off blogging about ice cream (and alcohol) until after labor day, so I won’t be writing about it unti lthen.

    PS Don’t forget to put in the magic passcode word on the bottom of the comments page, else your coments get shunted to the Junk Comments folder–from whence I rescued this fine comment–with the spam.

  5. Just a question about a sweetener alternative to Splenda (it gives me headaches). Why don’t you recommend the herbal sweetener stevia? It’s 100% natural, has no glycemic index, is 150 times sweeter than sugar, and is roughly 50% of the Japanese sweetener market.

    COMMENT from MD EADES: We have no problem at all with using stevia or Stevia Plus (with FOS) and make mention of it often. From a practical standpoint, it’s not as readily available to everybody as Splenda, although that’s really not too much of a problem with internet ordering. Most it’s because it’s just a little bit more difficult to use, owing to its being what’s called a ‘bitter sweetener’. There’s a pretty fine line sometimes between an amount that makes a given recipe just sweet enough and the tad bit more that sends it to bitter. We’ve even blogged on its as a suitable alternative to Splenda or saccharine and recommended several stevia cookbooks for those who want to use it.

  6. I, too, have been having way too much fun with the ice creams and the polyd this summer.

    Just a suggestion, but I’ve found it useful to make a sweetened polyd “heavy syrup” with a corn syrup consistency, that I then jar up and use as needed — by the tablespoonful – in my ice cream custard/base. I’ve been using almost equal amounts of the polyd heavy syrup and glycerine, with a titch of the vegetable gums, and am very happy with the results.

    COMMENT from MD EADES: Sounds interesting. Do you just mix the polyd, glycerine and gums? Do you use any water? Do you have to heat it as one does for a sugar simple syrup to get it to fully dissolve?

    I haven’t ever been a great fan of glycerine, mainly because I worry just a bit about its propensity for side effects and the fact that it can raise blood sugar, at least in type I diabetics. Thus you have to wonder about exactly how absorbable it really is. In the rest of us who can produce insulin on demand, would that mask a measureable blood sugar rise we might otherwise see from it? I don’t know the certain answer. I suppose if one has no ill effects (GI symptoms, stall of weight loss, loss of control of BP, blood sugar, or lipids, for instance) then why not.

  7. I make the polyd heavy syrup first (my sweetener blend of Erythritol, xylitol, sucralose, ace-k, stevia – a few tbsp – & 3:1 polyd and boiling water, simmer til thoroughly dissolved and sweet to taste. Jar, label 🙂 and cool. Then I make the ice cream base as usual & when the milk component is scalded, I stir in the heavy syrup & glycerine (no further sweetener required, proportions are proprietal :] but probably match yours fairly closely for the polyd). I beat the eggs & (with the scalded milk standing by for tempering) beat in the gums then temper with hot milk mixture QUICKLY (the gums have amazing viscosity in the egg yolks). Proceed as usual.

    This is the only thing I use glycerine for; it’s scoopability qualities make it worth while, though I have the same issues with glycerine (& fructose for that matter) vis a vis BP, blood sugar, GI symptoms, etc.).

    I’d like to comment that I’ve admired your (and Dan’s :] ) work for years, I first read you in 1998 or 1999 & found the seminal Protein Power to be the first reading that made technical SENSE to me. Thank you!

    COMMENT from MD EADES: Thanks so much, both for the information and for the compliment about our work. One thing, though, just FYI, his name is Mike. It’s a common and quite honest mistake that’s cropped up for years. My first name is Mary Dan (thus, the MD in the comments from above) and although our middle son is named Dan, he hasn’t contributed much to the march of science…at least to date and he’s 33, so I’m not holding my breath that he’ll take a turn toward science at this late date.

  8. I find myself making/using a polyd/erythritol syrup for more and more recipes these days, not just for ice cream. Experience has shown me that both granular erythritol and polyd require boiling temperatures to dissolve (higher even if the polyd is badly clumped). By making the syrup beforehand, I don’t have to worry whether or not the dessert reaches a sufficient temperature, because all the hard to dissolve ingredients are pre-dissolved.

    A pure erythritol/water solution will re-crystallize ferociously when cooled (like very old honey, but istantaneously and times 10). Fortunately polyd prevents this. With the right ratio of polyd to erythritol (my experimentation seems to reveal 4 to 1 to be the magic number), the erythritol doesn’t re-crystallize/stays dissolved when cooled.

    By using a minimal amount of water in the syrup, you can add a substantial amount of polyd/erythritol to a recipe without adding too much water in the process. The less water you make the syrup with, though, the trickier it gets, as the polyd will have a greater tendency to clump. If I stir the polyd/erythritol vigorously while adding the water, I find a 4:1:1 polyd:erythritol:water syrup integrates nicely when microwaved.

    The other nice thing is that polyd’s large molecular size makes it excellent at curtailing water activity, so you can store this syrup in the fridge for months without worrying about it going bad.

    COMMENT from MD EADES: Great advice, I’ll have to try it. Ain’t kitchen science fun?

