Glycemia and insulin in the meal response

The postprandial insulin response is one of the primary mechanisms through which the body maintains glucose homeostasis after a meal. Although it is commonly associated with the rise in glycemia following carbohydrate ingestion, we now know that insulin release is the result of a much more complex regulatory network. Glucose remains the primary physiological stimulus for insulin secretion, but amino acids, fatty acids, other energy substrates, and numerous non-nutritional factors also participate in modulating the response of pancreatic β-cells. These signals converge on shared intracellular mechanisms, resulting in highly dynamic and adaptable insulin secretion. Furthermore, the insulin response exhibits significant variability both between individuals and among the same pancreatic islets, reflecting differences in sensitivity to metabolic stimuli and in the secretory capacity of the β-cells themselves. This heterogeneity may help explain the varying individual susceptibility to the development of hyperinsulinemia, insulin resistance, and diabetes. Genetic, environmental, and pharmacological factors, food additives, and lifestyle habits can also modify insulin secretion, sometimes independently of glucose concentration, confirming that the postprandial response does not depend exclusively on glycemia but on the integration of multiple physiological signals (Kolic et al., 2026) [11].

For many years, the assessment of the insulin response was attributed almost exclusively to the amount of carbohydrates ingested. More recent evidence shows that this interpretation is incomplete. The parameter that most accurately describes the extent of insulin secretion after a meal is the glycemic load, as it simultaneously integrates the quality of carbohydrates, expressed by the glycemic index, and their actual consumed quantity. In a study conducted on over one hundred foods belonging to different categories and subsequently validated on mixed meals, the glycemic load proved to be the best predictor of both postprandial glycemia and insulinemia, explaining a significantly higher portion of variability compared to simple carbohydrate content. While the latter maintains a fair predictive capacity for individual foods, it loses much of its accuracy when foods are consumed within complex meals, in which proteins, fats, and fibers modify the digestion and absorption of nutrients. The glycemic load, by contrast, continues to maintain a significant association with the insulin response even in mixed meals, confirming itself as a physiologically more representative indicator of the body's actual insulin demand (Bao et al., 2011) [12].

These observations have been further reinforced by experimental studies in which meals characterized by different glycemic indices and different amounts of carbohydrates were compared under controlled conditions. It emerged that high-glycemic index foods produce increases in glycemia and insulinemia nearly double those of low-glycemic index foods when compared at the same quantity. However, when different meals are designed to have a similar glycemic load, the glycemic and insulin responses are surprisingly overlapping, despite significant differences in food composition and portion sizes. This indicates that the insulin response depends primarily on the combined effect of carbohydrate quality and quantity rather than the glycemic index alone (Galgani et al., 2006) [13].

A further step forward in understanding the postprandial insulin response comes from the development of the Food Insulin Index (FII), an index designed to directly estimate the insulin demand of foods. By analyzing a database including 147 common foods, it was observed that the glycemic load represents the best individual predictor of the food insulin index, followed by the glycemic index and the available carbohydrate content. Fats, proteins, and sugars also contribute to the insulin response, but to a significantly lesser extent. Even more interesting is the fact that the nutritional composition of foods alone can explain only about half of the variability in the insulin response, while the integration of glycemic response, sugar content, and proteins allows for explaining about 78% of the observed variability. This confirms that insulin secretion is the result of the interaction between multiple characteristics of the food and not just the amount of carbohydrates introduced (Bell et al., 2016) [12].

In overweight subjects with impaired glucose tolerance, higher postprandial insulin concentrations were observed after the consumption of both high- and low-glycemic index meals, suggesting a greater secretory demand on β-cells to maintain glycemic control. Despite these individual differences, low-glycemic index meals continue to result in a significantly lower insulin response compared to high-glycemic index meals in all categories of participants, demonstrating that carbohydrate quality maintains a decisive role even in the presence of pre-existing metabolic alterations (Perälä et al., 2011) [13].

