Lifestyle Modification in Prediabetes and Diabetes: A Large Population Analysis
by
Michael L. Dansinger
1,*, 
Joi A. Gleason
1,
Julia Maddalena
1,
Bela F. Asztalos
1,2 and
Margaret R. Diffenderfer
1,* 1
Boston Heart Diagnostics Corporation, 200 Crossing Boulevard, Framingham, MA 01702, USA
2
Jean Mayer USDA Human Nutrition Research Center on Aging, Tufts University, Boston, MA 02111, USA
*
Authors to whom correspondence should be addressed.
Nutrients 2025, 17(8), 1333; https://doi.org/10.3390/nu17081333
Submission received: 15 February 2025
/
Revised: 1 April 2025
/
Accepted: 4 April 2025
/
Published: 11 April 2025
(This article belongs to the Special Issue Impact of Lipids on Cardiovascular Health)
Abstract
:Background/Aims: Diabetes mellitus is a major cause of atherosclerotic cardiovascular disease (ASCVD). We examined a large population and tested the efficacy of a voluntary lifestyle program in prediabetic and diabetic subjects. Methods: Of 133,764 subjects, 56.3% were healthy, 36.2% were prediabetic, and 7.5% were diabetic. Fasting serum measurements of glucose, insulin, adiponectin, glycosylated hemoglobin (HbA1c), high-sensitivity C-reactive protein (hs-CRP), glycated serum protein (GSP), fibrinogen, myeloperoxidase (MPO), lipoprotein-associated phospholipase A2 (LpPLA2), as well as standard lipids, direct low-density lipoprotein cholesterol (LDL-C), and small dense LDL-C (sdLDL-C) were performed using standard automated assays. Follow-up sampling at 6–12 months occurred in 20.1% of the prediabetic and 22.2% of the diabetic subjects; of these, 12.2% of the prediabetic and 9.7% of the diabetic subjects participated in a voluntary, real-world, digital dietitian-directed lifestyle-modification program with a 10-year diabetes risk being calculated using a biochemical model (Framingham). Results: Prediabetic and diabetic subjects had significantly elevated triglycerides, sdLDL-C, and hs-CRP and decreased HDL-C. They were insulin resistant as compared to healthy subjects, but only diabetics had significant reductions in insulin production. Lifestyle modification significantly reduced diabetes risk by 45.6% in prediabetics and significantly increased (2.4-fold) the percentage of diabetics that were in remission at follow-up (8.2% versus 3.4%) with increased weight loss (6.5 versus 2.0 pounds). Lifestyle intervention resulted in significant favorable effects on many metabolic markers. Conclusions: The measurement of fasting glucose and insulin is essential for the detection of decreased insulin production in diabetics. A digital lifestyle program can have favorable effects on ASCVD risk factors and diabetic status.
1. Introduction
Current criteria for the diagnosis of diabetes mellitus in the United States are having a fasting plasma glucose ≥ 126 mg/dL (7.0 mmol/L) and/or a glycosylated hemoglobin (HbA1c) value ≥ 6.5% according to the American Diabetes Association [1]. Subjects with diabetes are known to be at significantly higher risk of developing atherosclerotic cardiovascular disease (ASCVD), kidney disease, diabetic retinopathy, and neuropathy [1,2]. Prediabetes has been defined as having a fasting plasma glucose of 100–124 mg/dL and/or having a HbA1c value of 5.7–6.4%; about one third of the middle-aged and elderly United States population have this condition [1]. Having prediabetes, being overweight or obese, being >45 years of age, having a parent or sibling with type 2 diabetes, being physically inactive (<3 times a week), having gestational diabetes (diabetes during pregnancy) or giving birth to a baby who weighed more than 9 pounds (4.1 kg), and/or race or ethnic background have all been associated with an increased diabetes risk according to the American Diabetes Association [1].
It is recommended that patients with prediabetes be placed on an intensive behavioral lifestyle intervention program, modeled on the Diabetes Prevention Program, to achieve and maintain 7% loss of initial body weight and increase moderate-intensity physical activity (such as brisk walking) to at least 150 min/week [1]. Intensive lifestyle intervention using exercise (mainly walking for 30 min/day or more) and caloric control (restriction of high fat, high sugar desserts) has been shown to reduce the risk of developing diabetes by more than 50% in subjects with prediabetes [1,2,3,4,5,6]. In a review of randomized trials in which 5238 prediabetics were enrolled for 2–6 years, about 15% developed diabetes in the treatment group versus about 26% in the control group, representing a 42% risk reduction [6]. However, such intensive programs involve multiple in-person sessions and have not been widely adopted in clinical practice.
