Risk factors and causes of type 2 diabetes

Major non-modifiable risk factors include:

• a positive family history — 3 times higher with one affected parent, 5–6 times for both [1];
• age over 45 years — the risk rises with each decade;
• high-risk ethnicity — South Asians, Africans, Hispanics, Native Americans;
• a history of gestational diabetes — an almost 10 times higher risk [2];
• polycystic ovary syndrome — 3 times higher risk [3].

Known genetic variants explain only 20% of susceptibility, with TCF7L2 conferring the greatest individual risk, as the figure below shows [4].

The main modifiable factors are abdominal obesity, a sedentary lifestyle and a diet rich in refined carbohydrates and saturated fats [8]. Abdominal obesity means a waist circumference over 102 cm in men and 88 cm in women (a fivefold higher risk) [5]. A sedentary lifestyle, meaning a lot of time spent sitting, doubles the risk [6] [7]. The main predictive biochemical markers are prediabetes (5–10% annual progression to diabetes), triglycerides above 250 mg/dL (2.8 mmol/L) and HDL below 35 mg/dL (0.9 mmol/L) [9] [10].

Obesity, especially the visceral distribution of adipose tissue, gradually increases resistance to insulin action [11]. This is due to the release of free fatty acids from adipose tissue. Once they reach muscle and liver, they interfere with insulin action through the accumulation of toxic lipid metabolites (e.g. ceramides, diacylglycerol). Hypertrophied (swollen) adipocytes secrete an altered profile of adipokines, creating an environment of low-grade chronic inflammation, which further disturbs glucose metabolism [12]. Diabetes risk increases with BMI. In general, each extra kg increases the risk, with wide individual variations [13].

Paradoxically, 10% of type 2 patients are of normal weight, but instead have increased visceral adiposity. It can be detected by DEXA (a scan that measures body composition) or MRI, tests rarely used in routine practice for this purpose. Some of them actually have type 1 diabetes, the LADA form. Losing 10% of body weight significantly improves insulin sensitivity, and a sustained loss of over 15 kg can induce remission in most recently diagnosed cases [14]. This is achieved through the reduction of lipotoxicity and the recovery of beta cell function.

Type 2 diabetes has a strong hereditary component. In twin studies, heritability is estimated at about 72% [15]. The risk is 2–3 times higher when one parent has diabetes and 5–6 times higher when both do [1]. Genetic transmission is complex, with over 400 common genetic variants identified. They explain, however, only 20% of susceptibility, the rest being rare variants with large effect or gene-environment interactions still undiscovered [1].

Familial aggregation reflects not only common genes but also a common environment, with similar eating habits, physical activity level, socioeconomic status and access to medical services. Epigenetics also plays an important role. Intrauterine exposure to maternal hyperglycemia programs fetal metabolism, increasing diabetes risk by 30% [1]. After birth, epigenetic modifications induced by diet and lifestyle can be passed on to offspring. Screening of families with an affected member identifies the presence of prediabetes in up to half of first-degree relatives. This allows preventive lifestyle intervention, which can reduce progression to diabetes by 58% [16].

A sedentary lifestyle, meaning a lot of time spent sitting, doubles diabetes risk, independent of the physical activity performed the rest of the time [6] [7]. The main mechanisms are decreased metabolically active muscle mass, reduced mitochondrial density and down-regulation of glucose transporters in muscle (a fall in their number). Each additional hour spent watching television increases diabetes risk by 10% [17].

Physical inactivity significantly alters the body's metabolism. After just three days of immobilization insulin sensitivity decreases by 30%, and after two weeks you can already develop glucose intolerance (prediabetes). Muscle contraction activates insulin-independent pathways of glucose uptake, which then persist for 1–2 days, thus explaining why moderate-intensity physical exercise reduces diabetes incidence. Interrupting sedentary behavior every 30 minutes with three minutes of light activity improves glycemic control, especially after meals.

Aging brings increased diabetes risk through inevitable physiological processes. These include the decline of muscle mass by 1% annually after the age of 30, progressive mitochondrial dysfunction and the accumulation of senescent cells. The latter secrete pro-inflammatory factors. Beta cell function decreases by ~0.5% annually after the age of 20. Diabetes prevalence increases rapidly with age, being under 1% at 20 years, 5% at 40 years, 15% at 60 years and up to 40% at 80 years [18].

Diabetes incidence peaks between 65–74 years, when physiological decline converges with the accumulation of risk factors. Diabetes in the elderly is marked by an insidious onset, often masked by other morbidities, an increased risk of hypoglycemia and a greater risk of complications. Paradoxically, diabetes with onset after 75 years has a better prognosis. It can be managed with glycemic targets chosen by health status rather than age, which prioritize quality of life and the avoidance of hypoglycemia over strict control. The HbA1c target is below 7.0–7.5% (53–58 mmol/mol) in healthy older adults and below 8% (64 mmol/mol) in those with coexisting illnesses [19]. Screening is done in all adults from age 35. In people who are overweight and have at least one risk factor, it is done at any age [10].

Ethnic differences in type 2 diabetes are a reality [20]. Compared with the white population, the risk is twice as high in African Americans and 2.5 times as high in Hispanics. It is three times as high in South Asians and Native Americans. Extreme prevalence rates, of 50%, are reached in the Pima Indians (USA). Asians have an increased diabetes risk at a BMI 5 kg/m² lower and an age 10 years younger. They additionally have a distinct phenotype, with more visceral adiposity, a more severe deficit in beta cell secretion and faster progression to insulin treatment [20].