The overall composition of the meal, such as the amount of fiber present, represents another determining element: in fact, in patients with type 2 diabetes, a breakfast characterized by both a high glycemic index and low fiber content produced the highest insulin response among all combinations studied, while increasing the fiber content significantly attenuated insulinemia even when the glycemic index remained unchanged. In parallel, low-glycemic index breakfasts resulted in greater suppression of ghrelin (a hormone involved in appetite regulation) without significantly changing the subjective perception of hunger. These results suggest that carbohydrate quality and fiber content act in a complementary manner in modulating the postprandial insulin response (Silva et al., 2015) [14].

Similar results have also been observed in the long term. After ten weeks of a low-glycemic index diet, healthy women continued to show lower peaks of glycemia and insulin compared to subjects fed a high-glycemic index diet, with no differences in energy expenditure or gastric emptying. The favorable effect therefore seems to depend primarily on the modulation of the glycemic and insulin response rather than on changes in energy metabolism. Furthermore, participants fed a low-glycemic index diet reported greater satiety and a lower desire for fat-rich foods, suggesting a possible benefit for appetite control as well (Krog-Mikkelsen et al., 2011) [15].

Therefore, the favorable effects of reducing the glycemic index or carbohydrate content do not seem to be limited to a single meal, but can extend to the entire day's intake. In a crossover study conducted under strictly controlled metabolic conditions, all nutritional strategies that involved a reduction in the glycemic index, carbohydrate content, or both resulted in a significant decrease in overall glucose and insulin exposure over the following twelve hours compared to a high-glycemic index, high-carbohydrate diet. The combination of a low glycemic index and low carbohydrate content produced the greatest reduction in the daily insulin response, while no significant differences emerged in subjective feelings of hunger or satiety. These observations confirm that intervening in the quality and quantity of carbohydrates represents an effective strategy for attenuating postprandial insulin stimulation throughout the day, potentially contributing to reducing the functional load imposed on pancreatic β-cells (Liu et al., 2012) [16].

If fasting plasma glycemia can be compared to a photograph that captures glucose metabolism at a precise moment, glycated hemoglobin (HbA1c) is more like a film that tells the story of what has happened over the last two or three months. This parameter is derived from the non-enzymatic glycation of hemoglobin present in red blood cells (the process by which glucose in the blood binds spontaneously to hemoglobin): the longer glucose remains elevated in the blood, the greater the amount of hemoglobin that is modified. Since erythrocytes have an average lifespan of about 120 days, the percentage of HbA1c reflects the body's average exposure to glucose during the preceding months, offering a much more stable measure than a single fasting glycemia test. Furthermore, unlike the latter, it does not require fasting, has lower daily biological variability, and is less influenced by acute lifestyle fluctuations or occasional events—characteristics that have favored its progressive introduction among international diagnostic criteria for diabetes. However, some clinical conditions remain, such as anemia, hemoglobinopathies, renal failure, pregnancy, and alterations in erythrocyte turnover, which can modify its accuracy and therefore require correct interpretation of the result (American Diabetes Association, 2026) [33]; (Kharroubi et al., 2014) [37].

This different physiological nature explains why fasting glycemia and HbA1c should not be considered alternatives, but rather complementary. The former measures the delicate balance between hepatic glucose production and basal insulin secretion at a specific time of day; the latter, instead, integrates thousands of glycemic oscillations that occur during weeks of daily life, including both fasting periods and responses to meals. Consequently, two individuals may have an almost identical fasting glycemia but different HbA1c values, simply because they experience profoundly different glucose exposures during the rest of the day. In this sense, HbA1c represents an indicator of the overall quality of glycemic control and the persistence of hyperglycemia, offering information that a single glycemia determination cannot provide.