Therefore, it is important to develop risk prediction tools to identify subjects at high risk for developing diabetes mellitus as well as to test high-risk subjects with an intervention that can easily be implemented. Investigators have previously developed models for predicting diabetes over 5–10 years using a variety of markers, including fasting glucose, body mass index (BMI), high-density lipoprotein cholesterol (HDL-C), parental diabetes history, and triglyceride levels [7,8,9,10,11,12,13]. Noble and colleagues reviewed data from 94 diabetes risk prediction models tested in a large number of individuals [9]. Follow-on studies that applied a risk score as part of an intervention aimed at reducing actual risk in people were sparse.
Recently we have developed a diabetes prediction model based on prospective data from the Framingham Offspring Study [13]. This model predicts the 10-year risk of developing diabetes and uses the following four biochemical parameters based on fasting blood sampling: (1) glucose, (2) glycated serum albumin, (3) adiponectin, and (4) triglycerides. The model has a C statistic of 0.897. As previously stated, lifestyle modification remains the cornerstone of therapy for both ASCVD and diabetes prevention, often using diets restricting calories, animal and trans fats and sugars, and an exercise program, along with medications if necessary [1,2,3,4,5,6]. We have previously developed and tested a lifestyle modification program that has been shown to be effective in promoting weight loss, as well as lowering serum triglyceride levels, reducing insulin resistance, and increasing HDL-C levels [14]. Our goals in this study were (1) to apply our Framingham biochemical diabetes risk model to a large population and (2) to assess the efficacy of a voluntary online lifestyle modification program in reducing ASCVD risk and improving diabetes status.
2. Materials and Methods
2.1. Human Subjects and Study Design
Studies were carried out as an anonymized, retrospective analysis of a total of 135,929 subjects whose physicians sent serum samples (after blood drawing, allowing samples to clot for no more than 30 min, centrifugation, and aliquoting), shipped on ice packs by over-night courier service, to Boston Heart Diagnostics (Framingham, MA, USA) for ASCVD risk assessment; the subjects also had testing 6–12 months later. Only data from fasting subjects were used. All data were anonymized prior to analysis, and such analyses are exempt from institutional review board review [exemption 4, https://grants.nih.gov/policy/humansubjects.htm (accessed on 9 September 2020), see open education resource website 45 CFR 46.104(d)]. Of these 135,929 subjects, 1.6% were excluded because they were listed as being diabetic and on insulin therapy. Of the remaining 133,774 subjects, based on Boston Heart testing over a 3-year period of time, 56.3% were classified as healthy (fasting glucose < 100 mg/dL), 36.2% were classified as having prediabetes (fasting glucose 100–125 mg/dL), and 7.5% were classified as having diabetes (fasting glucose ≥ 126 mg/dL or receiving diabetes medication, but not insulin) (Table 1).
All prediabetic and diabetic subjects were able to access a free, personalized diet and exercise plan electronically via an online portal, with online tools for tracking diet, exercise, and weight. The plan provided a single counseling session and review with a dietitian, as previously described [14]. It also provided multiple menus with recommended carbohydrate intakes ranging from 30% to 60% (median 45%) and saturated fat intakes ranging from 5% to 10% but no more than 7% for subjects with known ASCVD or low-density lipoprotein cholesterol (LDL-C) ≥ 160 mg/dL. The consumption of unrefined, natural foods such as vegetables, legumes, fruits, low-fat dairy, and whole grains was promoted. Total calorie intake was set to facilitate healthy weight as determined by the waist-to-height ratio. A 30% reduction in calories was recommended for those requiring weight loss. Caloric intake was estimated using the Mifflin–St. Jeor equation. Macronutrient and calorie targets were translated into food servings and personalized 7-day menus that were consistent with the dietary approaches to stop hypertension and with Mediterranean diets [14].
It is important to emphasize that this was not a randomized controlled trial but rather a retrospective analysis of our experience with a voluntary “real-world” lifestyle intervention program. The lifestyle evaluation occurred over a 6–12 month period based on laboratory values. There was no dietary compliance assessment. Only those subjects with no change in diabetic medications were included in the analysis of the efficacy of the lifestyle intervention program.
