Dietary factors linked to type 1

The timing of dietary exposures, as well as genetic background, may be important in the development of type 1 diabetes. The "up" arrows indicate associations with an increased risk of type 1, and the "down" arrows with a decreased risk.

Source: Slide courtesy of Dr. Jill Norris, Colorado School of Public Health.

The Details

Studies on Diabetes/Obesity and:


Links Between Nutrition and Diabetes/Obesity

Researchers have examined numerous dietary factors in relation to the development of diabetes and obesity. This page focuses on type 1 diabetes, but also includes some information on type 2 and obesity (a full examination of the dietary and nutritional factors linked to type 2 and obesity is beyond the scope of this webpage; for info see this review on prevention of type 2 via nutrition for example, Uusitupa et al. 2019). Dietary factors are either discussed below, or on other pages, such as Wheat and Dairy, Breastfeeding, Vitamin D Deficiency, or Nitrate and Nitrite. The gut ties many dietary factors factors together, and information on gut microbiota, inflammation, and permeability are on the Diet and the Gut page.

A number of nutritional factors may influence the development of type 1 diabetes or type 1-related autoimmunity. Omega-3 fatty acids may be protective against type 1 diabetes, but more studies would be necessary to confirm this finding. Eating high glycemic-index foods may accelerate the progression of type 1 diabetes, but this association should also be confirmed. Taking anti-oxidant supplements does not appear to reduce the risk of type 1 diabetes, but it is possible that a diet high in anti-oxidants may still be protective.

Timing of food introduction is also important. It seems that introducing food before 3 months of age is problematic, perhaps because the intestine is still immature and unable to handle these foods. 

There is a trial that is open to enrollment to see if vitamin D and omega 3 supplements can help prevent the development of type 1 diabetes (Ricordi et al. 2019). For information on enrollment, see the grassroots health D*Action project, or

Note that diet is linked to environmental chemical exposure: many chemicals enter our bodies via foods, such as pesticides (on or in the food), food packaging materials that leach out of packages into the food, or persistent chemicals that accumulate in the food chain. Numerous diets are associated with lower chemical exposure levels, such as: organic (Oates et al. 2014; Lu et al. 2006), fresh foods (limiting packaging) (Rudel et al. 2011), vegetarian (Ji et al. 2010), vegan (Arguin et al. 2010 Schecter et al. 2001), and the WHO recommended diet (Ax et al. 2015). Your nutrition can also impact the toxicity of environmental chemicals in your body (Hoffman et al. 2017); certain dietary components, such as omega 3 fatty acids, can counteract pollution-induced inflammation, for example (Hennig et al. 2018) or chemical-induced autoimmunity (Benvenga et al. 2020).

Note also that the effects of diet can vary, depending on the person. A large study that measured glucose responses to consistent meals found highly variable responses by person (Zeevi et al. 2015). In other words, your diabetes may vary!

The Details

Omega-3 Fatty Acids

Type 1 Diabetes

Dr. Jill Norris

Listen to Dr. Jill Norris discuss dietary factors in type 1 diabetes on the call, Type 1 Diabetes and the Environment, sponsored by the Collaborative on Health and the Environment (2014).
A trial of pregnant women with type 1 diabetes found that those given supplements of omega-3 fatty acids showed an increase in beta cell function over the course of the pregnancy (as measured by fasting c-peptide levels), as compared to those who did not take supplements (Horvaticek et al. 2017)! Meanwhile, the fatty acid levels of infants were associated with the later development of type 1 diabetes-related autoimmunity. Specifically, fatty acids derived from fish may be protective, as are those consumed via breastmilk as well. Higher omega-6 to omega-3 ratios were associated with an increased risk (Niinistö et al. 2017).

Norris et al. (2007) found that dietary intake of omega-3 fatty acids, found in fish, flax seeds, walnuts, soy, canola, and greens, is protective against the development of type 1 diabetes-related autoantibodies in U.S. children at genetic risk of type 1 diabetes. Omega-3s can reduce inflammation, and the lack of omega-3s in Western diets may predispose people to inflammation. Yet the same authors later found that omega-3 levels were not associated with later development of type 1 in these children (Miller et al. 2011). So, it is possible that omega-3s may be protective against type 1 autoantibody development, but be less significant later in the disease process. In a subsequent study, the authors found that genetic factors may be one factor affecting this process (Norris et al. 2014).

A large prospective study from 8 countries in Europe found that a higher fish consumption or higher levels of omega 3s in blood was protective against adult-onset diabetes in people who tested positive for autoimmunity (Löfvenborg et al. 2020). A longitudinal study from Finland found that levels of omega 3s in blood in early life, however, were not consistently associated with islet autoimmunity development. The average proportions of other fatty acids were, however, including pentadecanoic acid, heptadecanoic acid, stearic acid, and conjugated linoleic acid (Hakola et al. 2021). In the large, prospective, international TEDDY study, higher omega 3 fatty acids during infancy and higher conjugated linoleic acid in childhood were associated with a lower risk of islet autoimmunity, and some saturated fatty acids were linked to an increased risk (Niinistö et al. 2021).

An earlier study of the same children found that the mother's dietary intake of omega-3 fatty acids during pregnancy did not affect the risk of autoimmunity in her children (Fronczak et al. 2003). Another study from Norway also found that the mother's blood levels of fatty acids was not associated with the risk of type 1 diabetes in her children (Sørensen et al. 2012). And, a study from Finland also found no link between the mother's dietary intake of fatty acids during pregnancy and later type 1 diabetes development in her children (Niinistö et al. 2014). A meta-analysis found that omega-3 or omega-6 fatty acid supplementation in children did not affect the overall risk of preclinical or clinical type 1 diabetes, although omega-3 fatty acid intake in early life might reduce the risk of preclinical type 1 (Liu et al. 2018). The international TEDDY study found no difference in offspring autoimmunity risk between those whose mothers took vitamin D during pregnancy and those that did not (Silvis et al. 2019), nor did they find any associations between the mother's diet in late pregnancy and subsequent onset of autoimmunity or type 1 diabetes (Johnson et al. 2021). In children with new-onset type 1 diabetes who ate a Mediterranean diet and took vitamin D, omega 3s helped reduce the demand for insulin (Cadario et al. 2019).

Cod liver oil, however, taken during pregnancy, has been associated with a reduced risk of type 1 diabetes in offspring. Both omega-3 fatty acids and vitamin D are present in this oil, and either or both may play a role (Stene et al. 2000). Cod liver oil taken during the first year of life is also associated with a lower risk of type 1 diabetes (Stene et al. 2003).

Virtanen et al. (2010) found that the fatty acids associated with milk and ruminant meat fat consumption were associated with an increased risk of type 1 related autoimmunity In Finland. Linoleic acid, however, was associated with lower levels of autoimmunity, in children genetically at risk of type 1 diabetes. A more recent study from Finland found that in children with a genetic risk of type 1 diabetes, meat consumption was associated with a slightly higher risk and fish consumption with a slightly lower risk of both islet autoimmunity and type 1 diabetes, up to age 15 (Syrjälä et al. 2019). In adults, consumption of processed (but not unprocessed) red meat was associated with a higher risk of Latent Autoimmune Diabetes in Adults (LADA), especially in people with a genetic risk of type 1 diabetes (Löfvenborg et al. 2020).

In U.S. children who already had developed both antibodies and type 1 diabetes, omega-3 intake was associated with higher residual insulin production (C-peptide levels). This implies that omega-3s may help preserve the functioning of the insulin-producing pancreatic beta cells over time, even in people who already have type 1 diabetes (Mayer-Davis et al. 2013). 

