September 24, 2019

Is Insulin Messing with Your Skin?

Question: Is there anything in the body insulin doesn’t affect?

Answer: From what I’ve seen, no.

If you’re new to my blog or are a newcomer to the science of the clinically therapeutic effects of dietary carbohydrate restriction, you might think of insulin mostly as a blood sugar hormone. People with diabetes have blood sugar that’s too high, so they take insulin to bring it down. Simple, right? Not quite.

I’ve written in past posts that reading and learning I’ve done over the past few years has led to me to the perspective that lowering blood sugar is among insulin’s least important effects. (In fact, insulin isn’t even required to lower blood sugar at all. Your body can do that just fine without insulin…even in someone with type 1 diabetes. Type 1 diabetics do need insulin, just not solely for the purpose of lowering blood sugar. Details on all this here.)

I’ve written articles about insulin as a major factor in the development of gout, migraines, Alzheimer’s disease, PCOS, erectile dysfunction, benign prostatic hyperplasia (BPH, a.k.a. enlarged prostate), Parkinson’s disease, and more. The short list of things we know for certain are damaged by chronically elevated insulin and/or blood glucose (BG) includes the liver, kidneys, eyes, cardiovascular system (heart muscle and blood vessels), ovaries, the brain and nervous system. At this point, knowing what I now know about insulin, I don’t need someone to explain to me why insulin would affect any particular organ, gland, or tissue system; I need them to explain why it wouldn’t.

With this in mind, is it possible insulin is affecting your skin? You can’t see a fatty liver, polycystic ovaries or an enlarged prostate gland from the outside; you need special tests to determine for sure whether you have those. But what about acne, skin tags, psoriasis, and other things we can see just by looking at someone? Could insulin be playing a role here, too?

Tl;dr: If you want a brief summary of this article, read this. And if you have a few extra minutes and want to read one paper that will give you an enormous amount of insight into this topic, read this one. But if you come to my blog because you enjoy digging into the meaty details, stay here and keep reading. 

Before We Get Started…

I mention something called HOMA-IR several times in this post. If you don’t know what HOMA-IR is, take a couple minutes and read this article I wrote about it for Heads Up Health. (It’s much shorter than my normal blog posts, so fear not.) If you really don’t want to spare that time, what you need to know is that HOMA-IR is a mathematical relationship between your fasting glucose and insulin levels. The higher your HOMA-IR, the more insulin your body requires to keep your blood sugar where it is, so the less sensitive to insulin (i.e., the more insulin resistant) you are. This is a great way to identify the many (many!) people who have totally normal glucose at the expense of super-elevated insulin—an important concept I wrote about here. You also need to keep in mind that HOMA-IR is based on fasting levels of glucose and insulin, and there are plenty of people whose fasting insulin is normal, but it skyrockets after a meal (especially a high-carb meal), and then stays elevated for several hours. So there are people who have a normal HOMA-IR yet are nonetheless living with multiple signs & symptoms of chronic hyperinsulinemia. This is not reflected in lab tests when fasting insulin registers as normal. As far as I know, there are no universally accepted cutoff points for HOMA-IR, but a general starting point would be < 1 for optimal insulin sensitivity, > 1.9 indicates mild insulin resistance, and > 2.9 indicates severe insulin resistance. Okay, on with the show!


For people living with acne, skin tags, psoriasis, lichen planus, and other visible skin disturbances, these issues can severely diminish self-esteem and quality of life. But these things are far more than skin deep. A staggering amount of research indicates they are the surface manifestations of damaging hormonal imbalances on the inside—imbalances that could eventually result in type 2 diabetes, cardiovascular disease and other long-term consequences of chronically elevated insulin, a.k.a. metabolic syndrome. 

Most of us are accustomed to thinking that greasy foods equal a greasy face. We assume that acne, especially in teenagers, comes from indulging in pizza, potato chips and french fries. But what about other skin conditions that we don’t associate with these foods—things like hidradenitis suppurativa and increased pigmentation? We might also assume that skin breakouts come from poor hygiene—either not cleaning well enough or touching the face with dirty or oily hands. But in the industrialized world, we don’t exactly live in squalor. Could poor hygiene really be the cause of this stuff, or might there be a much better, more scientific explanation?

Before we look more closely at some of the research supporting a role for insulin in skin problems, I need to clarify that while insulin resistance (IR) is typically associated with people who are overweight, obese, or have type 2 diabetes (T2D), chronic hyperinsulinemia can and does occur in people of “normal” body weight and who are not diabetic. Consider the fact that neither body weight, nor BMI, nor even body fat percentage, are part of the official diagnostic criteria for metabolic syndrome (MetSyn). They’re not taken into consideration at all! You can have severe metabolic problems even if you’re at a “normal” weight. (I wrote about this here.) Fasting glucose is a criterion, but the diagnosis of MetSyn can be made with just three of the criteria (such as elevated blood pressure, high triglycerides and low HDL), so you can be diagnosed with it even if your fasting BG is normal. Many of the long-term consequences from T2D (retinopathy, nephropathy, vascular disease) are thought of as consequences of chronic hyperglycemia, but there is a massive epidemic of patients with normal blood glucose with chronically high insulin, and the elevated insulin, even in the absence of elevated glucose, drives a number of chronic non-communicable illnesses. This is a massively underrecognized problem that underlies numerous conditions not typically associated with insulin at all. (I co-authored a paper with Dr. Jason Fung about this. Read it here.)