Precisely because of these characteristics, in recent years HbA1c has taken on an increasingly relevant role in the screening and diagnosis of type 2 diabetes. However, numerous studies conducted in different populations demonstrate how its diagnostic performance depends heavily on the threshold value adopted. A vast systematic review with meta-analysis including 37 studies highlighted that the traditional 6.5% cut-off, currently recommended by international guidelines, possesses extremely high specificity (97.3%), thus proving very effective in confirming the diagnosis when the test is positive. Sensitivity, however, stops at 50%, indicating that about one diabetic subject in two might not be identified in the early stages of the disease. The authors observed that by lowering the threshold to about 6.03%, sensitivity would increase significantly while maintaining good specificity, suggesting that the traditional 6.5% value was chosen by prioritizing diagnostic certainty over the ability to intercept all cases early (Kaur et al., 2020) [38].

Similar observations also emerge from studies conducted in different ethnic groups. In a Palestinian Arab population, for example, the threshold value that guaranteed the best balance between sensitivity and specificity was 6.3%; however, the authors deemed it appropriate to maintain the international 6.5% cut-off, as the higher specificity reduces the risk of making incorrect diagnoses. Even more interesting was the finding that HbA1c and fasting glycemia identified only a partial overlap of diabetic subjects, while concordance was particularly low in the diagnosis of prediabetes. This confirms how each test explores different aspects of the pathophysiology of glucose metabolism and how neither, used in isolation, is able to identify the entire population at risk (Kharroubi et al., 2014) [37].

The epidemiological implications of using HbA1c are also extremely relevant. A representative survey of the Canadian population demonstrated that the use of this parameter allows for the identification of a significantly higher number of people with undiagnosed diabetes compared to fasting glycemia alone. Furthermore, the subjects identified through HbA1c tended to be younger on average and less obese than those identified by fasting glycemia alone, suggesting that this biomarker can intercept persistent metabolic alterations at a relatively early stage of their evolution. While this characteristic expands prevention opportunities, it also inevitably increases the number of people eligible for monitoring and intervention programs, with significant organizational consequences for healthcare systems (Rosella et al., 2015) [39].

The most interesting aspect of HbA1c, however, concerns its prognostic capacity. It is not enough to identify a high number of subjects with prediabetes; it is equally important to understand which of them will actually develop diabetes in the following years. In a large prospective cohort of nearly 18,000 Korean adults followed for over five years, subjects classified as prediabetic based on HbA1c showed a significantly higher probability of progressing to diabetes compared to those identified exclusively by fasting glycemia; therefore, while selecting a smaller share of individuals, HbA1c identified a biologically more vulnerable population. From a pathophysiological point of view, this result appears plausible, as HbA1c integrates glycemic excursions throughout the entire day, including postprandial ones, providing a more faithful representation of the progressive impairment of β-cell function and the increase in insulin resistance (Kim et al., 2016) [40].

A similar conclusion emerges from a prospective study with over eleven years of follow-up, in which HbA1c was compared with the two-hour glycemia from an OGTT. Although the latter was confirmed as the best absolute predictor of diabetes development in the general population, HbA1c showed very similar performance in subjects with abdominal obesity, in whom the sensitivity of the test increases significantly. The authors also highlighted that OGTT and HbA1c do not necessarily identify the same individuals: post-load glycemia mainly reflects an early alteration of pancreatic β-cell function and peripheral insulin sensitivity, while HbA1c describes chronic exposure to hyperglycemia. Once again, a fundamental concept emerges: no test, considered in isolation, is able to fully describe the complexity of glucose metabolism (Yoo et al., 2020) [41].

Overall, glycated hemoglobin has profoundly changed the way metabolic health is evaluated, shifting the focus from the simple measurement of glycemia at a given moment to the chronic exposure of tissues to glucose. While offering important advantages in terms of practicality, analytical stability, and prognostic capacity, it does not completely replace other glycemic markers, as each of them describes different aspects of the pathophysiology of diabetes. A truly complete evaluation therefore requires the integration of multiple diagnostic tools. Among these, the test that best explores the body's ability to manage a glucose load and that allows for the identification of metabolic alterations often still invisible to both fasting glycemia and HbA1c is the plasma glycemia measured two hours after the oral glucose tolerance test, the subject of the next subsection.