In a case study, a 14 year old boy with new-onset type 1 diabetes was treated with high-doses of omega 3s and vitamin D. These seem to have helped preserve what was left of his beta cell function (Baidal et al. 2016). In another case study, An 8 year old child took high-dose vitamin D and omega-3 fatty acids, and shows near-normal blood sugar levels 1.5 years after diagnosis, only taking 1.5-2 units of insulin per day (Cadario et al. 2017). The same authors found success with at least one other person as well (Cadario et al. 2018).

A Swedish study found that higher fatty fish consumption was associated with a reduced risk of Latent Autoimmune Diabetes in Adults (LADA, i.e., type 1 diabetes in adults), but did not affect the risk of type 2 diabetes. Similar results were found with omega-3 intake and fish oil supplementation (Löfvenborg et al. 2014).

The DAISY study of type 1 diabetes found that oxylipins, which are derived from polyunsaturated fatty acids, are associated with type 1 development (some increased risk, some decreased) (Buckner et al. 2021).

Omega 3 and other polyunsaturated fatty acids may also protect against the ability of environmental chemicals to cause inflammation that could lead to autoimmunity, as well as protect against organ damage (Moloudizargari et al. 2019). Omega 3s may be useful for both prevention and treatment of numerous other autoimmune diseases as well (Li et al. 2019).

Fish oil may also be helpful for treatment of diabetic neuropathy (Lewis et al. 2017Mirhashemi et al. 2016; Yorek 2018).

Metabolic Syndrome, Type 2 Diabetes, Gestational Diabetes, and Obesity

A meta-analysis of 13 randomized controlled trials found that fish oil supplements reduced insulin resistance in children (Hou et al. 2020).

A group of people with metabolic syndrome (a group of conditions common in people with diabetes) were given omega-3 fatty acid supplements or a placebo for six months. Those taking the supplements were found to have lower markers of autoimmunity and inflammation, as well as more weight loss, compared to people who did not take the supplements (Ebrahimi et al. 2009). Another trial, of people with type 2 diabetes, found that omega 3s improved lipid levels and inflammation (Mazaherioun et al. 2017). However, a systematic review and meta-analysis of 83 randomized controlled trials concluded that omega-3s, omega-6s, or total poly unsaturated fatty acids had little or no effect on the prevention or treatment of type 2 diabetes (Brown et al. 2019).

A large, worldwide meta-analysis of 28 studies found that in men, fish consumption did not affect the risk of developing type 2 diabetes. In women, fish consumption increased risk (Pastorino et al. 2021). I would guess this is probably due to the chemicals in the fish.

Prospective, randomized, double-blind, placebo-controlled clinical trials from Iran found that women with gestational diabetes who took omega-3s had better outcomes than those who did not, including better glucose levels, triglycerides, LDL-cholesterol, and HDL-cholesterol concentrations (Jamilian et al. 2020Jamilian et al. 2017Jamilian et al. 2016; Rajabi-Naeeni et al. 2020Taghizadeh et al. 2016). A trial from Finland found that neither fish oil nor probiotics nor their combination lowered the risk of gestational diabetes in overweight or obese pregnant women (Pellonperä et al. 2019). A review of studies on fish oil supplements and gestational diabetes (which, for some reason, only included 2 studies...) found that DHA-enriched fish oil had no effect on gestational diabetes prevention, but did reduce insulin resistance in women with gestational diabetes. Not surprisingly, they conclude there is not enough evidence to support or refute the use of fish oil supplements during pregnancy (Ostadrahimi et al. 2016). A meta-analysis of four randomized controlled trials assessing the influence of omega-3 fatty acid and vitamins in gestational diabetes found that omega-3 fatty acid and vitamin E or D was associated with lower fasting plasma glucose, insulin resistance, and triglyceride levels (Li et al. 2020).

Adequate intake of omega-3s during pregnancy may also decrease the risk of obesity in the offspring. Higher levels of omega-6 fatty acids in relation to omega-3s in umbilical cord blood has been associated with higher obesity in children at age 3 (Donahue et al. 2011). A trial found that infants who received omega-3s had lower insulin resistance and lower waist circumference at age 5 (See et al. 2018). However, a very large study from around the world found that high fish intake during pregnancy (more than 3x per week) was associated with an increased risk of childhood obesity (Stratatkis et al. 2016). Whether this could be due to the chemicals in fish, I don't know... Another study from Mexico found that higher levels of omega-3s and omega-6s during pregnancy were associated with lower height (but not other metabolic measures, e.g., not glucose levels etc.) in offspring around the time of puberty (Al-Hinai et al. 2018). Again, I don't know what would explain this finding.

Omega 3 fatty acids may also be protective against type 2 diabetes, according to a long-term study of Finnish men, even when mercury levels were taken into account (Virtanen et al. 2014). A randomized, controlled trial is now investigating whether omega 3s or curcumin (found in the spice turmeric) can prevent type 2 diabetes (Thota et al. 2016).

Gestational Diabetes

In China, a trial of women with gestational diabetes who took 40,000 IU of vitamin D and 8,000 mg of omega 3 fatty acids twice a day for 6 weeks had lower fasting blood glucose, insulin, insulin resistance, triglycerides, total cholesterol, LDL, and VLDL cholesterol levels and beta cell function was markedly improved (Huang et al. 2021).

Diabetes Complications

A meta-analysis of 10 randomized controlled trials of people with type 1 and 2 diabetes found that 24 weeks of omega 3 supplements improved markers of kidney disease (Chewcharat et al. 2020). Omega-3 fatty acid supplementation for 12 weeks in people with diabetic nephropathy (kidney disease) had favorable effects on insulin levels, triglycerides and cholesterol (Soleimani et al. 2017). Supplementation also helped reduce triglyceride levels in people with type 2 diabetes (Fayh et al. 2018). A randomized controlled U.S. trial found that omega-3 supplementation for 5 years (with and without vitamin D) in people with type 2 diabetes did not help prevent a decline in kidney function (de Boer et al. 2019).

While some studies have found no benefit (e.g., ASCEND Study Collaborative Group 2018), others argue that the beneficial cardiovascular effects of omega-3 supplementation in people with diabetes are likely only found with daily doses above 2000 mg (Tenenbaum and Fisman 2018). 

Laboratory Studies

In laboratory animals, flax oil during pregnancy protects the offspring from the negative effects of maternal diabetes and high blood sugar in the womb (Correia-Santos et al. 2015Correia-Santos et al. 2014). A proper fatty acid ratio during pregnancy and lactation can prevent diabetes in the offspring of NOD mice, an animal model of type 1 diabetes (Kagohashi and Otani 2015). Another study of NOD mice also shows that omega 3s drastically reduce the development of diabetes, blocked autoimmunity, and normalized glucose levels (Bi et al. 2017). Omega-3s also protect beta cells from the effects of a high-fat diet or chemicals that cause beta cell dysfunction and death (Wang et al. 2015). In rats, animals who received both vitamin D and omega 3 fatty acids had lower blood sugar levels than untreated rats after islet transplantation (Gurol et al. 2016).

Chemicals and Omega-3s

The presence of environmental contaminants in food may also play a role in the effects of nutritional factors. Some contaminants may interfere with the beneficial effects of foods. For example, in an animal study linking insulin resistance to persistent organic pollutants, the researchers concluded that beneficial aspects of omega-3 fatty acids in salmon oil could not counteract the harmful effects caused by the persistent organic pollutants in that oil (Ruzzin et al. 2010). In rats, fish oil contaminated with POPs-- despite overall beneficial effects-- led to lower antioxidant capacity and more oxidative stress as compared to uncontaminated oil (Hong et al. 2015). Whether the overall effects are positive or negative, it does appear that POPs reduce the healthy effects of fish oil.