Insulin and the Skin

It’s under-recognized among medical professionals—and even more so among patients—that a wide array of skin issues is caused or exacerbated by chronic hyperinsulinemia. The list includes but is not limited to acne, skin tags, acanthosis nigricans, psoriasis, cellulitis, and hidradenitis suppurativa. One paper referred to these as “cutaneous manifestations of obesity and the associated metabolic syndrome,” but what about patients with these skin issues who are not obese? Again, hyperinsulinemia is not a condition exclusive to individuals who are overweight or obese. Metabolic syndrome should not be ruled out in patients at what would be considered a “healthy” body weight who present with one or more of these skin issues. (If you’ve been around my blog for a while, you know I am constantly harping on this point because it is highly personal to me, having struggled with weight for most of my life and also being aware of the terrible stigma overweight people face in healthcare.)

The influence of insulin on the various hormonal abnormalities in PCOS is well known, and PCOS brings with it several skin manifestations such as oily skin, acne, and hirsutism, so there’s good reason to believe chronic hyperinsulinemia could be driving other skin issues as well, and not just in women. “Paleo” diet authority Loren Cordain and the beloved Drs. Mike and Mary Dan Eades (authors of Protein Power Lifeplan) have been writing about this since at least 2003, and their paper wasn’t the first to make some of these associations, so this isn’t exactly breaking news. Their paper is one of my all-time favorites. Every time I read it, I notice something I’d glossed over in the past—something important that I skipped over only because my knowledge that at the time wasn’t sufficient for me to really make sense of it. As my knowledge has expanded and deepened and I revisit some of my favorite older papers, I’m often astounded at how people were writing about certain mechanisms and pathways decades ago and these critical bits of information still have not reached most doctors’ offices. But I digress.

Let’s take a closer look at the role of insulin in select skin issues.

Skin Tags

Skin tags (medically called acrochordons) are an interesting thing to explore with regard to the role of insulin in skin issues. They typically appear on the neck, in the armpits, or near the groin. While we might see these as aesthetically unpleasing, and they can be uncomfortable due to itching, pain, or rubbing against clothing, most of us would consider them benign in terms of being a medical problem. However, research indicates these are anything but benign. Study authors have described skin tags as “a cutaneous marker for impaired carbohydrate metabolism,” and others have noted the same thing, saying that skin tags are “a cutaneous sign of impaired carbohydrate metabolism, hyperlipidemia, liver enzyme abnormalities and hypertension” – all issues that cluster with insulin resistance. In a nutshell, if you have skin tags—especially multiple skin tags—you probably have chronic hyperinsulinemia, even if you have no other obvious signs or symptoms.

In a case-control study of subjects with two or more skin tags,  compared to age- and gender-matched controls with no skin tags, subjects with skin tags had significantly higher mean levels of fasting glucose, triglycerides and liver enzymes (ALT, AST, GGT and alkaline phosphatase), along with higher blood pressure and lower HDL—all indicative of metabolic disturbances tied to chronic hyperinsulinemia. Out of 110 subjects with skin tags, a 2-hour oral glucose tolerance test (OGTT) showed that 56 of them (51%!!) had overt type 2 diabetes, which probably came as a shock to most of them. (10 of 110 people with no skin tags were also diagnosed…, not great for them, either.)

In a separate case-control study comparing subjects with at least three skin tags to controls matched for age, sex and BMI, subjects with skin tags had a higher frequency of T2D (23% among cases versus 8.5% for controls). There was a positive correlation between the total number of skin tags and mean fasting plasma glucose, leading the study authors to write, “With regard to the importance of early diagnosis of diabetes, we recommend a high level of suspicion for impaired carbohydrate metabolism in patients with skin tag.” No correlation was found between the number of skin tags and BMI, meaning that heavier people didn’t have more skin tags than people at lower body weights. Again, it’s not about body weight; it’s about insulin levels, and in some individuals perhaps glucose as well. Many people probably think skin tags only happen in people who are obese and assume this is a consequence of obesity. But first of all, they don’t only occur in obese people, and second, what if the obesity isn’t causing the skin tags, but rather, some other factor is causing both? (And you get one guess as to what that other factor is.)

These are not isolated findings. Another case-control study compared three groups of subjects with multiple skin tags (30 subjects at a normal BMI, 30 overweight and 30 obese) to healthy controls (no skin tags) matched for age, sex and BMI. HOMA-IR was significantly higher in all groups of skin tag patients compared to BMI-matched controls, with 71% of cases meeting the criteria for metabolic syndrome – seventy-one percent! Yet again we see that body weight is not the driving factor in development of skin tags, because individuals at the same BMI do not show equal development of skin tags.

It should be noted, though, that among the skin tag patients, higher BMI was correlated to increased number and greater extent of skin tags. This finding was echoed in a later study of nondiabetic subjects who were overweight or obese and had skin tags (77%) or other skin issues: acanthosis nigricans (97% of subjects), keratosis pilaris (42%), or plantar hyperkeratosis (38%). The degree of obesity was significantly associated with acanthosis nigricans, skin tags and plantar hyperkeratosis, while number of skin tags, acanthosis nigricans distribution and neck severity score were significantly and independently associated with insulin levels. These subjects were nondiabetic; so here again we see it’s not the glucose; it’s the insulin. (There’s a reason I called my multi-part blog series “It’s the Insulin, Stupid!”)

The study authors wrote that skin tags (and acanthosis nigricans, which I’ll get to in a bit) “should be considered clinical markers of hyperinsulinemia in nondiabetic, obese patients.” This is an important point that I’ll revisit a couple times and come back to toward the end of this post. For now, I’ll just say that it hints at completely non-invasive ways to assess metabolic health. Do you have skin tags? Especially multiple skin tags? Whether or not you are overweight or obese, guess what? You’re probably hyperinsulinemic and are on your way to type 2 diabetes, cardiovascular disease, kidney failure, and more, if you don’t do something about this.

