Fasting plasma glucose level in the 90-99 mg/dL range and risk of type 2 diabetes onset: Panasonic population-based cohort study 2 - Evidence review
In the Panasonic cohort, fasting glycemia between 90 and 99 mg/dL was found to be associated with a progressively higher risk of type 2 diabetes, even in people with obesity. Independent cohorts support the association, but clinical observation does not mean causality, nor does it define diagnostic thresholds or appropriate interventions for everyone on its own.
- Published 12/11/2026
- Updated 12/11/2026
- Article history
ABSTRACT
The analyzed study evaluated whether fasting plasma glycemia values still below 100 mg/dL might signal a higher future risk of type 2 diabetes. In a retrospective cohort of 37,148 Japanese workers without diabetes and with an initial glycemia below 100 mg/dL, followed from 2008 to 2018, the risk increased progressively when moving from values equal to or lower than 89 mg/dL to values between 90 and 99 mg/dL.
The research contributes to the passage from “The Choice to Live in Health or Illness” by Dr. A. Colonnese and R. Panzironi, according to which the high end of glycemia considered normal can be a prognostic indicator of metabolic risk, even in people with obesity. Independent studies in US, Mediterranean, and Israeli populations show a consistent association. However, observational cohorts do not prove that a small increase in glycemia directly causes diabetes, nor do they establish a universal clinical threshold on their own.
Analyzed study
Fasting plasma glucose level in the range of 90-99 mg/dL and the risk of the onset of type 2 diabetes: Population-based Panasonic cohort study 2
DOI: 10.1111/jdi.13692
PMID: 34624178
PMCID: PMC8902401
The research question was whether, in individuals with fasting glycemia below 100 mg/dL, values between 90 and 99 mg/dL were associated with a higher probability of developing type 2 diabetes over time. The hypothesis was that the risk does not appear suddenly only upon exceeding diagnostic thresholds, but increases gradually even in the high range of values commonly considered normal.
The relevance to paragraph 2.4.1 is precise. In "The Choice to Live in Health or Illness" by Dr. A. Colonnese and R. Panzironi, the publication is cited to support the claim that, between 90 and 99 mg/dL, the increase in fasting glycemia is associated with a progressively higher risk of diabetes and that this trend is also observable in participants with obesity. The study can document a prognostic association within the context of the examined cohort. However, it cannot demonstrate that glycemia is the sole cause of diabetes or that the same results apply identically to all populations.
Materials and methods of the original study
The authors conducted a retrospective cohort study, which is an analysis of health data collected in the past and observed over time. They used the annual medical check-ups of Panasonic employees in Japan from 2008 to 2018. All participants underwent annual blood tests and completed a standardized questionnaire on habits and lifestyle.
Of the 140,590 workers evaluated in 2008, those who did not have consecutive check-ups, already had diabetes, had an initial glycemia between 100 and 125 mg/dL, or had missing data were excluded. The final population included 37,148 people with an initial glycemia of less than 100 mg/dL. Incident diabetes was defined as a fasting glycemia equal to or greater than 126 mg/dL, the use of antidiabetic medication, or a history of diabetes reported by the participant.
Glycemia was divided into six categories: equal to or less than 89, 90-91, 92-93, 94-95, 96-97, and 98-99 mg/dL. The main analysis used Cox regression, a method that estimates the relative rate at which an event occurs over time. The results are expressed as hazard ratio or HR, which is the ratio between the rate of diabetes onset in the groups compared. The models were adjusted for age, sex, body mass index, blood pressure, lipids, smoking, dietary habits, alcohol, and physical activity. A separate analysis was also performed on people with obesity, defined as a body mass index equal to or greater than 25 kg/m² according to the criteria adopted by the authors.
Results
During the observation period, 1,028 of the 37,148 participants developed type 2 diabetes. At baseline, those who later developed diabetes had, on average, higher glycemia compared to those who did not: 92.69 versus 89.35 mg/dL. They also had, on average, a higher body mass index, blood pressure, and triglycerides, while HDL cholesterol, commonly known as “good cholesterol,” was lower.