While fasting glycemia describes metabolic balance under basal conditions and glycated hemoglobin reflects the body's chronic exposure to glucose, plasma glucose measured two hours after a 75-gram oral glucose load (2-hour Plasma Glucose, 2hPG) evaluates the body's ability to handle a true metabolic challenge. During the Oral Glucose Tolerance Test (OGTT), the glycemic regulation system is subjected to a sudden increase in glucose availability, requiring rapid and coordinated intervention from pancreatic β-cells, insulin-sensitive peripheral tissues, and the liver. Glycemia measured two hours after the load therefore represents a dynamic indicator of metabolic health, as it simultaneously reflects insulin secretion capacity, tissue sensitivity to insulin, the suppression of hepatic glucose production, and peripheral glucose uptake, especially in muscle tissue. Unlike fasting glycemia, which observes metabolism in a state of equilibrium, the OGTT evaluates how the body reacts to a condition of increased functional demand (effectively simulating the metabolic response to a meal), revealing alterations that may remain completely invisible under basal conditions (Zhang et al., 2025) [42].

This characteristic explains why post-load glycemia is often able to identify very early metabolic alterations. In the initial stages of the natural history of diabetes, many individuals still maintain fasting glycemia and glycated hemoglobin levels that are apparently within normal parameters, despite having already developed a reduced capacity to manage a glucose load. In these subjects, the first manifestation of the progressive loss of β-cell function and the increase in peripheral insulin resistance is precisely the excessive increase in glycemia after a meal or after a glucose tolerance test. The OGTT therefore allows for the interception of a phase of metabolic deterioration that frequently precedes the alterations detectable through the other two main glycemic markers.

Available evidence shows that at-risk subjects identified through the OGTT present a significantly more compromised metabolic profile compared to those classified exclusively by HbA1c. In a study conducted on the population of Qatar, the concordance between HbA1c and OGTT in the diagnosis of prediabetes was modest, with less than 60% of participants classified in the same way by both tests. More importantly, subjects identified via OGTT showed significantly more compromised pancreatic β-cell function, a higher prevalence of metabolic syndrome, a worse overall metabolic profile, and a higher probability of developing diabetes compared to subjects classified as prediabetic solely on the basis of HbA1c. Conversely, many individuals positive only for HbA1c had substantially preserved β-cell function, suggesting that the two tests do not simply identify different populations, but also different stages of the pathophysiological evolution of the disease (Iskandar et al., 2019) [43].

Similar results have also been observed in a cohort of Thai patients at high metabolic risk. In this population, the OGTT identified a prevalence of dysglycemia (a condition characterized by altered regulation of blood glucose levels) significantly higher than HbA1c: this means that while HbA1c was very reliable when positive (reaching the indicative value of 6.5%), it left a considerable proportion of subjects already affected by glucose metabolism alterations undiagnosed at lower values. The authors emphasize that this limited sensitivity is particularly relevant in Asian populations, where postprandial glycemia alterations often represent the first manifestation of metabolic deterioration. Delaying diagnosis therefore means losing a precious window of time during which lifestyle interventions are particularly effective in preventing progression to overt diabetes (Thewjitcharoen et al., 2019) [44].

Not all studies, however, have highlighted a clear superiority of a single parameter. In a large multi-ethnic Asian population residing in Singapore, HbA1c and fasting glycemia showed overall comparable performance as first-line tests for diabetes screening. However, the authors observed that integrating the two tests significantly improved the identification of subjects with diabetes and prediabetes compared to using a single marker. In particular, an HbA1c value equal to or greater than 6.1%, followed by the assessment of fasting glycemia in positive subjects, allowed for the identification of a population characterized by a very high prevalence of diabetes or impaired glucose tolerance documented via OGTT. This confirms that the assessment of glucose metabolism benefits from an integrated approach, in which each test contributes to describing different aspects of the disease's pathophysiology (Lim et al., 2018) [45].