Can you stand to read about another one?
“Multiple skin tags appear associated with abnormalities in glucose/insulin metabolism.” “Multiple skin tags should raise suspicion of insulin resistance or hyperinsulinemia.” Researchers know this, so why don’t doctors?

Clinical and metabolic glucose/insulin characteristics of men with multiple skin tags (8 or more) on the neck were compared with a control group with few or none. Both groups were divided in two subgroups according to normal or abnormal lab test findings. In the subgroup with normal lab findings, the number of skin tags varied from 8-33, whereas in those with abnormal laboratory findings the range was 9-65. (Sixty-five skin tags? YIKES!)

Eight or more skin tags were related with statistically significant glucose/insulin abnormalities including fasting hyperinsulinemia, postprandial hyperinsulinemia (high insulin after meals), and postprandial hyperglycemia. In the multiple skin tag group, testing showed 77% of them to have insulin resistance, basal/fasting hyperinsulinemia, postprandial hyperinsulinemia, glucose intolerance, or type 2 diabetes. (In the control group with few or no skin tags, 33% of subjects showed some of the same lab test abnormalities, which tells us that even among seemingly “healthy” people, there’s still a high prevalence of metabolic dysfunction!)

If you really want to see how bad things are, the researchers noted that only 15% of subjects with some of these “metabolic abnormalities” showed no cutaneous signs of glucose/insulin alterations. So 85%—eighty-five percent!—did have visible signs of hyperinsulinemia. Here’s the kicker, though: their definition of “cutaneous expression of glucose/insulin alterations” was 9 or more skin tags on the neck, acanthosis nigricans, or a waist circumference greater than 95 centimeters (~37.4 inches). Nine or more skin tags? That’s their threshold? So let me get this straight: someone can have eight skin tags on their neck alone, and not be considered to be showing cutaneous signs of chronic hyperinsulinemia. Okay, sure, LOL. You can’t make this stuff up, folks!

One of the studies I mentioned earlier determined that compared to people matched for age, sex, and body mass who had no skin tags, subjects with at least 3 skin tags had a higher frequency of type 2 diabetes (23% vs. 8.5%) – in fact, over fifty percent of the subjects with 30 or more skin tags were shown to have diabetes. Over fifty percent! (But also…thirty skin tags? I’m repeating myself here, but: YIKES!)

The authors wrote, “With regard to the importance of early diagnosis of diabetes, we recommend a high level of suspicion for impaired carbohydrate metabolism in patients with skin tag.” Let’s focus on the first part of that sentence: With regard to the importance of early diagnosis of diabetes… Early diagnosis. For some people out there, skin tags might be the first sign of “impaired carbohydrate metabolism.” Maybe the only sign. We don’t like the way skin tags look, but this is actually good news: If someone has skin tags, this is very likely an early warning sign of deranged carbohydrate metabolism. They don’t have to wait for this to become full-blown T2D or MetSyn. They don’t have to wait until their HbA1c is 11.5 or until they’ve done permanent damage to their eyes, kidneys, or neurons. They can nip this carb problem in the bud now. 

And for the cherry on top, once again, there was no correlation between subjects’ BMI and their number of skin tags. Say it with me, dear readers: it’s not about weight, it’s about insulin. In this case, it was also about glucose, because there was a modest positive correlation between mean fasting glucose and total number of skin tags.

How does this work? What’s the mechanism? Well, insulin is an anabolic hormone. It makes stuff grow. Some of us know only too well it grows the hell out of our fat cells. It also makes the prostate gland bigger. It causes cysts to grow in the ovaries. So it’s not really a shock that it can make random bits of skin grow larger and extend out from the body. I haven’t looked much into the mechanism beyond this admittedly overly simplistic explanation. But the thing is, we don’t have to know how insulin causes skin tags. We just know it does, and we know how to correct this. (Actually, there are some mechanisms I’ll cover in more detail later on. Keep reading.)

As an aside, are you a parent? Do you have young children? Please know that these “cutaneous manifestations” of insulin resistance are not exclusive to adults. They happen in children, too, so if you have a child with skin tags or acanthosis nigricans, you might want to consider doing something about their carbohydrate intake. (Can kids do keto? Yes! I did a video on this topic which you can watch here.)

Acanthosis Nigricans

There’s zero debate that most cases of acanthosis nigricans (AN) come from chronic hyperinsulinemia. One paper’s authors went so far as to say, “the presence of this skin lesion is a clinical surrogate of laboratory-documented hyperinsulinemia.” Meaning, if someone has AN, you can basically assume they have chronically high insulin. (AN has multiple potential causes, though. Hyperinsulinemia isn’t always the cause, but it’s the most common one.)

AN is more common in people who are overweight or obese, but AN can also occur in people at a normal weight. (Please imagine quotation marks around “normal” anytime I talk about a normal weight, because there really is no such thing as a normal weight.) What is AN? It’s discoloration of skin, mostly in places where there are skin folds, like the armpits, knees, knuckles, elbows, eyelids, and back of the neck (especially in obese people, who might have a fold there where others don’t), but it can happen in lots of other places, too. The skin in these areas turns some variation of black, brown, or ashy gray, and becomes rough in texture. With a warning that you might not want to see these, here are some images of what this looks like.  

(Aside: I am infuriated that the American Academy of Dermatology says AN is “not harmful.” What, now? What? Okay, fine. I will grant you that the AN, itself, is not “harmful,” but what causes AN most definitely is harmful out the wazoo. Perhaps this organization should change its name to the Jupiter Academy of Dermatology, because they are clearly not living on planet Earth.)