The central finding was a progressive increase in risk associated with glycemia. After adjusting for other measured factors, each 1 mg/dL increase in the 90-99 mg/dL range was associated with a 9% increase in the rate of diabetes onset. Compared to the group with glycemia at or below 89 mg/dL, the HR was 1.53 for 90-91 mg/dL, 1.76 for 92-93 mg/dL, 1.89 for 94-95 mg/dL, 3.17 for 96-97 mg/dL, and 3.41 for 98-99 mg/dL. In practical terms, in the 98-99 mg/dL group, diabetes appeared at a rate more than three times higher than in the reference group; all these comparisons were statistically significant.
The direction of the association remained similar in people with obesity. In this subgroup, compared to glycemia at or below 89 mg/dL, the HRs were 1.56 for 90-91 mg/dL, 1.82 for 92-93 mg/dL, 2.05 for 94-95 mg/dL, 3.53 for 96-97 mg/dL, and 3.28 for 98-99 mg/dL. The absolute incidence also increased: in people with obesity, it rose from 3.97 cases per 1,000 person-years in the group at or below 89 mg/dL to 15.48 cases per 1,000 person-years in the 98-99 mg/dL group.
Interpretation
The most immediate implication is that a fasting glycemia still below 100 mg/dL does not necessarily identify individuals with the same future risk. In the studied cohort, rising within the 90-99 mg/dL range was associated with a higher probability of a future diagnosis of diabetes, showing an increasing gradient rather than a jump observable only after a single threshold.
The segment of the narrative directly documented by the study is: higher fasting glycemia, even if below 100 mg/dL → greater future risk of type 2 diabetes. The analysis in people with obesity reinforces the fact that the association was present even in this subgroup, after statistical adjustment for other available characteristics. However, it does not prove that the association is completely independent of any possible effect of excess weight, because statistical adjustment does not eliminate unmeasured factors, measurement errors, or changes that occurred during follow-up.
The authors hypothesize that, in some people, higher but non-diagnostic values may reflect early alterations in insulin secretion or insulin sensitivity. This is a biologically plausible interpretation, but the study did not measure insulin, oral glucose tolerance tests, or glycated hemoglobin; therefore, it cannot establish which mechanism produced the increased risk.
Independent supporting evidence
A large U.S. cohort examined 46,578 adults with an initial glycemia below 100 mg/dL. After adjusting for age, sex, body mass index, blood pressure, lipids, smoking, cardiovascular disease, and hypertension, each 1 mg/dL increase in fasting glycemia was associated with a 6% increase in the risk of diabetes. Participants with values of 95-99 mg/dL had an HR of 2.33 compared to those below 85 mg/dL. This directly confirms the passage in the book stating that the high end of normoglycemia can be associated with a higher future risk, even after statistical consideration of other metabolic factors [1]. The limitation is that this is also an observational cohort and that the participants were members of a U.S. healthcare system, not Japanese workers.
A Mediterranean cohort of 13,845 apparently healthy adults, followed for 108,061 person-years, identified 307 new cases of diabetes. Compared to the group with glycemia of 51-82 mg/dL, the 91-99 mg/dL range was associated with an adjusted HR of 2.03; the 95% confidence interval was 1.18-3.50. The 83-90 mg/dL range, however, did not show a statistically conclusive estimate, with an HR of 1.42 and a confidence interval of 0.42-4.74 [2]. This result confirms the direction of the association but also signals that estimates may be less precise in the lower ranges or when the number of events is limited.
In an Israeli cohort of 13,163 young men with initial glycemia below 100 mg/dL, 208 incident cases of diabetes occurred over 74,309 person-years. The risk increased progressively starting from glycemia values equal to or greater than 87 mg/dL, after adjusting for age, family history, body mass index, physical activity, smoking, and triglycerides. The association was also particularly high when values of 91-99 mg/dL coexisted with obesity or elevated triglycerides [3]. This evidence is consistent with the Panasonic observation in people with obesity, but it does not prove that the effect of glycemia is identical in women, the elderly, or populations with different characteristics.
Conflicting evidence
No directly conflicting evidence of sufficient relevance has been identified; this does not prove the absence of contrary findings. Differences between studies primarily concern the shape of the relationship and the value at which the risk appears to increase most clearly. In the Mediterranean cohort, for example, the association was statistically clear for 91-99 mg/dL but not for 83-90 mg/dL [2]. This suggests that the continuum should not be interpreted as an identical, linear, and universally quantifiable curve across every population.
Furthermore, the available results are derived from observational studies with definitions of diabetes, follow-up duration, age, sex, healthcare context, and obesity distribution that are not perfectly overlapping. For this reason, the association between high-normal glycemia and future risk is consistent, while the precise threshold and the magnitude of the increased risk remain context-dependent.