DHA, one of the omega-3 fatty acids found in fish, appear to protect the immune system (in animals) from low-level mercury exposure, and may reduce the risk of autoimmune diseases associated with that exposure (Gill et al. 2015). A pilot study found that DHA supplementation is safe in infants with genetic risk of type 1 diabetes (Chase et al. 2015).

Fish is one source of omega-3 fatty acids, but according to an editorial in the American Journal of Clinical Nutrition (AJCN), it may be better to rely on plant-based sources instead (Feskens 2011). Studies on fish consumption and type 2 diabetes are inconsistent: some show that higher dietary intake of omega 3s decreases the risk of type 2, some show no connection, and some even show that higher fish consumption increases the risk of type 2 diabetes (Djousse et al. 2011; Villegas et al. 2011). It may be that the chemicals in fish can explain these inconsistencies. A study shows that plant-based omega 3s have different effects than marine-based omega 3s in relation to type 2 diabetes (Brostow et al. 2011), possibly due to the contaminants present in fish. The FDA now recommends that pregnant and breastfeeding women eat fish that are low in chemicals such as mercury (Wenstrom 2014).

Perfluoroalkyl substances (PFASs, e.g., found in Scotchguard and teflon) are associated with lower essential fatty acids in pregnant women, vital to fetal growth (the female babies also weighed less) (Kishi et al. 2015). Whether this association is causal is not known, but would be alarming if so. In fact, animal studies are attempting to figure out mechanisms by which PFASs affect fatty acids in the fetus (e.g., Lee et al. 2015).

Other Fats

Flax seeds are one source of omega-3 fatty acids. Refined fish oil is another source, but unrefined fish oil may contain higher levels of chemicals.
There is confusion over the links between saturated fats and health. While the general opinion is that saturated fats are "bad" for your health, the research is not so clear. Using heart disease as an example, recent studies have found no link between dietary saturated fats and heart disease. However, higher levels of saturated fat in your blood are related to an increased risk of heart disease, as well as type 2 diabetes. We might assume that the saturated fat in our diet would influence the saturated fat in our blood, but it appears that dietary carbohydrates may be even more important than dietary fats in determining blood fat levels. People with metabolic syndrome who consumed various diets, high or low carb and high or low saturated fat, showed no relation between fat consumption and blood fat levels, but did have a relation between a high carb diet and high fat in their blood (Volk et al. 2014).

All saturated fatty acids do not necessarily increase the risk of type 2 diabetes. Some types are associated with an increased risk, but others with a decreased risk (Forouhi et al. 2014). Rodents fed cream from pasture-raised cows had better metabolic outcomes than those fed regular cream, for example (Benoit et al. 2014). An interesting Swedish study found that high-fat dairy products (but not low-fat) actually were associated with a decreased risk of type 2 diabetes. Both high and low fat meat products were associated with an increased risk (Ericson et al. 2015). (Meat has been linked to type 2 diabetes in other studies as well; what exactly in the meat is problematic is still being worked out (Kim et al. 2015). Organic meat appears to be higher in beneficial fatty acids than commercial meat (Średnicka-Tober et al. 2016).

A higher consumption of omega 6 fatty acids has been linked to a lower risk of type 2 diabetes, in 3 large prospective studies from the U.S. (Zong et al. 2019), and in a study from Finland (Yary et al. 2016) Actually, an analysis of the data from 20 prospective studies on omega 6s found that linoleic acid (found in nuts and vegetable oils) has long-term benefits for the prevention of type 2 diabetes (Wu et al. 2017). In general, nut consumption shows some positive trends for reducing the risk of type 2 diabetes (Ntzouvani et al. 2019).

Higher levels of certain trans fats are associated with a higher risk of diabetes in U.S. adults, as well as higher glucose and insulin levels, higher insulin resistance, and higher HbA1c (a measure of long-term blood glucose control) (Liu et al. 2018). Mothers who consumed higher levels of trans fats had an increased risk of excess body fat, and so did their breastfed infants (Anderson et al. 2010).

Short-chain fatty acids may be protective against type 1 diabetes, in findings from the TEDDY study, a large international long-term study (Vatanen et al. 2018). Virtanen et al. (2011) found a weak protective effect of butter and low-fat margarine eaten during pregnancy and the development of type 1 related autoimmunity in the offspring.

Laboratory Studies

In rodents, feeding them a high-fat diet is an easy way to make them obese. Saturated fats can cause inflammation, which in turn causes insulin resistance. Meanwhile, unsaturated fats do not (Wen et al. 2011).

Can the effects of a high fat diet be passed down to subsequent generations? In animal studies, a high-fat diet that causes obesity in mothers can affect the metabolism and weight of her offspring. But what about a high fat diet in fathers? In one study, the female offspring of heavier father rats (fed a high-fat diet) had defects in their insulin and glucose levels, like their fathers. Unlike their fathers, they were not heavier than the controls (Ng et al. 2010). Other researchers fed mice a high fat diet with fat composition similar to a standard Western diet, and then bred them and fed them the same diet for multiple generations. Over four generations, the offspring became gradually heavier, and developed higher insulin levels, despite not eating more calories. The diet was associated with changes in gene expression (Massiera et al. 2010).

Different researchers have also found similar results-- that a high-fat diet in pregnant mice resulted in two generations of their offspring having a higher body size and increased insulin resistance. In the third generation, however, only the females developed the high body size, and only via their fathers (not mothers). Again, the changes were found to be associated with gene expression (Dunn and Bale 2011).

Another fatty acid found in many foods, butyric acid, can cause detrimental effects on liver cells and pancreatic beta cells (alone, and in conjunction with arsenic). The diabetes medication metformin may be protective against these effects (Ahangarpour et al. 2017).

Sweeteners, Carbs, and The Glycemic Index

Carbohydrates and the Glycemic Index and Type 1 Diabetes

I'm not going to get into all the discussion about low-carb diets and type 1 diabetes management here-- it seems to me that's an individual choice. Yes my blood sugar is better on a low-carb diet, but it's hard to stick with consistently over the long-term. Can a low-carb diet help prevent type 1 diabetes? Perhaps. There is a lot of anecdotal evidence, and research is starting to catch up. For example, one study found, "A low-carbohydrate high-fat diet initiated promptly after diagnosis provides clinical remission in three patients with type 1 diabetes" (Bouillet et al. 2019). ("Clinical remission" is also known as the "honeymoon" period, where beta cell function tends to improve for a period of time after diagnosis and usually after insulin is initiated).

Here is a cool study that looked at people without diabetes, and how their blood glucose responded to various meals and amounts of carbohydrates. The results showed that there is a lot of individual variation: some people's blood sugar did not rise after meals (or only rose slightly), and others rose quite a lot (almost up to 200 mg/dl). The authors looked at the gut microbiome and determined that an individual's microbiome helps predict how a person responds to food, and that other characteristics can also be used to help predict responses, beyond just carbohydrate or calorie content (Mendes-Soares et al. 2019).

The glycemic index is a measurement of how high a certain food raises blood glucose levels after it is eaten. Foods that have a high glycemic index will cause blood glucose to rise more, triggering insulin production (in people who still produce insulin), then leading to falling blood glucose levels. One prospective study has found that a higher glycemic index diet leads to a faster progression to type 1 diabetes. The group of people on this diet, however, did not have higher levels of autoantibodies, showing that the diet may affect disease progression but not disease initiation. The mechanisms involved may include oxidative stress, caused by high blood glucose levels after meals, or perhaps insulin resistance. Whatever the mechanism, a high glycemic index diet may place additional stresses on beta cells that are already under an autoimmune attack (Lamb et al. 2008).