How does AN work? Why would chronically elevated insulin cause this? According to one paper, “The endocrine origin of this condition is beyond doubt. Insulin and insulin-like growth factor-1, and their receptors on keratinocytes are obviously involved in the complex regulations leading to the peculiar epidermal hyperplasia.” Basically, chronically high insulin is like the body sprinkling insulin all over the skin (from the inside) like a fertilizer. Fertilizer helps things grow, and so does insulin. Constant exposure to high levels of insulin causes keratinocytes or dermal fibroblasts (skin cells, basically) to grow and/or proliferate. This seems to be the primary mechanism: “Elevated insulin concentrations result in direct and indirect activation of IGF-1 receptors on keratinocytes and fibroblasts, leading to proliferation.” (The same mechanism is likely at work in skin tags.) In a nutshell, insulin affects approximately a bajillion other hormones, some of which have roles in regulating skin cells. (We saw in the PCOS post that insulin affects testosterone, luteinizing hormone, follicle stimulating hormone, sex hormone binding globulin and more, so why wouldn’t it affect a slew of other hormones, proteins and enzymes?)

As I mentioned for skin tags, children are not immune to AN, and it’s not solely seen in children who are overweight or obese. One study’s authors “propose that all children, adolescents, and youth with AN be screened for IR irrespective of BMI.” In comparing subjects between ages 2-24, those with AN had substantially higher HOMA-IR than those matched for age and sex. (All healthy/non-AN subjects were at a normal weight. Subjects with AN were included regardless of BMI, so there were some normal weight, and some overweight or obese.) The clincher here is insulin: Among the AN cases, all subjects had high fasting insulin but normal glucose. Average fasting glucose was in the high 80s (mg/dL; approx 5.0 mmol/L), but look at insulin: In the low-carb world, we like to see fasting insulin under 10 µIU/mL. Among the children and young people with AN, mean fasting insulin was 15.6, 18.8, and a whopping 27.6 in subjects who were at a normal weight, overweight, or obese, respectively. Even the normal weight ones had high fasting insulin! And just to reiterate, the blood glucose was normal in all cases. All of them! It really is the insulin, stupid.

The HOMA-IR also increased between each group: from normal weight to overweight to obese, mean HOMA-IR went from 1.98 to 2.38 to 3.38, respectively. (For clarity, they defined insulin resistance as HOMA-IR greater than 1.8 and some doctors I know consider anything above 2.9 to indicate severe insulin resistance), so you can see that the normal weight subjects were very slightly insulin resistant, overweight subjects were more so, and obese subjects were severely insulin resistant. Among the subjects with AN, based on HOMA-IR, insulin resistance was found in 28%, 50%, and 81% of subjects who were normal weight, overweight, or obese. Tell me again why fasting insulin is not a routine part of standard bloodwork? Should a teenager’s armpits have to turn black before they find out they have a massive insulin problem? Should they have to develop full-blown diabetes, for crissake? And remember, HOMA-IR is based only on fasting insulin and glucose. My guess is that somewhere close to 100% of subjects with AN have post-prandial hyperinsulinemia, so even though their fasting level is normal, it skyrockets after they eat and then stays high for most of the day. But the fact that 81% of obese people with AN have severe insulin resistance based on the fasting level alone tells us all we need to know here.

We see the same thing in a study of children ages 10-18. 62% of those with AN had “high” insulin resistance. This rose to 80% in children who had AN and also a high BMI. (As usual, though, in this study, insulin resistance was determined by HOMA-IR. Would those 62% and 80% figures be higher if they had measured post-prandial insulin?) Conclusion: “AN can be used as a screening method to identify children at risk of DM Type-2-since those who have high IR have a high possibility of having DM Type-2 in future. Hence, early screening and simple, but effective interventional strategies can be instituted at this age, which may prevent or delay diabetes in the long run.” Indeed. I can think of at least one really good interventional strategy, and I’m sure you can, too.  ;-)

A paper from 1992twenty-seven years ago—includes this gem: “Data suggest that acanthosis nigricans is an easily detected empirical marker for elevated risk of type II diabetes. The lesion can appear long before the onset of glucose intolerance. Thus, including acanthosis nigricans screening in a comprehensive disease-prevention program can help identify people at risk for type II diabetes prior to the actual onset of glucose intolerance, as well as individuals with undiagnosed diabetes.” Just like we saw with skin tags, our skin gives us clear signs of what’s going on inside. People don’t have to wait until they’re on the verge of losing organs, limbs, or going blind before they have a screaming sign that Something. Is. Wrong. AN is a serious canary in the coalmine. It can tip you off to a potentially terrible situation while it’s still in an early stage, when you can turn it around before it becomes terrible.     

One last word on AN: There are other causes for AN besides hyperinsulinemia. This is why I said “most cases” are insulin-related. So if you or a loved one have AN, it’s not a slam dunk that you are chronically hyperinsulinemic. That is one potential cause that is easily confirmed or ruled out. If it’s ruled out, work with a doctor to determine the cause.


I mentioned acne earlier and I think it’s important to look a little more closely at the role of chronic hyperinsulinemia here, too. It’s fairly well known, even in mainstream medicine, that AN is a sign of insulin problems. I think it’s less known with skin tags, and much less known with acne. Like I said, acne is often blamed on a bad diet, particularly in teenagers. Acne can most definitely result from hormonal imbalances unrelated to insulin (hello, puberty!), and can of course be a manifestation of a food sensitivity. Plenty of people know their skin breaks out when they eat certain foods. But what about when there’s seemingly no explanation for it?

Considering the influence of insulin on so many other hormones, acne does result from a hormonal disturbance. But what’s causing this disturbance? Women with PCOS often have oily skin and/or acne. And what causes PCOS? Right.