Summary table of evidence
The table distinguishes the study under review from the independent cohorts used to evaluate the same scientific narrative.
| Reference | Study type | Population or model | Main result | Direction of evidence | Strength of evidence |
|---|---|---|---|---|---|
| Original study | Retrospective cohort | 37,148 Japanese workers without diabetes | Between 90 and 99 mg/dL, the risk of diabetes increases progressively; similar trend in people with obesity. | direct confirmation | moderate |
| [1] | Retrospective cohort | 46,578 US adults with baseline glycemia below 100 mg/dL | Each 1 mg/dL increase is associated with an HR of 1.06; 95-99 mg/dL: HR 2.33 compared to less than 85 mg/dL. | direct confirmation | moderate |
| [2] | Observational cohort | 13,845 apparently healthy Mediterranean adults | Glycemia of 91-99 mg/dL is associated with an HR of 2.03 for diabetes compared to 51-82 mg/dL. | direct confirmation | moderate |
| [3] | Observational cohort | 13,163 young Israeli men with baseline glycemia below 100 mg/dL | Risk increases progressively from 87 mg/dL; glycemia of 91-99 mg/dL combined with obesity is associated with higher risk. | indirect confirmation | moderate |
Confirmation of the evidence
Claim in the book
In section 2.4.1, “The Choice to Live in Health or Illness” by Dr. A. Colonnese and R. Panzironi, it is stated that, in the Panasonic cohort, the risk of diabetes increases progressively between 90 and 99 mg/dL of fasting glycemia and that a similar trend is observed in participants with obesity.
Fasting glycemia 90-99 mg/dL → progressive increase in future risk of type 2 diabetes → association also present in the subgroup with obesity
The publication subject to the Evidence Review was used in the book to support the following passage: within the 90-99 mg/dL range, higher fasting glycemia is associated with a progressively greater probability of developing type 2 diabetes, even among people with obesity.
The correspondence between the original publication and the book excerpt is direct: the Panasonic study precisely measured this association, showing increasing HRs in the 90-91, 92-93, 94-95, 96-97, and 98-99 mg/dL categories, with a similar trend in the subgroup with obesity. Independent literature reinforces the same narrative. In the United States, higher normal glycemia was associated with a progressive increase in risk, even after adjustment for body mass index and other metabolic factors [1]. In a Mediterranean population, the 91-99 mg/dL range was associated with approximately double the risk compared to lower values [2]. In young Israeli men, the risk increased even within the normoglycemic range, and the combination of 91-99 mg/dL glycemia with obesity was associated with a particularly high risk [3]. Overall, these independent cohorts support what is reported in “The Choice to Live in Health or Illness” by Dr. A. Colonnese and R. Panzironi: fasting glycemia is a continuous prognostic indicator, but it is not in itself causal evidence nor a universal diagnostic threshold.
Overall level of confirmation
STRONG CONFIRMATION
The passage is documented directly in the original cohort and confirmed by multiple independent cohorts in different populations. The confirmation is strong on a prognostic-observational level; however, causal demonstration and the definition of a single threshold value valid for all individuals remain limited.
Conclusions
The Panasonic study clearly supports the association between higher fasting glycemia in the 90-99 mg/dL range and an increased future risk of type 2 diabetes. This result is also consistent within the subgroup with obesity. The independent cohorts make the narrative quite well-supported from a prognostic standpoint. It remains necessary to distinguish a risk indicator from a proven cause: these data do not establish that every increase in glycemia causes diabetes, nor do they, on their own, indicate a threshold for diagnosis or treatment.
Overall assessment
The narrative is well supported by independent cohorts: higher fasting glycemia values, even when below 100 mg/dL, are associated with a greater future risk of diabetes. The link is also confirmed in contexts involving obesity, but it remains observational and does not define a causal relationship or a universal clinical threshold.
References
Editorial note
Artificial Intelligence was used solely as a tool to support editorial and documentary work. All content, interpretations, bibliographic citations and conclusions have been verified, reviewed and approved by the author, who assumes full scientific responsibility for them.
The manuscript was also subject to independent peer review prior to publication.
To ensure maximum transparency, the article's complete history — including the dates of submission, review, acceptance and publication, as well as information on the peer review process — is available on the dedicated Article History page.