Note, however, that even low glycemic foods could be potentially problematic. Researchers found that dozens of types of low glycemic foods react with islets and autoimmune targets related to type 1 diabetes (Kharrazian et al. 2017).

Sugar and Type 1 Diabetes

Is sugar consumption linked to type 1 diabetes development? Most people think it is not, but some recent research may start to question that assumption. Babies introduced to sugar-sweetened beverages (including fruit juice) during the first year of life developed type 1 diabetes 5 months faster than children who started drinking these beverages at an older age (Crume et al. 2014). A long-term study from Colorado found that among those who had already developed autoimmunity or were of high genetic risk, sugar intake increased the risk of developing type 1 diabetes. Sugar intake was not associated with the original development of autoimmunity however, implying that sugar may exacerbate the later stage development of type 1 (Lamb et al. 2015). This study also found that children with autoimmunity who had a lower intake of linoleic acid, niacin, and riboflavin, and a higher intake in total sugars, had increased risk for progressing to type 1 diabetes (Johnson et al. 2020).

Daily intake of more than 2 sweetened beverages per day is associated with the development of LADA, Latent Autoimmune Diabetes in Adults (essentially type 1 but in adults), as well as type 2 diabetes (Löfvenborg et al. 2016), which may depend partially on genetic background (Löfvenborg et al. 2019).

Sugar and Type 2 Diabetes, Obesity, and Metabolic Syndrome

The consumption of sugar-sweetened beverages has been associated with type 2 diabetes, obesity, gestational diabetes, and metabolic syndrome. A meta-analysis of a 11 prospective studies (of over 300,000 people) found that those who consumed 1-2 sweetened beverages per day had a 26% greater risk of developing type 2 diabetes than those who consumed fewer than one serving per month. The risk was 20% higher for developing metabolic syndrome. Sugar-sweetened beverages include soft drinks, fruit drinks, iced tea, and energy/vitamin water drinks (Malik et al. 2010). Another meta-analysis found that sugar-sweetened beverages are associated with higher fasting glucose and insulin levels (McKeown et al. 2018). A meta-analysis of 17 articles found that consumption of total sweetened beverages, sugar-sweetened beverages, and artificially sweetened beverages was associated with an increased risk of metabolic syndrome (Zhang et al. 2020).

Long-term studies have also found links between sweetened beverages and type 2 diabetes (Hirahatake et al. 2019; O'Connor et al. 2015). The intake of these beverages is also associated with coronary heart disease (Malik 2017). U.S. adolescents with higher added sugar intake had a higher risk of metabolic syndrome, no matter their BMI, total calorie intake, or physical activity level (Rodríguez et al. 2016). U.S. adults who drank more sugar-sweetened beverages (or fruit juice) over a 4-year period had an increased risk of developing type 2 diabetes, and replacing those drinks with coffee, tea, or water reduced the risk (while artificially sweetened beverages increased risk, see the section below for more on that topic (Drouin-Chartier et al. 2019).

The consumption of sugar-sweetened soft drinks (but not diet soft drinks) prior to pregnancy is associated with an increased risk of gestational diabetes (Donazar-Ezcurra et al. 2018). In the U.S., in the offspring of women who had gestational diabetes, breastfeeding was protective against obesity, but only if offspring have a lower intake of sugar-sweetened beverages (breastfeeding is protective even in children who drink sugar-sweetened beverages if their mothers did not have gestational diabetes) (Vandyousefi et al. 2019).

In laboratory animals, a high-sugar diet during pregnancy and lactation led to metabolic effects in two subsequent generations. In grand-daughter mice, it led to higher fasting blood glucose levels and increased their food intake, while in grand-son mice, it increased brown fat and HDL cholesterol levels (Školníková et al. 2020).

What About Sugar?

For the latest science on the effects of sugar, fructose, HFCS, sugar-sweetened beverages on health-- including diabetes and obesity, see:, a website developed and maintained by scientific researchers in the field.
High-fructose corn syrup is another sweetener linked to obesity. Rats given access to high-fructose corn syrup gained more weight than those given access to sucrose, despite eating the same number of calories (Bocarsly et al. 2010). Goran et al. (2013) review the evidence that high fructose exposure during development (in the womb and during infancy) can "act as an obesogen by affecting lifelong neuroendocrine function, appetite control, feeding behaviour, adipogenesis, fat distribution and metabolic systems. These changes ultimately favour the long-term development of obesity and associated metabolic risk." In lab animals, fructose causes high blood sugar, glucose intolerance, decreased insulin secretion, and increased glucagon secretion (showing it affects both beta cells and alpha cells in the pancreas) (Asghar et al. 2017). A high-sugar diet also increases obesity, glucose intolerance, insulin resistance, and other metabolic changes in rats, despite the rats eating fewer calories on this diet (Oliveira et al. 2020).

A pretty convincing intervention study found that replacing fructose/sugars with other starchy carbohydrates (like bread) improved metabolism in children in just 10 days, including lower blood pressure, weight, triglycerides, and LDL cholesterol. Glucose tolerance improved, as did insulin levels. Sugar levels were reduced from 28% of calories (!) to 10%, so not even eliminated, just reduced. The children in this study were obese or had metabolic syndrome (Lustig et al. 2016). Additional studies by the same authors found that the children had improved markers of cardiovascular risk (Gugliucci et al. 2016), and developed less liver fat, less visceral fat, and better insulin function (Schwarz et al. 2017). 

Some researchers suggest that reducing added sugar by 20% in the U.S. would yield $10 billion in lower medical costs by 2035, and significantly lower the incidence of type 2 diabetes, heart disease, liver disease, and obesity (Vreman et al. 2017).

Artificial Sweeteners and Type 1 and Type 2 Diabetes and Obesity

A systematic review of human and animal studies found that overall, using low-energy sweeteners in place of sugar tends to lead to lower caloric intake and lower body weight in both children and adults (Rogers et al. 2016), but another review finds that the evidence is not entirely clear (Young et al. 2019). Interestingly, consumption of artificial sweeteners is associated with higher body weight and type 2 diabetes, but when used in trials for weight loss, they may help aid weight loss (Sylvetsky and Rother, 2018). A long-term study found that greater consumption of artificially-sweetened beverages is associated with an increased risk of type 2 diabetes after 30 years of follow-up (as was consumption of sugar-sweetened beverages) (Hirahatake et al. 2019).
There also may be other health effects of artificial sweeteners, including effects on the gut microbiome, activating sweet taste receptors, and influencing hormones (Rother et al. 2018). For example, chemical artificial sweeteners can increase glucose intolerance by changing the gut microbiota in mice-- and humans as well (Suez et al. 2014). In humans, some gut microbiota changes linked to artificial sweeteners are not beneficial (Farup et al. 2019). In mice, artificial sweeteners can cause body weight gain (Bian et al. 2017). In mice, developmental exposure to small amounts of artificial sweeteners had drastic effects on the offspring's gut microbiome (Olivier-Van Stichelen et al. 2019); artificial sweeteners during development also affect offspring metabolism in mice (Plows et al. 2020).  In humans, maternal consumption of artificially sweetened beverages during pregnancy is associated with a higher risk of offspring obesity/overweight at age 7 (Zhu et al. 2017).

In people with type 1 diabetes and healthy controls, artificially sweetened beverages affect levels of certain gut hormones, but not in people with type 2 diabetes. It's not entirely clear what the significance of this is for health, as these hormones are generally thought to be beneficial for diabetes and obesity (Brown et al. 2012).

I tend to use stevia as a sweetener, which is a plant-based compound. In animals, it can affect gut microbiota, but didn't affect glucose metabolism or weight gain (Nettleton et al. 2019). It may have other health effects, but in general seems mostly safe (Samuel et al. 2018).