The differences between various ethnicities reflect the complex interaction between genetic predisposition, evolutionary adaptations, epigenetic factors and socioeconomic determinants. "Thrifty" genes favored the laying down of fat reserves and were an advantage when food was scarce, but become disadvantageous in the modern environment. An example of an epigenetic factor is fetal programming through maternal malnutrition followed by postnatal caloric excess. Socioeconomic determinants include unequal access to healthy food, medical services and education [1]. Treatment response also varies between ethnicities. Asians respond better to DPP-4 inhibitors (a class of diabetes medications taken as tablets). Native Americans have an increased risk of diabetic chronic kidney disease, requiring intensified screening for it [21].

Ultra-processed foods, rich in added sugars, trans fats and sodium, increase diabetes risk by at least 10% for each 10% increase in caloric intake from this category [22] [23]. The explanation lies in the high glycemic index that leads to chronic hyperinsulinemia, the high caloric density that promotes overeating and the additives that disturb the intestinal microbiome. Sugary drinks increase diabetes risk by 20% per daily serving, through a massive fructose intake [24]. Natural fruit juices (100%, no added sugar, but rich in the sugar naturally present in the fruit) increase it less, by about 7% per daily serving [25]. Fructose induces increased lipid production in the liver, where the lipids are deposited and cause insulin resistance.

Processed red meat (bacon, salami, sausages) increases diabetes risk by 37% per 50 g daily [24]. This effect comes from the intake of heme iron, nitrates/nitrites and advanced glycation products, which induce oxidative stress and inflammation. Refined grains and white rice (over 5 servings weekly) double the risk relative to whole grains, through the loss of fiber and B vitamins. At the other extreme, optimal consumption of whole grains, vegetables, fruit and dairy products is associated with a 42% lower risk [24].

Chronic psychosocial stress activates the hypothalamic-pituitary-adrenal axis (the system through which the brain tells the adrenal glands to release stress hormones) and the sympathetic nervous system. This increases cortisol and catecholamines [26]. These hormones induce insulin resistance through several mechanisms. They stimulate hepatic glucose production, inhibit muscle glucose uptake, release free fatty acids from adipose tissue and redistribute body fat toward a dangerous visceral pattern [27]. Chronically elevated cortisol is associated with a doubling of diabetes risk. Behavioral adaptation mechanisms can amplify diabetes risk.

Chronic stress leads to emotional eating, with a preference for calorie-dense foods ("comfort foods"). Next come a sedentary lifestyle through fatigue and lack of motivation, but also insufficient or fragmented sleep, which alters glucose metabolism. Reduced adherence to healthy lifestyle recommendations adds to all of this [26]. Occupational stress increases diabetes incidence, and depression raises diabetes risk by about 18% [28]. The link runs both ways: the onset of diabetes raises the risk of later depression by about 24% [27] [29]. Stress reduction interventions (e.g. mindfulness) improve glycemic control, with a fall in HbA1c of 0.5 percentage points, thus demonstrating the therapeutic potential of stress management.

Smoking increases type 2 diabetes risk by about 37% in active smokers and by almost 60% in those who smoke at least 20 cigarettes a day [30]. In former smokers the risk remains raised by 14–20% [30]. The effect is dose-proportional. Nicotine induces resistance to insulin action through sympathetic activation and the release of free fatty acids from adipose tissue. Carbon monoxide induces mild chronic tissue hypoxia. Cadmium and polycyclic hydrocarbons induce oxidative stress at the pancreatic level [31]. Smoking cessation can initially increase diabetes risk temporarily through weight gain, but the long-term benefit exceeds this transient risk [30].

Excessive alcohol consumption (over two units daily or a lot at once) increases diabetes risk through chronic pancreatitis, hepatic steatosis and malnutrition [32]. Beer and sweet alcoholic drinks confer a greater risk through their added carbohydrates. Stopping alcohol in former heavy drinkers gradually reduces diabetes risk, but it will remain higher than in the general population for the rest of their lives. The Mediterranean pattern of consumption (steady, very small amounts) slightly lowers the risk associated with alcohol compared to the Nordic pattern (episodic, excessive) [32].

Glucocorticoids (cortisone and related medications, for example prednisone) are the most diabetogenic. With prolonged treatment, they slightly raise blood glucose in two-thirds of patients and induce diabetes in 20% of them [33]. Glucocorticoids raise blood glucose by stimulating hepatic glucose production, inhibiting insulin secretion and action and redistributing fat toward a Cushingoid pattern (trunk and back of the neck). Diabetes risk increases with the dose (from 7.5 mg prednisolone or equivalent) and the duration (from 3 months). None of the medications below should be stopped or changed without consulting the doctor who prescribed it. Glucocorticoid-induced diabetes may persist after their discontinuation in approximately one quarter of cases [33]. Statins slightly increase diabetes risk through reduced insulin secretion and muscle glucose uptake.

The cardiovascular benefits are, however, so great that continuing them without interruption is well worth it [33]. Atypical antipsychotics (olanzapine, clozapine, quetiapine) increase the risk threefold through weight gain (on average 10 kg) and a direct toxic effect on beta cells [33]. Thiazide diuretics in high doses (over 25 mg hydrochlorothiazide) increase the risk by 30% through potassium loss. Non-selective beta-blockers mask hypoglycemia and slightly reduce insulin sensitivity. Protease inhibitors (HIV treatment), tacrolimus (post-transplant medication) and nicotinic acid in pharmacological doses are other classes with an associated diabetes risk. They require glycemic monitoring at initiation and possibly dose adjustments [33].