Indeed, in a study looking specifically at women with acne, compared to women without acne, those with it were found to have higher levels of fasting insulin, fasting glucose, HOMA-IR and area under the curve for glucose and insulin measured during an OGTT. (“Area under the curve,” or AUC, is the total amount of glucose or insulin secreted during the time period measured.) One of the hallmarks of PCOS is elevated androgens (male characteristic sex hormones). To eliminate this potential confounder because we already know women with PCOS or elevated androgens will probably have acne, the researchers analyzed the data a second time, this time excluding patients with elevated androgens. Even so, the women with acne still had higher fasting insulin, HOMA-IR and AUC for insulin and glucose, so PCOS or not, elevated androgens or not, elevated insulin drives acne in women. The authors’ conclusion? “Anti-insulin drugs may be an adjunctive treatment of female acne.” Um, yeah. Or how about an insulin-reducing diet?

And what about men? Grown men—not just teenagers—get acne. In a study of adult men with acne (men age 20 or older), prevalence of insulin resistance was double that in men without acne (22% compared to 11%). However, here again, insulin resistance was determined by HOMA-IR, which you know by now is based solely on fasting levels of insulin and glucose. How much higher would this prevalence be if they had assessed post-prandial levels? The authors wrote, “Postadolescent male patients with acne more commonly have insulin resistance. This resistance may be a stage of prediabetes, and the patients may develop hyperinsulinemia or type 2 diabetes in the future. These patients should be followed up for a prolonged time to determine whether they develop conditions associated with insulin resistance.” Well, yes, we could “follow these patients for a prolonged time” to see who gets diabetes or fatty liver, but how about this crazy idea: how about we correct the hyperinsulinemia so they don’t have to get diabetes? We can sit by, wait a while and watch as this trainwreck happens right before our eyes, or we can wake the conductor up so he doesn’t drive this train straight off the cliff into a ball of flames. I mean, call me crazy, but really.

A different study in young men with acne showed that compared to men without acne, men with it had more features of metabolic syndrome and carbohydrate intolerance, such as higher blood pressure, higher BMI, lower HDL, larger waist circumference, higher HOMA-IR, higher fasting insulin & glucose, and higher insulin & glucose during an OGTT. In acne subjects with a BMI < 24.9 (“normal”), the only two factors that correlated with acne were low HDL and high insulin at the 2-hour mark of an OGTT. The study authors said that the low HDL and high insulin were “independent predictors of acne” in the men at a normal BMI. The fact that fasting insulin level was not predictive, but the insulin level after the OGTT was, is important: it shows us the shortcomings of relying solely on fasting levels. These young guys think they’re healthy and their doctors probably do, too (especially because of the normal BMI), but what if their acne is being driven by a Mountain Dew habit, or a penchant for beer and nachos? And what if those dietary preferences are leading to something far more serious than acne?

What about mechanisms? The study authors wrote, “IGF‐1, by stimulating androgen production from the gonadal and adrenal glands, is responsible for the comedogenic [pore-clogging] effects of androgens, growth factors and corticosteroids.” Here we need to understand that insulin itself stimulates IGF-1 like crazy, partly directly and partly indirectly, via reducing synthesis of IGF binding proteins. Less of the binding protein means more of this hormone is free in the bloodstream to do what it does, like make stuff grow.

They also wrote, “Young male patients with acne resistant to common therapies may have an impaired metabolic profile and decreased insulin sensitivity.” And, “In male patients with acne resistant to common therapies, a possible diagnostic/therapeutic algorithm is the following: evaluate serum glucose and insulin levels, then HOMA‐IR, and finally OGTT.” Do you think the average dermatologist has any clue that for many acne patients (not all, but many), they don’t need lotions and potions, creams and ointments; they need to lower their insulin levels, and we all know of at least one damn effective way to do that.


I referenced the paper by Dr. Cordain and the Drs. Eades earlier, and I also said I would go into a bit more detail on specific mechanisms that might explain how chronically high insulin drives various skin issues. Well, here we are. This paper has the fabulous title, Hyperinsulinemic diseases of civilization: More than just Syndrome X. I give you my highest recommendation to read the whole thing if you really want to learn this stuff. (The paper is easy to understand; you don’t need a PhD to make sense of it.) They go a bit deeper into the IGF and IGF binding protein issue, explaining that this cluster (high insulin, reduced IGF binding proteins and elevated free IGF) is a major contributor to acne and cellular proliferation in general. (There are a few different forms of IGF binding proteins. I explained in the men’s health post that one of these is reduced [alongside elevated insulin] in prostate enlargement. It appears the same is true in some forms of acne. Insulin, either directly or indirectly, makes our fat tissue grow, makes the prostate gland grow, and makes skin grow, both in the form of skin tags and the “eruptions” we see as acne.)

These authors also have a fabulous explanation of the role of elevated androgens in acne, especially as it relates to the sebum in pimples and might explain why acne is especially severe in teenage boys, whose testosterone levels are surging. Take chronically elevated insulin and its role in both increasing IGF and reducing synthesis of IGF binding proteins, plus the androgens stimulating production of sebum, and you have a recipe for oily skin and pimples—as well as skin tags, acanthosis nigricans, and a host of other issues associated with “hyperkeratinization” and pathological changes to the functioning of epidermal/skin cells:

“The reductions in IGFBP-3 stimulated by elevated serum insulin levels or by acute ingestion of high-glycemic carbohydrates may also contribute to unregulated cell proliferation in the follicle. Hyperinsulinemia causes overexpression of the epidermal growth factor receptor (EGF-R) by elevating plasma non-esterified fatty acids, and also induces production of transforming growth factor-beta (TGF-beta1). Increased concentrations of EGF and TGF-beta1 depress localized keratinocyte synthesis of IGFBP-3, and thereby increase the availability of free IGF-1 to its keratinocyte receptors which in turn promotes keratinocyte proliferation. In addition, low plasma levels of IGFBP-3 induced by hyperinsulinemia may reduce the effectiveness of the body’s natural retinoids to activate genes that would normally limit follicular cell proliferation. Consequently, hyperkeratinization of sebaceous follicles may result synergistically from both elevations in free IGF-1 and reductions in IGFBP-3. Elevated sebum production, essential to the development of acne, is stimulated by androgens. Both insulin and IGF-1 stimulate the synthesis of androgens in ovarian and tissues.”