Fruit, Vegetables, and Fiber

One study found that eating vegetables daily during pregnancy reduced the risk of a child's developing type 1-associated autoimmunity (Brekke and Ludvigsson 2010). Virtanen et al. (2011) found a weak protective effect of a few foods eaten during pregnancy and the development of type 1 related autoimmunity in the offspring, including berries; most foods showed no association.

According to a large international longitudinal study (the TEDDY study), dietary fiber in childhood is not associated with the development of autoimmunity or type 1 diabetes (Beyerlein et al. 2015). However, the Finnish Type 1 Diabetes Prediction and Prevention Study, another prospective birth cohort of children with genetic susceptibility to type 1 diabetes, found that those who had a high intake of dietary fiber (as well as oats and gluten) during childhood (up to age 6) had an increased risk of islet autoimmunity (Hakola et al. 2019).

Virtanen et al. (2006 and 2011) found that introducing fruits, berries, and roots early (around 3-4 months of age) was associated with development of type 1 diabetes-related autoimmunity in genetically at-risk children. It is likely that introducing any food before 3 months of age could be problematic, due to an immature gut.

In mice, a high fiber diet produced fatty-acids that were protective against type 1 diabetes development (Mariño et al. 2017).

Fiber is also linked to a reduced risk of type 2 diabetes (InterAct Consortium 2015).

Fresh fruit consumption, meanwhile, has been linked to a lower risk of diabetes in Chinese adults, as well as a lower risk of death or complications in those with diabetes (Du et al. 2017).

Children who ate a diet high in vegetables and grains, and low in refined cereals and sweet beverages during early childhood had a reduced risk of celiac disease (which is common in people with type 1 diabetes) (Barroso et al. 2018). 

Some researchers argue that the Mediterranean diet, which is high in fiber and omega 3s, might help prevent the development of type 1 diabetes, in part via its effect on the gut microbiota (Calabrese et al. 2021).

Low Blood Sugar from Lychee Fruit-- Huh?

This isn't diabetes but it is pretty interesting. The article, Dangerous Fruit: Mystery of Deadly Outbreaks in India Is Solved (by Ellen Barry, 2017, New York Times), describes how kids were dying -- with very low blood sugar-- and it turns out it was caused by eating lychee (litchi) fruit on an empty stomach. Apparently lychee fruit contain chemicals (hypoglycin) that inhibit the body's ability to synthesize glucose, causing low blood sugar, and in some cases leading to death (Shrivastava et al. 2017). A follow-up study identifies the chemical methylenecyclopropyl glycine as the source of the problem, and rules out pesticides on the fruit as being the culprit (Asthana et al. 2019).

Coffee, Chocolate, and Alcohol

Some Good News: Coffee

Numerous human and animal studies have found that higher coffee consumption is associated with a lower risk of type 2 diabetes (Akash et al. 2014Carlström and Larsson 2018; Kolb et al. 2021Muley et al. 2012). Coffee is also associated with a lower risk of obesity (Lee et al. 2019). Decaf coffee and caffeinated tea have also been associated with a decreased risk (Huxley et al. 2009), although another study found that only green tea -- not black tea -- was associated with a decreased risk (Iso et al. 2006).

However, coffee consumption may be associated with an increased risk of autoimmunity and type 1 diabetes in adults (latent autoimmune diabetes in adults, LADA) (Löfvenborg et al. 2014)-- at least in people with higher genetic risk of type 1 (Rasouli et al. 2018). Interestingly, coffee has different associations with different autoimmune diseases, in that it is associated with a decreased risk of some and increased risk of others (Sharif et al. 2017). Coffee is also associated with a higher risk of metabolic syndrome in people with type 1 diabetes (Stutz et al. 2018). And, maternal intake of caffeine during pregnancy is associated with an increased risk of obesity in their children (Li et al. 2015). In rats, maternal intake of caffeine during pregnancy reduces beta cell mass, but increased glucose tolerance in adult offspring, especially female offspring (Kou et al. 2017). Virtanen et al. (2011) found a weak protective effect of drinking coffee during pregnancy and the development of type 1 related autoimmunity in the offspring.

More Good News: Chocolate

Chocolate, meanwhile, protects pancreatic beta cells and reduces insulin resistance and high blood sugar, delaying diabetes -- at least in rats (Fernández-Millán et al. 2015; Martin et al. 2016). In a large, long-term study of physicians, the more chocolate eaten (well, up to 2 servings per week), the lower the risk of type 2 diabetes-- although only in younger and normal body-weight men (Matsumoto et al. 2015). Dark chocolate also may help to prevent metabolic syndrome (Leyva-Soto et al. 2018). In pregnant Japanese women, higher chocolate consumption was associated with a lower risk of gestational diabetes as well (Dong et al. 2019).

And even better-- children with type 1 diabetes who ate dark chocolate (25 grams, 2-5 times/week) had better blood sugar control than those who ate milk chocolate or no chocolate (Scaramuzza and Zuccotti, 2015). Adults with type 2 diabetes also benefited from eating dark chocolate (Jafarirad et al. 2018). A review of the use of dark chocolate in people with diabetes finds that it may slowing the progression to type 2 diabetes, lessen insulin resistance, and help prevent cardiovascular complications in people with diabetes (Shah et al. 2017). Fun fact: air pollution makes rats eat more chocolate (da Silveira et al. 2018).

Not As Much Good News: Alcohol/Wine

A systematic review and meta-analysis found that moderate alcohol consumption is associated with a reduced risk of type 2 diabetes, especially in women and non-Asians (Knott et al. 2015). A longitudinal U.S. study found that higher alcohol consumption was associated with a lower risk of diabetes in both women and men, and especially in those with a higher BMI (He et al. 2019). However, some have found that high alcohol consumption in early adulthood increases the risk of type 2 diabetes (Yu et al. 2019), even when the amount consumed is later reduced (Han et al. 2019). And chronic alcohol consumption may be linked to an increased risk of type 2 diabetes (reviewed in Kim et al. 2015). So there may be a non-linear relationship between alcohol and diabetes, which may vary by sex, BMI, race, and more. High alcohol consumption also interacts with heavy metal exposures to increase the risk of type 2 diabetes, at least in occupationally exposed Chinese men (Yang et al. 2019).

As for adult-onset type 1, it's more complicated. Moderate alcohol consumption is associated with a reduced risk of insulin resistance, but a higher risk of autoimmunity. Overall, the risk of adult-onset type 1 was lower with more alcohol consumption, but that only held true in people with low antibody levels (Rasouli et al. 2014). 

Resveratrol, the stuff found in red wine, has been found to help ameliorate the progression of autoimmune diseases in some cases (Oliveira et al. 2017).

Drinking while pregnant may be problematic, however. In rats, while low-level prenatal alcohol exposure did not affect fasting or blood glucose levels after a glucose tolerance test in offspring, there was evidence of insulin resistance in alcohol-exposed male offspring at 6 months of age (Nguyen et al. 2019).

Zinc and Other Trace Elements

While many metals can be toxic at high levels, they can be necessary for life in very low levels (e.g., calcium, iron, etc.). (See the Heavy Metals page for a discussion of metal exposures at higher levels; the levels discussed here are all at very low levels, in the trace level range).