The most important evidence in favor of post-load glycemia likely comes from a recent prospective study conducted on over 14,000 US adults followed for nearly ten years. When fasting glycemia, HbA1c, and two-hour glycemia were considered separately, all three parameters were associated with an increase in total and cardiovascular mortality. However, when analyzed simultaneously in the same statistical model, only two-hour plasma glucose maintained independent prognostic value for both clinical outcomes. Fasting glycemia completely lost statistical significance, while HbA1c retained only limited predictive capacity for cardiovascular mortality under specific conditions. According to the authors, this result indicates that the glycemic response after an oral glucose load encapsulates pathophysiological information that is not fully captured by either basal glycemia or HbA1c. The 2hPG integrates pancreatic β-cell function, peripheral insulin sensitivity, the liver's ability to suppress glucose production, and the speed with which peripheral tissues remove glucose from the bloodstream, likely representing the most complete parameter for describing overall metabolic health (Zhang et al., 2025) [42].

Overall, two-hour plasma glucose represents much more than a simple diagnostic criterion for diabetes. It constitutes a true functional test of the body's ability to maintain glucose homeostasis in response to an intense metabolic stimulus, allowing for the identification of pathophysiological alterations that are often still invisible to fasting glycemia and glycated hemoglobin. If basal glycemia describes metabolism in a resting state and HbA1c tells the glycemic history of the previous months, two-hour glycemia reveals how the metabolic system reacts to a real challenge. It is precisely this ability to highlight the first signs of metabolic dysfunction that explains its high prognostic value and the central role it continues to play in assessing the risk of diabetes, cardiovascular disease, and mortality. The growing scientific interest in even earlier diagnosis of glycemic regulation alterations has finally brought attention to an emerging parameter: plasma glucose measured just one hour after an oral glucose load (1hPG), which numerous recent findings indicate as a potentially even more sensitive biomarker for the very early stages of metabolic deterioration.

For many years, the assessment of glucose metabolism has focused primarily on three parameters: fasting glycemia, glycated hemoglobin, and glycemia measured two hours after an oral glucose tolerance test. In recent years, however, a growing number of studies have highlighted that a significant portion of metabolic alterations still escapes these traditional diagnostic criteria. From this observation, interest has grown in plasma glucose measured one hour after the intake of 75 grams of glucose (1-hour Plasma Glucose, 1hPG), a parameter that appears to detect the deterioration of glycemic regulation at an even earlier stage. The importance of this evidence is such that in 2024, the International Diabetes Federation (IDF) published a specific Position Statement proposing the integration of 1-hour glycemia among the tools for identifying intermediate hyperglycemia and type 2 diabetes. According to this document, a 1-hour glycemia of 155 mg/dL (8.6 mmol/L) or higher identifies individuals who are apparently normoglycemic but already exposed to a significantly increased risk of progression toward diabetes, while values of 209 mg/dL (11.6 mmol/L) or higher are highly indicative of overt diabetes and require confirmation via a second diagnostic test (Bergman et al., 2024) [46].

From a pathophysiological perspective, 1-hour glycemia possesses characteristics that clearly distinguish it from other glycemic markers. It reflects the immediate metabolic response to a glucose load, simultaneously integrating first-phase insulin secretion, peripheral insulin sensitivity, the functional capacity of pancreatic β-cells, and the speed at which the body removes glucose from the bloodstream. These very mechanisms represent the first components to deteriorate during the progression toward type 2 diabetes. Consequently, 1-hour glycemia tends to increase when fasting glycemia, HbA1c, and even 2-hour glycemia are still within normal limits. A recent meta-analysis including over 35,500 participants from different ethnic groups confirmed that 1-hour glycemia correlates more closely with insulin sensitivity, the Matsuda index (a parameter used to estimate the body's sensitivity to insulin based on glucose and insulin values measured during an OGTT), and β-cell function compared to conventional glycemic markers, explaining why it represents one of the earliest indicators of the onset of the loss of glucose homeostasis (Ahuja et al., 2021) [47]; (Bergman et al., 2024) [46].