So yeah: like I said earlier, insulin influences so many other hormones, proteins, and enzymes that at this point, I don’t need someone to explain why insulin would affect basically every single organ, tissue, and gland in the body; I need them to explain why it wouldn’t.

Here’s another gem, this one from a paper titled Role of insulin resistance and diet in acne:

“Even though acne is considered an androgen-dependent disease, occurence [sic] of acne doesn't correlate with plasma androgen levels. Increased serum levels of IGF-1 have been observed in adult women and men with acne, giving rise to the possibility of the role of GH [growth hormone], hyperinsulinemia, and IGF1 in acne. A positive correlation between the mean facial sebum excretion rate and serum IGF-1 levels has been demonstrated in post-adolescent acne patients. Cappel, et al. demonstrated that IGF-1 levels correlate with severity of acne in women.”

With all this in mind, findings from dietary trials in people with acne will come as no surprise. In one such trial, young men with acne followed either a high or low-glycemic load (GL) diet for 7 days. The high-GL diet was 15% protein, 55% carbs, 30% fat; the low-GL diet was 25% protein, 45% carbs and 30% fat, so the low-GL diet was still relatively high in carbs, but the glycemic load was lower than the other diet and the protein was a bit higher by percentage, which is also an important factor. I would not consider the low-GL diet a low-carb diet (at 45% of calories from carbs, it clearly wasn’t), but even so, here’s what happened—and remember, this was after only 7 days: compared to subjects on the high-GL diet, subjects on the low-GL diet showed significant decreases in HOMA-IR and increases in IGF binding proteins. (On the high-GL diet, subjects’ HOMA-IR actually increased, and IGFBPs decreased.) This was a very small study—only 5 subjects on the high-GL diet and 7 on the low-GL diet, and it has other weaknesses too, but subjects were sequestered in a housing facility for the duration of the study, so we know exactly what they ate—something rare for a diet study. Can we conclude anything ironclad from this? No. Is it intriguing? Heck yes. Only 7 days, 45% of calories from carbs, and biomarkers predictive of acne still got better. Imagine what might’ve happened if the diet was actually low in carbs and continued for longer. These measurements likely would’ve improved even more, and as a result, I would expect substantial improvement in the acne. If you happen to have your own “success story” regarding LCHF/keto and acne, now you understand why this way of eating helped your skin.

This study was published in 2008. Apparently the authors of a 2017 paper missed it because they had some speculation about associations between carbohydrate intake, glycemic load, and acne. Results were based on a 5-day food record, which we know is shaky and unreliable, but there are also blood tests to rely on here: compared to subjects without acne, subjects with moderate-to-severe acne reported consuming foods with a greater total carb content, greater available carb (meaning, lower fiber), and a higher glycemic load. Blood tests revealed subjects with acne had higher fasting insulin and IGF-1, greater insulin resistance, and lower levels of sex-hormone binding globulin.   

Many people notice that dairy foods trigger their acne. Some people choose to consume it anyway, and I can’t say I blame them. (Life without cheese is no life at all. Not for me, anyway.) Some notice that goat or sheep dairy is less problematic than cow dairy. I don’t have a dairy intolerance, so I can’t say. For those who do have a dairy sensitivity that manifests as acne or some other skin issue, perhaps it’s due to the influence of dairy on IGF-1, particularly dairy foods that are also insulinogenic. 

This is worth noting because most of us low carbers loooove our dairy, right? Does the influence of dairy on IGF-1 mean we have to give up our beloved cheese, sour cream, heavy cream, butter, and cream cheese? Probably not. A paper looking at “evidence for acne-promoting effects of milk and other insulinotropic dairy products” concluded that “restriction of milk consumption or generation of less insulinotropic milk will have an enormous impact on the prevention of epidemic western diseases like obesity, diabetes mellitus, cancer, neurodegenerative diseases and acne.” So I don’t think we need to fear dairy. If you’re familiar with the work of nutrition research pioneer Weston A. Price, then you know that he found healthy, robust, long-lived populations who were dairy eating CHAMPS. So it’s not dairy, per se, that causes acne (or any other “disease of civilization”). It’s probably dairy in the context of a diet that’s already insulinogenic. Including cheese, butter, ghee, cream, sour cream or cream cheese in a very low carb diet is likely fine. Even plain, unsweetened yogurt is likely A-okay for most people, and I know many low-carbers who eat substantial amounts of this and remain lean and healthy. But if you drink fluid milk, and especially skim milk or sugar-sweetened yogurt? I would suspect problems from this. Bottom line: you probably already know if you have a problem with dairy. If you’ve been including dairy regularly on a low-carb/keto diet and you’re happy with your health, body size, and skin appearance, then keep calm and dairy on.

Other Skin Conditions

This post is already insanely long, so rather than go into detail on insulin/IGF and other skin issues, I’ll mention some briefly in case any of you reading this happens to have one of these or knows someone who does. But if you have a skin condition I don’t mention here, and no one knows what the cause of it is, you can’t go wrong assuming insulin probably has something to do with it.