Type 1 Diabetes

A few studies have found that higher zinc levels in drinking water (again, not TOO high-- still in the trace level range), may be protective against type 1 diabetes. For example, Zhao et al. (2001) found that higher levels of zinc and magnesium were associated with lower rates of type 1 diabetes in southwest England. In Norway, a study found that higher zinc levels in water was associated with a lower risk of type 1 diabetes, but the association was not statistically significant (Stene et al. 2002). In Finland, a study found that low zinc levels in drinking water was associated with a higher incidence of type 1 diabetes (Samuelsson et al. 2011). And a study form Canada found that zinc and calcium intakes during the year before diagnosis were marginally protective against type 1 diabetes risk in youth (Benson et al. 2010). In Sardinia, Italy, trace elements like zinc, copper, chromium, manganese in stream sediments were associated with a lower risk of type 1 diabetes (Valera et al. 2015). 

While not all studies have found a link between type 1 and zinc (e.g., Estakhri et al. 2011), overall reviews of the evidence linking zinc to both type 1 and type 2 diabetes-- a link that has been suggested for almost 70 years-- suggests that zinc may be a potential therapy for both diabetes prevention and blood glucose control (Chimienti 2013; Maret 2017; Wang et al. 2020). 

There are also studies of people who already have type 1 diabetes.

A review finds that low levels of magnesium are linked to poor glucose control in people with type 1 diabetes (Rodrigues et al. 2020).

In Taiwan, in people who already have type 1 diabetes, lower zinc levels and higher copper levels were correlated with higher blood glucose levels (a higher HbA1c). In addition, people with type 1 had lower zinc and higher copper levels than those without diabetes (Lin et al. 2014). Greek children and adolescents with type 1 diabetes with lower magnesium levels had a higher HbA1c (Galli-Tsinopoulou et al. 2014). Egyptian children with type 1 diabetes with low magnesium levels have a higher HbA1c and higher cholesterol levels (Shahbah et al. 2016). Another Egyptian study found that children with type 1 diabetes have low levels of various trace elements, including selenium, zinc, magnesium, and copper (Alghobashy et al. 2018).

In Florida, people with long-standing type 1 diabetes had higher copper levels than those without diabetes. While copper is an essential trace element, too much can be harmful. The essential elements manganese, zinc, and selenium were lower in those with type 1 (levels of additional metals also differed; see the Heavy Metals page for info) (Squitti et al. 2019). As this study was cross-sectional, and the people had type 1 for an average of 24 years, it does not really say much about diabetes development. It would be interesting to know, however, if having type 1 affects levels of these metals, or vice versa, or if higher or lower levels of these metals could contribute to disease progression or complications.
A study compared trace metal levels in people with type 1 diabetes, type 2 diabetes, and without diabetes. They found that those with type 1 diabetes had lower levels of chromium, manganese, nickel, lead, and zinc, and those with type 2 diabetes had lower levels of chromium, manganese, and nickel, as compared to people without diabetes (Forte et al. 2013). 

Type 2 Diabetes

A systematic review and meta-analysis of 16 studies found that a moderately high dietary zinc intake could reduce the risk of type 2 diabetes, but that higher levels were associated with an increased risk (Fernández-Cao et al. 2019). Additional studies have also found associations between various trace element levels in people with type 2 diabetes vs those without diabetes (Hansen et al. 2017; Kant et al. 2020Simić et al. 2017; Zhang et al. 2017). In people with type 2 diabetes, trace element levels are linked to lipid levels (Wolide et al. 2017). Low magnesium levels have been linked to pre-diabetes in Dutch adults, for example (Kieboom et al. 2017), and low chromium levels linked to type 2 diabetes and pre-diabetes in Chinese adults (Chen et al. 2017). These minerals may interact with other dietary factors to influence risk. For example, higher magnesium intake is linked to a lower risk of type 2 diabetes, especially in people with a poorer quality diet (low fiber, higher glycemic index) (Hruby et al. 2017). A review finds that deficiencies in zinc, selenium, and copper (which are essential but also toxic at high levels) are linked to type 2 diabetes and insulin resistance, and may also be linked to diabetes complications (Bjørklund et al. 2019). Obesity is also a related factor (Morais et al. 2019).

High iron levels or iron metabolism markers are associated with type 2 diabetes (Huth et al. 2015; Liu et al. 2020Orban et al. 2014), diabetes complications (Mojiminiyi et al. 2008), as well as with insulin resistance in people without diabetes (Krisai et al. 2016).

Laboratory Studies
In test tube studies, zinc deficiency and arsenic exposure, both independently and in combination, adversely affect pancreatic beta cells (Cao et al. 2019).

In rodents, the gasoline additive MTBE interferes with zinc and glucose levels (Saeedi et al. 2017), as well as with gut microbiota and glucose tolerance (Tang et al. 2019). Also in rodents, zinc helps prevent diabetes-induced kidney damage (Yang et al. 2017). Zinc also reduces intestinal inflammation in rats with diabetes (Barman and Srinivasan, 2019).

Exposure During Development

A study from Denmark found that zinc levels at the time of birth were not associated with the later development of type 1 diabetes (Kyvsgaard et al. 2016).

In Norway, iron supplementation during pregnancy was associated with a higher risk of type 1 diabetes in the offspring (Størdal et al. 2018). A study found that higher iron intake (via infant formula or supplements) in the first four months of life was associated with a higher risk of developing type 1 diabetes (Ashraf et al. 2010). Another found higher iron levels in blood around the time of birth was associated with a higher risk of developing type 1 by age 16 (Kyvsgaard et al. 2017). A review found that dietary iron was associated with an increased risk of type 1 diabetes, but not iron in drinking water (Søgaard et al. 2017). On the other hand, a large study from Denmark found that iron supplements in early life reduced the risk of type 1 diabetes, but supplementation during pregnancy did not affect risk (Thorsen et al. 2019).

Gestational Diabetes

Higher iron levels during pregnancy are linked to glucose intolerance in the mother (Zein et al. 2015) and gestational diabetes (Bowers et al. 2016; Fernández-Cao et al. 2017; Khambalia et al. 2016; McElduff 2017; Rawal et al. 2017).

Meanwhile higher calcium intake is associated with a lower risk of gestational diabetes (Osorio-Yáñez et al. 2017).

Nicotinamide and Other Vitamins and Antioxidants

Nicotinamide is a component of vitamin B3 that has been shown to protect against diabetes in animals, and prevent beta cell damage in the laboratory (Gale et al. 2004). Even better, one study found that it prevented the development of type 1 diabetes in children with type 1-associated autoantibodies (Elliott et al. 1996).

On the basis of these and other studies, a large, double-blind, placebo-controlled trial was conducted in Europe, the U.S. and Canada, called the European Nicotinamide Diabetes Intervention Trial (ENDIT). This trial gave nicotinamide to first degree relatives of people with type 1 diabetes who already had developed type 1-associated autoantibodies. Unfortunately, it found no difference in the development of diabetes between the two groups during the 5 year follow-up period. The study gave high doses of the vitamin, up to 3 g/day (30-50 times higher than the RDA) (Gale et al. 2004). Read this study and you can almost feel the disappointment-- we can identify who is at risk of developing type 1 but we can’t do a thing about it.

Lead and Cadmium Levels Are Associated With Lower Nutrient Levels

lead and cadmium

In U.S. adults, lead and cadmium levels are associated with higher levels of oxidative stress (serum gamma-glutamyltransferase, GTT), and lower levels of the nutrients vitamin C, vitamin E, and carotenoids.
Another double-blind, placebo controlled study in Sweden gave high doses of anti-oxidants (including nicotinamide, vitamin C, vitamin E, Beta-carotene, and selenium) to people after they were already diagnosed with type 1 diabetes and also found that they had no effect in protecting the beta cells against the damage of free radicals (Ludvigsson et al. 2001). There is no evidence linking the anti-oxidants alpha- or beta-carotene levels and the development of type 1 related autoimmunity in another study as well (Prasad et al. 2011).