The same meta-analysis also defined the diagnostic performance of 1-hour glycemia with high precision. By analyzing 15 observational studies and 2,705 new cases of diabetes diagnosed via OGTT, the authors identified three possible threshold values, pinpointing the value of 11.6 mmol/L (209 mg/dL) as the best compromise between sensitivity and specificity. With this cutoff, 1-hour glycemia identified 92% of diabetes cases; particularly interesting is the observation that many of the subjects apparently classified as false positives already exhibited impaired glucose tolerance and high metabolic risk, suggesting that 1-hour glycemia can identify a biologically very early stage of the disease, preceding conventional diagnostic criteria (Ahuja et al., 2021) [47].

Longitudinal evidence further strengthens this interpretation. In a study including over 1,300 initially normoglycemic individuals followed for more than five years, 1-hour glycemia proved to be the best predictor of the future onset of both prediabetes and diabetes, clearly outperforming fasting glycemia, 2-hour glycemia, and even glycated hemoglobin. Even more significant is the fact that adding other glycemic parameters did not further improve the predictive capacity of 1-hour glycemia, suggesting that it already encapsulates most of the prognostic information provided by the entire tolerance test. From a pathophysiological standpoint, this appears consistent with the fact that the glycemic value recorded after one hour represents the moment when early insulin response and peripheral sensitivity are most severely tested (Sai Prasanna et al., 2017) [48].

Similar results have been observed by directly comparing 1-hour glycemia with HbA1c. In a study conducted on subjects at high metabolic risk, a 1-hour glycemia of 155 mg/dL or higher showed a concordance with the reference test (OGTT) approximately four times higher than diagnostic criteria based on HbA1c. The discriminatory capacity was also significantly better, while correlations with insulin sensitivity, β-cell secretory function, and 2-hour glycemia were decidedly stronger compared to glycated hemoglobin. These results confirm that 1-hour glycemia more faithfully describes the initial pathophysiological alterations that characterize the progression toward diabetes, whereas HbA1c tends to reflect a more advanced and stable phase of dysglycemia (Jagannathan et al., 2016) [49].

The most recent evidence further confirms this shift in perspective. In a large Brazilian cohort including nearly 1,800 oral glucose tolerance tests, 1-hour glycemia identified 27.8% additional cases of intermediate hyperglycemia or diabetes compared to fasting glycemia, 26.7% more than HbA1c, and 16.8% more than 2-hour glycemia (Castro et al., 2025) [50].

Taken together, this evidence outlines a substantial change in the assessment of metabolic health. While fasting glycemia describes basal metabolism, glycated hemoglobin documents chronic glucose exposure, and 2-hour glycemia evaluates the body's ability to recover after a glucose load, 1-hour glycemia appears to capture the moment when the metabolic system begins to lose its functional efficiency. It is precisely this ability to identify still-reversible alterations that led the International Diabetes Federation to propose its integration into future diagnostic criteria. Therefore, 1-hour plasma glucose could represent one of the most important tools for anticipating a diabetes diagnosis and initiating preventive interventions at a stage when metabolic damage is still largely modifiable.

The most recent evidence suggests that the very concept of "normal glycemia" deserves deeper reflection. While for many years it was believed that metabolic risk began only beyond specific diagnostic thresholds, numerous studies today indicate that the transition from health to disease does not occur suddenly, but along a biological continuum characterized by a gradual increase in risk. Even small differences in fasting glycemia, while remaining within the range commonly considered physiological, seem to reflect metabolic alterations already in progress and associated with clinically relevant consequences (Nichols et al., 2008) [51]; (Brambilla et al., 2011) [52].