  9. My apologies for calling Mike by Dan, some kind of slip (Freudian? Nah). I must admit almost blushingly to thinking of you (two) somewhat collectively, generally as the Doctors Eades… 🙂

    And I’m with Scott on this one, “syrups” are cropping up with increasing frequency as a texturizer *and* flavour vehicle in my kitchen — not just sweets, either. 😉

    COMMENT from MD EADES: No problemo. We’ve lived, worked, played, and travelled together for so many years, that we practically are one person.

  10. “…if most of the sorbitol is converted to something else. What is it converted to, do you know?”

    I recently stumbled across the original Patent (# 3876794) for polydextrose. Here are two quotes relating to the final fate of the converted sorbitol:

    “The inclusion of a food acceptable polyol such as sorbitol in the saccharide-carboxylic acid reaction mixtures prior to polycondensation yields superior products. In most cases, 90% or more of the polyol cannot be isolated from the condensation product, demonstrating that is has been chemically incorporated in the polymer.”

    “Where sorbitol or other polyol is incorporated in the polymerization mixture such agent tends to be incorporated at the end of the polymer chain, in which case molecular weight determinations based on end-group methods prove inaccurate.”

    COMMENT from MD EADES: Thanks for the infor, but I don’t see that this patent pertains to what happens to polyols in the human GI tract. It’s completely about a new method of manufacture that inlcudes using polyols in the polymerization of glucose or maltose to make polyD for food use.

  11. I’m not referring the GI tract here. I’m addressing your question as to the fate of the original 10% sorbitol in the manufacturing process of polyd. In August, you had asked me:

    “COMMENT from MDEADES: Thanks for all the great info about polydextrose and erythritol. It could just be the fiber bolus that causes most of the GI effects, if most of the sorbitol is converted to something else. What is it converted to, do you know? Just out of curiositiy.”

    This was a follow up to that question. The sorbitol, at least 90+ percent of it, becomes one more link in polyd’s molecular chain.

    COMMENT from MD EADES: So I did. Sorry, I’ve slept since then ;D

  12. I arrived at your site while searching for recipes for peanut brittle using just erythritol as the sweetener and brittlizer. When I read your use of polydextrose in ice cream, with its sorbitol content and its problem with intestinal upset, namely gas and diarrhea, I thought this sounds crazy. I was happy to find erythritol with its promise of not creating these problems. I think the other -tols should not even be used in the food industry for causing digestive problems.
    Instead of using poly-D, I would call TIC Gums and find the perfect gum mixture to perfectly adjust the mouth feel of my ice cream.

    Comment from MD Eades:  Thanks for the suggestion.

  13. Hello there! I’m on a low carb sugar free ice cream making mission, and I learned a lot from reading your post, and the comments as well!
    I would love to know the base recipe you’re referring to, to get a better idea of how to incorporate poly-d into the recipes. And to give it a try of course! But unfortunately the link no longer works. I tried to search for it (on Google and on the blogs) but could not find it.
    It would be wonderful if you could provide an updated link, or copy and paste the recipe- I would be so grateful! Thank you so much 🙂

    1. Hi, Kate! Thanks so much for alerting us to the broken links! I’ve gone in and found a different source and re-linked the post with the recipe. Let us know how it works for you and what you think.

      1. Thank you so much! I just made my very first home made ice cream. Dairy free and sugar free, and it is so easily scoop-able and beautifully creamy right out of the freezer! I tweaked your recipe with some tricks I had learned elsewhere, and I’m so very pleased!! This is way better than the (rock hard) sugar free ice creams from the store!

        For you and your readers, I make my own dairy free heavy cream (then half and half) for substitutions. Blended to emulsify 2/3c silk unsweet soy with 1/3c vegetable oil for every 1 cup of replacement heavy whipping cream. Then used 7oz of that and 7oz more soy to make half and half. (You can also do this with Ripple milk to make it soy free, coconut free, nut free, and low carb!)

        I used 1/2c allulose and 1/2c Swerve confectioners to replace the sugar. Using xylitol, allulose, or bochasweet is another trick to sugar free ice cream. I added the 1/3c PolyD of course. I also mixed 1/4 tsp xanthan gum into the sugars before adding, and 1 Tbsp vegetable glycerine to the eggs before tempering.

        Followed all your directions. The base wasn’t sweet enough for me at the end, so I whisked in 4 tsp NatriSweet Monk Fruit Extract powder. About the equivalent of 2 cups of sweetness. Stevia extract would also work well here to add sweet without bulk, but I’m more partial to monk fruit.
        I also used LorAnn’s cake batter and cream cheese icing super strength extracts to make a birthday cake flavor.

        Chilled over night, churned for about 35 minutes in my Cuisinart 2qt ICE-30BC. Poured and froze over night, and done! So scoopable, so creamy, so delicious, and 2g net carbs per 1/2c serving! My diabetic dairy allergic self is so pleased with the result! All thanks to your help- Thank you! And now I’m going to go share what I’ve learned with others so they can be this happy 🙂

        1. Happy to share your success. One caveat, though, is the use of ‘vegetable oil’ in the recipe. If by that you mean what’s normally called ‘vegetable oil’ and is corn, cotton seed, rapeseed, and other industrially processed seed oils, that would not be a healthful choice. Could it be substituted with maybe Macadamia Nut oil? Or coconut oil? Just a thought. Thanks for the input.

Leave a Reply

Your email address will not be published. Required fields are marked *