However! We have some good news finally. A large, prospective, international study of children at genetic risk of type 1 diabetes (the TEDDY study) found that children who had higher levels of vitamin C in their blood had a lower risk of islet autoimmunity (although not a lower type 1 diabetes risk) (Mattila et al. 2019). A further TEDDY study also discusses this finding, as well as other nutritional factors (Li et al. 2020).

Uusitalo et al. (2008) also found that if pregnant women took anti-oxidants and trace minerals (including retinol, beta-carotene, vitamin C, vitamin E, selenium, zinc, or manganese) during pregnancy, there was no effect on the risk of the child's developing type 1-related autoimmunity. Pregnant women with type 1 diabetes have a higher risk of complications if they are deficient in vitamin C (Juhl et al. 2017). A Finnish study also found no association between maternal vitamin C or iron intake during pregnancy and the risk of islet autoimmunity or type 1 diabetes in the offspring (Mattila et al. 2021). And, vitamin supplements during pregnancy do not appear to be associated with the offspring's risk of developing celiac disease (an autoimmune condition common in people with type 1 diabetes) (Yang et al. 2017).

While these studies did not find promising results concerning anti-oxidant supplements, they also did not find that these supplements did any harm. But wait, the story might be more complicated...

Free radicals may play a role in the inflammatory process that destroys the beta cells in type 1 diabetes (Ludvigsson et al. 2001) (see the oxidative stress page for more information about its potential role in type 1 diabetes). Therefore, anti-oxidants have been thought to protect the body from oxidative stress due to the production of free radicals. But, there is some animal evidence that anti-oxidant supplements may also increase insulin resistance, showing that the relationship may not be so simple. When the researchers gave certain mice an anti-oxidant, they were more likely to become insulin resistant (Loh et al. 2009). Perhaps this finding could help explain why anti-oxidant supplements have not been found to be protective against type 1 diabetes.

For type 2 diabetes and obesity, a review finds "marginal" benefits for supplementing with antioxidants including zinc, lipoic acid, carnitine, cinnamon, green tea, and possibly vitamin C plus E. The evidence is weaker for supplementing with omega-3s, coenzyme Q10, green coffee, resveratrol, and lycopene (Abdali et al. 2015). However dietary intake of antioxidants is associated with a reduced risk of type 2 diabetes, although once people reached a certain level, there was a plateau where more wasn't better (Mancini et al. 2018).

Czernichow et al. (2009) found that anti-oxidant supplements were not protective against metabolic syndrome. Yet they also found that the people who had the highest levels of some anti-oxidants (beta-carotene, vitamin C, and vitamin E) in the beginning of the study, presumably due to a diet rich in plant foods, did have a lower risk of developing metabolic syndrome.

For people with diabetes and kidney disease, high doses of vitamin E had beneficial effects on markers of kidney injury, inflammation, and oxidative stress, but did not affect fasting blood glucose or insulin resistance (Khatami et al. 2016). A review of antioxidant supplements in people with diabetes show a possible reduced risk of retinopathy (Tabatabaei-Malazy et al. 2019).

Beta-carotene may be protective against the effects of phthalates, at least according to a U.S. study, as well as animal studies. In particular, it may help reduce the risk of insulin resistance (Li et al. 2019). Polyphenols (found in plant-based foods) may be protective against the effects of various environmental chemicals, such as BPA, phthalates, and persistent organic pollutants (Żwierełło et al. 2019).

Other Vitamins

Folic acid (folate), a B vitamin, when given to pregnant and lactating rats in high doses, caused insulin resistance in their offspring (Keating et al. 2015). In humans, folic acid supplements during pregnancy may increase the risk of gestational diabetes (Li et al. 2019Zhu et al. 2016). Another study also found high folate and low vitamin B12 levels during pregnancy were associated with gestational diabetes (Lai et al. 2018). (According to a systematic review, low vitamin B12 levels are associated with gestational diabetes as well (Kouroglou et al. 2019).) Another study also found high folate levels associated with an increased risk of gestational diabetes (Xie et al. 2019). However, an additional study found that higher habitual intake of folate before pregnancy was associated with a decreased risk of gestational diabetes (Li et al. 2019). (Since folic acid also decreases the risk of certain birth defects, it may be important to have the right amount. Ask your doctor please).

A deficiency in vitamin A causes high blood sugar and loss of pancreatic beta cells in adult mice (Trasino et al. 2015).

Food Processing: AGEs and Food Additives

Advanced Glycation End products (AGEs) are found in heat processed foods such as grilled meat, and have been linked to type 1 and type 2 diabetes in animal studies. They appear to predispose people to oxidative stress and inflammation, and may affect the fetus if the mother consumes them during pregnancy. A study has found that the level of AGEs that a mother eats are correlated with insulin levels in the baby. It found that if mothers have high AGE levels, and infant food is high in AGEs, it may raise the risk of diabetes in the offspring (Mericq et al. 2010). In animals, a diet with lower levels of AGEs during development protects the pancreatic islets as compared to a diet with higher AGE levels (Borg et al. 2018). Another study found that AGEs may also increase the risk of metabolic syndrome in mice as well as humans (Cai et al. 2014). Interestingly, AGEs may also be endocrine disruptors (Ravichandran et al. 2019). A chemical present in many foods and a precursor to AGEs is 3-Deoxyglucosone (3DG), which in rats increases intestinal permeability, leading to oxidative stress of the beta cells and beta cell dysfunction (Zhou et al. 2018).

Heating food can also produce acrylamide, a toxic chemical used in industry, found in cigarette smoke, and also sometimes found in food such as potato chips and french fries. It is associated with lower insulin levels in US adults (Lin et al. 2009). Industrial workers exposed to higher levels of acrylamide have higher rates of death from diabetes (Swaen et al. 2007). Acrylamide has also been associated with obesity-related measurements in human studies (Huang et al. 2018), and causes obesity in mice (Lee and Pyo, 2019). A pregnant woman's dietary acrylamide intake is associated with higher overweight/obesity in her children throughout childhood (Kadawathagedara et al. 2018). In animals, exposing either young or adult rats to acrylamide decreased beta cell mass and increased alpha cell mass (Stošić et al. 2018a; Stošić et al. 2018b). It also causes high fasting blood sugar levels, impaired glucose tolerance, and damaged islets in rats (Yue et al. 2019).

Chronic exposure to the caramel color food additive 4-methylimidazole, found in colas, induced high insulin levels and low glucose levels in mice (Rekha et al. 2018).

I do not have time to review all the food additives linked to diabetes or obesity. I will mention that MSG is a potential obesogen (Shannon et al. 2017; Shannon et al. 2018), as is BHA (Sun et al. 2019). DOSS, another food additive (and a laxative, and an oil dispersant), is also a potential obesogen (Bowers et al. 2016). A review of food additives finds that many of them have effects on the immune system that could contribute to metabolic disease like diabetes, obesity, or metabolic syndrome (Paula Neto et al. 2017). 

Some authors argue that food additives (including glucose, salt, emulsifiers, organic solvents, gluten, microbial transglutaminase, and nanoparticles) increase intestinal permeability, activate autoimmunity, and are responsible for the increasing incidence of autoimmune disease (Lerner and Matthias 2015).

In mice, the commonly used food preservative propionate, at levels eaten by humans, causes weight gain and insulin resistance, and when given to humans in a randomized, double-blind, placebo-controlled trial, affected glucagon and other hormone levels linked to diabetes (Tirosh et al. 2019).

Food additives are also linked to celiac disease (Mancuso and Barisani, 2019).