One of the largest studies available, conducted on over 46,000 initially normoglycemic adults, highlighted that every increase of just 1 mg/dL in fasting glycemia below 100 mg/dL is associated with a 6% increase in the risk of developing type 2 diabetes. Subjects with values between 95 and 99 mg/dL had a risk more than twice as high as those with glycemia below 85 mg/dL, even after adjusting for obesity, blood pressure, lipid profile, smoking, and cardiovascular disease. The risk, therefore, does not depend exclusively on other metabolic factors, but is independently associated with the blood glucose concentration itself (Nichols et al., 2008) [51]. Similar results were also observed in a Mediterranean population of nearly 14,000 individuals, in which values between 91 and 99 mg/dL were associated with approximately double the risk of developing diabetes compared to subjects with lower glycemia, confirming that the high end of normoglycemia is already a significant prognostic indicator (Brambilla et al., 2011) [52].

Further confirmation comes from the large Panasonic cohort, which included over 37,000 Japanese workers followed for ten years. The authors observed that, within the range between 90 and 99 mg/dL, every increase of just 1 mg/dL in glycemia resulted in a 9% increase in diabetes risk. The increase was progressive, with no evidence of a critical threshold, reaching a risk more than three times higher in subjects with values of 98-99 mg/dL compared to those with glycemia at or below 89 mg/dL. The same trend was also found in participants with obesity, suggesting that this relationship exists independently of excess body weight (Munekawa et al., 2022) [53].

These observations have been further reinforced by a recent analysis conducted on over 220,000 individuals from China and Japan. Using non-linear statistical models, the authors demonstrated a J-shaped relationship between normal fasting glycemia and future diabetes risk. Above specific inflection points—threshold values beyond which the risk began to increase more rapidly, identified around 81 mg/dL in the Chinese population and 94 mg/dL in the Japanese population—a marked increase in risk was observed. These results suggest that glucose metabolism begins to deteriorate well before exceeding the thresholds currently used to define prediabetes, making fasting glycemia a continuous biomarker of risk rather than a simple fixed parameter (He et al., 2023) [54].

The importance of these observations does not concern only the risk of developing diabetes. The international EpiDREAM study, which involved nearly 19,000 men and women from 21 countries, demonstrated that cardiovascular risk and mortality increase progressively along the entire glycemic continuum. Moving from normoglycemia to impaired fasting glucose or impaired glucose tolerance, up to newly diagnosed diabetes, the incidence of cardiovascular events and deaths grows constantly. Even more significant is the fact that every 1 mmol/L (≈ 18 mg/dL) increase in fasting glycemia is associated with a 17% increase in the risk of cardiovascular events or death, highlighting how even modest increases in glycemia can have relevant prognostic consequences (Anand et al., 2012) [55].

A similar conclusion had already emerged from the historic DECODE study, which followed nearly 30,000 Europeans for about eleven years. The authors observed that the relationship between glycemia and mortality is not characterized by a simple threshold beyond which risk increases suddenly, but rather by a predominantly J-shaped trend. For cardiovascular mortality, in particular, glycemia measured two hours after an oral glucose tolerance test (OGTT) showed a progressive increase in risk along the entire spectrum of glycemic values, without an obvious point of discontinuity. Furthermore, post-load glycemia identified high-risk individuals more effectively than fasting glycemia alone, suggesting that cardiovascular damage may develop long before the onset of overt diabetes (DECODE Study Group, 2003) [56].

Taken together, these findings profoundly change the way we interpret glycemic values. The traditional distinction between "normal" and "pathological" glycemia remains essential for making a diagnosis, but it does not adequately describe the biological progression of risk. Glucose metabolism appears, in fact, as a continuum in which every increase in glycemia reflects, at least in part, a progressive loss of metabolic balance. This means that prevention should begin long before the diagnostic thresholds for prediabetes or diabetes are exceeded, by identifying early those apparently normoglycemic subjects with values at the upper end of the range, who already present a significantly increased risk of developing diabetes, cardiovascular disease, and premature mortality. In this light, glycemia represents not only a diagnostic parameter but a dynamic indicator of the metabolic health status of the entire organism.