Psoriasis: it’s recognized that people with psoriasis often have comorbidities such as cardiovascular disease, but also T2D, metabolic syndrome, non-alcoholic fatty liver disease, and obesity. What kills me is that multiple papers I’ve come across imply that psoriasis is the cause of metabolic syndrome and those other issues (or, at least, it “predisposes” people to them), rather than realizing that, mechanistically, it makes way more sense that psoriasis is yet another result of chronic hyperinsulinemia. One paper even said that psoriasis increased the prevalence of metabolic syndrome 3-fold and predisposes to the development of T2D and hypertension. Seriously? Come on. I can think of a few ways chronic hyperinsulinemia/MetSyn can predispose to psoriasis, but not the other way around. Thank goodness at least one paper acknowledges that metabolic syndrome is a “beginning step in the ‘psoriatic march.’” (Unfortunately, it says “progressive adiposity and the resulting metabolic syndrome” are the beginning steps, but as I’m always asking, what about all the people with psoriasis who are not overweight? And what if the “progressive adiposity” results from metabolic syndrome, and not the other way around? *Sigh.*) If you have psoriasis, you have pretty much nothing to lose by trying low-carb/keto.

Hidradenitis suppurativa (HS): I suspect few of us are familiar with this condition, so with a warning that some of these images might be disturbing, here’s what it looks like (NSFW!). According to the Mayo Clinic, HS is “a skin condition that causes small, painful lumps to form under the skin. The lumps can break open, or tunnels can form under the skin. The condition mostly affects areas where the skin rubs together, such as the armpits, groin, buttocks and breasts.” Also according to Mayo Clinic, “The exact cause of hidradenitis suppurativa isn't known. […] Experts think it could be connected to hormones […] excess weight and metabolic syndrome also might play a role.” Color me shocked. They recommend a slew of different treatment options and medications, but acknowledge that “No single option has been proven to be completely reliable.” LOL. How about keto? Why not try a diet that corrects metabolic syndrome when MetSyn/hyperinsulinemia appears to drive HS just like it drives everything else I’ve written about here?

Compared to age- and sex-matched controls, HS patients have higher rates of T2D, PCOS, dyslipidemia, hypertension, obesity, fatty liver, and other issues we would expect in people with chronic hyperinsulinemia. Just like people with acne, why should young people with HS remain totally unaware that their condition is being driven by hyperinsulinemia, and that unless they correct this, they’re staring down the barrel of diabetes, cardiovascular disease, kidney failure, foot amputation, or blindness when they’re older? They can try all the creams and pills and ointments out there, or they can correct the underlying problem. According to one paper, “Metabolic disorders including obesity and metabolic syndrome are the most common associated conditions observed in patients with hidradenitis suppurativa.” I bet some people have the knee-jerk assumption that HS occurs in overweight people because their skin rubs together so much, leading to the broken skin, pustules, etc. But maybe it’s the hyperinsulinemia driving the obesity and the HS. I mean, how do we explain HS in someone who’s not overweight? HS also appears to have a strong autoimmune component, and it’s pretty remarkable how many autoimmune conditions improve dramatically with keto. Just like with psoriasis, there are papers claiming that HS causes or predisposes to MetSyn. Helloooo? What’s the mechanism for that? On the other hand, having come this far in this post, you can easily propose mechanisms by which MetSyn predisposes for HS.

There’s a book about this condition, called The Hidden Plague: A Holistic Field Guide to Managing Hidradenitis Suppurativa & Other Skin and Autoimmune Conditions. I can’t vouch for it because I haven’t read it, but it’s about using a Paleo/Primal-style diet for this issue. Mark Sisson vouches for it (in fact, he published it first), and I’m a huge fan of Mark, so that’s good enough for me. If it’s more of a Paleo/Primal approach, it’s likely not keto, which is fine, but if I were living with this condition, I would try keto first and then liberalize my diet over time. 

Okay, I guess I lied, and I did go into some detail on these other skin conditions. What are some other skin issues that might be driven by chronic hyperinsulinemia, and that it would be worth trying keto for? How about rosacea, lichen planus, atopic dermatitis, seborrheic dermatitis, and anything else mentioned here: Metabolic syndrome and the skin: a more than superficial association. (What a great title for a paper, huh?)

Diabetes Drugs for Skin Conditions

In my posts on PCOS and the “male equivalent of PCOS,” I shared links to studies assessing the efficacy of metformin and other diabetes drugs for PCOS and erectile dysfunction. If you’ve read those, it will not surprise you at all to find out that they’re studying metformin and rosiglitazone in patients with acne, skin tags, acanthosis nigricans, eruptive xanthomas, hidradenitis suppurativa and more, and results have generally been favorable, although not always. The fact that so much research has been done studying the effects of metformin on difficult-to-treat dermatological conditions—and that so much of it is so promising—indicates that there clearly is a connection between these issues and insulin and/or blood glucose—and that researchers know it. So why isn’t this helpful little nugget reaching the offices of family physicians and perhaps more importantly, dermatologists? *Shrug.* Your guess is as good as mine.

In a study of men with acne “resistant to common treatments,” compared to subjects using a conventional anti-acne cream, subjects who used the cream and followed a low-glycemic diet (also considered low-calorie, at 1500-2000 cals/day) plus metformin for 6 months, subjects in the diet & metformin group had much greater improvement in skin appearance. (Check out this picture of one of the subjects -- huge change!) 