Feast, Famine, or Fast

Dutch Famine Studies

Dutch famine studies

Studies of the Dutch famine illustrate the importance of timing in developmental processes. The timing of in utero famine is associated with different later-life disease outcomes.
Most of us are not worried about famine anymore (crossing fingers), but there are some interesting studies that show that experiencing a famine in early life can increase the risk of type 2 diabetes later in life. The Dutch famine during 1944-45 provided an opportunity to study this phenomenon. People who were children during this famine have an increased risk of type 2 diabetes later in life (van Abeelen et al. 2012). If mothers experienced the famine while pregnant, their offspring show impaired insulin secretion and lower glucose tolerance in their 50s (de Rooj et al. 2006a; de Rooij et al. 2006b; Ravelli et al. 1998). The offspring also seem to prefer fatty foods in adulthood, although their total caloric intake was not different (Lussana et al. 2008). Their female offspring also had a higher risk of increased weight and more fat deposition later in life (Stein et al. 2007), as well as higher cholesterol levels and triglycerides (Lumey et al. 2009).  For some health effects, the risk varied depending on the time during pregnancy that the famine occurred (e.g., 1st vs 3rd trimester). But for type 2 diabetes, the risk was increased during any period of gestation (Roseboom et al. 2011). 

In French women, food deprivation during early life in World War 2 was associated with a higher risk of developing type 2 diabetes and high blood pressure in adulthood. Effects were stronger for individuals exposed at younger ages (Mink et al. 2020).

In a study of the Ukraine famine, the more severe the famine during pregnancy, the higher the risk of type 2 diabetes in the offspring. Early gestation seems to be the most susceptible time (Lumey et al. 2015). In a study of the Chinese famine of 1959-62, those exposed to famine during fetal life or childhood had a higher risk of diabetes and higher average blood sugar levels (HbA1c) in adulthood than those unexposed. Those exposed during adolescence or adulthood did not have a higher risk after adjusting for other factors (Wang et al. 2017). Those exposed to famine as a fetus or in childhood also had beta cell dysfunction as adults (Wang et al. 2020), and those exposed prenatally had a higher risk of type 2 diabetes as adults, irrespective of genetic risk (Wang et al. 2021).

In Bangladesh, young adults exposed to famine prenatally were underweight but still had higher blood glucose levels after a meal. Those exposed to famine after birth were more likely to be overweight, and had higher fasting blood glucose levels (Finer et al. 2016).

People exposed prenatally to the Ethiopian famine of 1983-5 had almost a 3 times higher risk of metabolic syndrome in adulthood than people unexposed. Those exposed had higher waist circumference, diastolic blood pressure, triglyceride levels, and fasting blood glucose (Arage et al. 2020).

In the Ukraine and Hong Kong, people with type 2 diabetes who were exposed to perinatal famine have an increased risk of proliferative diabetic retinopathy (Fedotkina et al. 2021).

For a free full text review of famine during pregnancy and subsequent metabolic changes in the offspring, see Stein et al. 2019. A systematic review and meta-analysis of 57 studies found that "Exposure to severe malnutrition or famine in childhood was consistently associated with increased risk of cardiovascular disease (7/8 studies), hypertension (8/11), impaired glucose metabolism (15/24) and metabolic syndrome (6/6) in later life. Evidence for effects on lipid metabolism (6/11 null, 5/11 mixed findings), obesity (3/13 null, 5/13 increased risk, 5/13 decreased risk) and other outcomes was less consistent. Sex-specific differences were observed in some cohorts, with women consistently at higher risk of glucose metabolism disorders and metabolic syndrome." (Grey et al. 2021).

Transgenerational Effects

Researchers are now looking into whether these effects continue to the 2nd generation, that is, the grandchildren of women who were pregnant during the famine. They have found that the children of fathers (who were exposed in the womb) were heavier and more obese than those unexposed (Veenendaal et al. 2013). A study from China found that prenatal exposure to famine was associated with high blood glucose and type 2 diabetes in the first generation, and high blood glucose in the second generation of offspring as adults-- especially if both parents were exposed in the womb (Li et al. 2017). 

One more interesting finding is that excess food during a boy's "slow growth period" (before puberty) is associated with an increased death rate from diabetes in his grandchildren (but only his son's children, not his daughter's). This evidence is from studies done in the Överkalix region in Sweden, using data collected since the 1890s. Some sort of nutrition-linked mechanism passed through the male line is likely the cause, but we don't know what it is (Kaati et al. 2002). This 2014 article in the science journal Nature describes this and other studies, and explains a possible link with epigenetic mechanisms: Epigenetics: The sins of the father (Hughes 2014).

Laboratory Studies

In animals, we can also see the effects of famine, or at least lower food consumption. In rodents, prenatal food restriction followed by a high-fat diet after weaning led to numerous changes in metabolism, including insulin levels, changes in islets, and glucose intolerance (Xiao et al. 2017). Primates whose mothers ate less food during pregnancy, but then a normal diet after birth, had higher fasting glucose, fasting insulin, and insulin resistance-- in all, metabolic changes that could predispose them to type 2 diabetes (Choi et al. 2011). The authors note that these results are similar to those found in rodents and sheep as well. In sheep in fact, the grand-piglets of sheep who experienced too much or too little food during pregnancy had metabolic effects such as excess weight and lipid disturbances. In other words, these effects may be able to be passed down to multiple generations (Gonzalez-Bulnes et al. 2014).

To mimic malnutrition in rodents, researchers can use a protein-deficient diet. When researchers fed mother rats such a diet, they found higher rates of diabetes in the offspring. They also found that one of the offspring's genes was "silenced"-- a gene associated with type 2 diabetes development. Nutrition, then, may have effects on gene expression that are linked to type 2 diabetes development (Sandovici et al. 2011). Interestingly, the offspring of mother mice fed a low protein diet have both lower body weight/fat mass and higher food intake throughout life than controls. The researchers found that these changes were associated with gene expression (Jousse et al. 2011).

Animals experiments also show that the risk of obesity and diabetes can be affected by the nutrition of prior generations. Both obesity and malnutrition can increase the risk of diabetes in grandchildren of lab rats. Malnutrition through the maternal line had a stronger effect than obesity through the paternal line (Hanafi et al. 2016).

If prior generations have experienced under-nutrition for multiple generations (50, in this case), undernourished rats have low birth weight, insulin resistance, and higher visceral fat, higher insulin levels, and higher susceptibility to chemical-induced diabetes. These abnormalities are not reversed after 2 generations of normal nutrient feeding. Again, epigenetic mechanisms may be behind this pattern (Hardikar et al. et al. 2015).

On the other hand, another animal experiment shows that if mothers experience a change of diet that promotes obesity, their children will be fatter, not surprisingly. However the grandchildren and great-grandchildren seem to recover from the obesity, no matter what their diet. That implies there is some possibility of reversing these trends (Tait et al. 2015). Similarly, if mothers have an obesity-promoting diet while pregnant, their offspring have worse outcomes, but when they receive a normal diet after birth, they partially recover (Li et al. 2015).


In mice, a 4-day fast-mimicking diet helped to regenerate beta cells, in mouse models of both type 1 and type 2 diabetes, as well as in human islets from people with type 1 diabetes (Cheng et al. 2017).

Feast AND Famine

In some developing countries, mothers may be both undernourished (in micronutrients) and overnourished (with gestational diabetes), exposing the offspring to increased disease risk through multiple pathways. These factors may play a role in the diabetes epidemic in India, for example (Krishnaveni and Yajnik 2017).


To download or see all the references on nutrition and other diet-related pages, including breastfeeding, cow's milk, gluten, and more, see the collection Diet, nutrition, gut, microbiome and diabetes/obesity in PubMed.