In a study of metformin in patients with hidradenitis suppurativa, first of all, 75% of subjects were found to have insulin resistance. Metformin therapy led to improvements in 36 out of 53 total subjects (68%), with 7 subjects having “quiescent disease” – basically a remission or period of inactivity. That’s pretty great, but instead of using this drug, how about a diet that induces most of the same effects as metformin, but also does a ton of other beneficial things for the body? (And maybe even eliminates the need for metformin in the first place?) 25% of subjects showed no improvement, but remember, there was no diet change; only metformin. I wonder how they might have fared on a very low-carb diet. ;-)

Some of the studies employing diabetes drugs for skin problems have shown little to no improvement in the conditions. What a shock…Most of these drugs are intended to improve blood glucose levels, some of them by actually increasing insulin levels. This is exactly what you would NOT want in these situations. If you don’t address the fundamental cause—chronically elevated insulin—then you can’t expect to have all that big an impact on the problem. (This is why so many people with T2D get put on more and more medications at ever increasing doses, yet continue to get worse. Zero addressing of the actual problem.

What this entire post points to is that dermatologists are in a unique position to identify chronic hyperinsulinemia in its early stages, long before it progresses to full-blown metabolic syndrome, type 2 diabetes or cardiovascular disease. One paper states this in no uncertain terms: “Being aware of such clinical signs and the underlying systemic disorders may facilitate earlier diagnoses, thereby permitting earlier of therapy initiation and prevention of long-term sequelae. In this process, dermatologists are key figures in the early detection of MetS and its clinical manifestations.” 

If these cutaneous issues are early warning signs of MetSyn, doctors can inform their patients that elevated insulin is driving their skin issues and that there’s something they can do about it, such as adopting a low-carb or ketogenic diet. Just as erectile dysfunction of otherwise unknown origin may be the first clinical sign of metabolic disease, these skin problems can be the canary in the coalmine indicating pathological processes going on internally, particularly in people with few other biomarkers that would indicate this nefarious metabolic situation.

Fasting insulin and HOMA-IR testing should be routine in a dermatologist’s office. These doctors can offer patients an endless array of creams and ointments, laser therapies and light treatments, but perhaps patients might be better helped not by what they put on their skin, but by addressing what’s happening internally, inside and underneath the skin. Visible manifestations of underlying hormonal imbalances can’t be permanently corrected by topical application of medicinal products. The root cause must be addressed: if the driving factor is a hormone imbalance, then the hormone imbalance needs to be corrected.

Ketogenic Diets for Skin Issues

I’m not aware of any clinical research specifically studying ketogenic diets for skin problems. (The studies I covered in this post were lower glycemic diets but not truly low carb.) But anecdotal reports from living, breathing humans who’ve done low-carb, ketogenic, or carnivore diets tell us that yes, these diets are pretty great for helping to improve skin appearance. Here are a handful of such accounts from Twitter – acne, cystic acne, butterfly rash from lupus:

Perhaps you have your own “success story” of a noticeable improvement in your skin after adopting one of these ways of eating. If so, share it in the comments!

P.S. I’m considering writing a small e-book about this…keto for skin! This is a huge topic, and no one’s writing about it!

Disclaimer: Amy Berger, MS, CNS, NTP, is not a physician and Tuit Nutrition, LLC, is not a medical practice. The information contained on this site is not intended to diagnose, treat, cure, or prevent any medical condition and is not to be used as a substitute for the care and guidance of a physician. Links in this post and all others may direct you to, where I will receive a small amount of the purchase price of any items you buy through my affiliate links.


  1. How about wrinkles? Maybe the book will include a chapter about the effects of insulin or high sugar on collagen.

  2. Wow! It might take me a while to think about all the implications and follow all the links, but THANKS! This was a massive amount of work and research, and I'm glad you're back.

  3. Thanks Amy! Yep, my Rosacea which was treated for many years with antibiotics and all kinds of topicals pretty much disappeared when I went Keto/low carb and has maintained for several years.
    Now wondering if onset of lichen sclerosus in my 30's was associated with high insulin levels? (had a ton of skin tags back then....)

  4. This is some interesting stuff! Personally it lends weight to my belief that my problem is insulin insufficiency - the only thing on your list that I suffered from is a small patch of psoriasis on one elbow and occasionally on the other one.

    All my skin problems resulted from high glucose - skin, eye, sinus, gum and fungal infections, and I caught every cold, flu and "food poisoning" bug going as the glucose killed my immune system. None of this happens during the last fifteen low carb/keto years. I wish I'd know earlier.

    There was a recent TV news item on endometriosis. The family interviewed were fat and looked like they had PCOS which makes me wonder if insulin could be implicated in endometriosis also, making tissue grow where it shouldn't?

  5. I am glad I read this blog, Amy, since I listened to the video a couple of days ago. Good to have time to ponder and reflect on this topic. It certainly has applied to my own situation, as I've had skin tags and always wondered how to get rid of them. Following low carb/keto for a year now, and seeing much clearer skin, and shrinking skin tags, which makes me happy. I am 71, and wish I'd known about the insulin situation 20 years ago, but better late than never :) ...

  6. I'm 82 and diagnosed with type 2 diabetes 30+ years ago and just found a person who explains what I have been thinking for years!.
    Y O U !

  7. I had skin tags in the typical spots. Doing carnivore since December 2019 (mostly beef, some eggs, very minimal dairy). The old typical skin tags disappeared - but recently multiple (10+) smaller new ones appeared in new spots. Thoughts?

    1. I can't really speculate without knowing anything else about your situation. Might be best to see a doctor if you're concerned, just to rule out anything problematic. (Has anything at all changed in your diet lately? Anything about your physical environment?)

    2. I don't think you need anything out of the ordinary. Do be sure to request fasting insulin, though -- that is not a standard test that is typically run, so if you want to have it, you have to specify that you want it.