2.2. Laboratory Measurements
The fasting serum or red blood cell specimens sent to Boston Heart Diagnostics were measured for glucose, insulin, C-peptide, high sensitivity C-reactive protein (hs-CRP), HbA1c, direct LDL-C, direct small dense low-density lipoprotein cholesterol (sdLDL-C), and HDL-C using assays obtained from Roche Diagnostics (Indianapolis, IN, USA) and for adiponectin and glycated serum protein (GSP) using assays obtained from Diazyme Laboratories (Poway, CA, USA). All assays were performed on Roche COBAS analyzers and had within- and between-run CVs of <4.0%, as previously described [15,16,17,18,19,20,21]. With regard to the measurement of serum glucose, we have compared the values obtained with those obtained using plasma and have found that the results were virtually identical (<2.0% difference), consistent with the Roche package insert for this assay. For all populations, homeostasis model of insulin resistance (HOMAIR) values were calculated as [fasting insulin (µU/mL) × fasting plasma glucose (mg/dL)]/405, while homeostasis model of insulin production (HOMAβ) values were calculated as [360 × fasting insulin (µU/mL)]/[fasting plasma glucose (mg/dL) − 63], as previously described [22]. As shown in Figure 1, we plotted the homeostasis model of insulin sensitivity (HOMAs) values, calculated as [(1/HOMAIR) × 100], versus HOMAβ values in all subjects. We also plotted the reciprocal of this value multiplied by 100, or as [(1/HOMAIR) × 100], for the same subjects as a measure of insulin sensitivity (HOMAS) [Figure 1].
2.3. Statistical Analysis
For all populations, biochemical variables are expressed as median values with the interquartile range (25–75th percentile). Variables that were not normally distributed, including adiponectin, glycated albumin, hs-CRP, insulin, and triglycerides, were log-transformed before statistical analysis. Categorical variables are provided by percentage.
All variables between subjects that were healthy, prediabetic, or diabetic were compared using the non-parametric Kruskal–Wallis test. To predict diabetes risk in healthy subjects and prediabetic subjects in the Boston Heart Diagnostics population, we used a biochemical model that included fasting glucose, fasting triglycerides, adiponectin, and glycated serum protein with a C statistic of 0.897, as described [13].
3. Results
3.1. Comparison of Healthy, Prediabetic, and Diabetic Subjects—Age and Weight
As shown in Table 1, the percentage of women was decreased in the prediabetic group (−26.4%) and in the diabetic group (−35.4%), compared to the healthy group, indicating a sex effect. Prediabetic and diabetic men (+11.5%; +15.4%, respectively) and women (+15.5%; +15.4%, respectively) were significantly older than healthy subjects, indicating a significant age effect. As expected, BMI values were significantly higher in prediabetic and diabetic men (+7.1%; +10.7%) and women (+15.4%; +30.8%), indicating a strong obesity effect as compared to healthy subjects. Similar differences were seen for body weight.
3.2. Comparison of Healthy, Prediabetic, and Diabetic Subjects—Glucose Homeostasis
As shown in Table 1 and Table 2, HbA1c values in prediabetic and diabetic men (+3.6%; +30.9%) and women (+5.6%; +33.3%) were significantly higher than in healthy subjects. Fasting serum glucose values in prediabetic and diabetic subjects by selection were significantly higher in men (+15.2%; +68.5%) and women (+16.7%; +70.0%) as compared to healthy subjects. GSP in prediabetic and diabetic subjects was significantly higher in men (+4.0%; +52.5%) and women (+1.0%; +43.9%) than in controls, especially in the diabetic group. Fasting insulin levels were significantly higher in prediabetic and diabetic subjects in men (+44.4%; +88.9%) and women (+62.5%; +125.5%) than in controls. Fasting C-peptide values in prediabetic and diabetic subjects were similarly significantly higher in men (+39.1%; +65.2%) and women (+60.0%; +90.1%) than in controls.
The greatest difference seen, however, when comparing prediabetic and diabetic subjects with healthy subjects, was for markers of insulin resistance. HOMAIR values were markedly higher in prediabetic and diabetic subjects in both men (+75; +260%) and women (+112.0%; +306.0%) as compared to controls. With regard to insulin production, we only observed significant differences for diabetic subjects versus controls. Calculated HOMAβ values in prediabetic and diabetic men (−6.6%; −46.9%) and women (+0.3%; −37.9%) were most significantly reduced, versus controls, in diabetic subjects. In Figure 1, we have plotted HOMAs (insulin sensitivity, the reciprocal of insulin resistance multiplied by 100) on the vertical axis versus insulin production (HOMAβ) on the horizontal axis in all subjects. The data clearly indicate that while most diabetics are insulin resistant, a significant number also have evidence of decreased insulin production and relative insulin deficiency. Moreover, the median level of HOMAβ in diabetic subjects was about the same as the 25th percentile value in healthy subjects.
3.3. Comparison of Healthy, Prediabetic, and Diabetic Subjects—Inflammation
The most striking differences between prediabetic and diabetic subjects versus controls for inflammation markers, as shown in Table 1 and Table 2, were observed for hs-CRP, with much smaller differences for adiponectin, fibrinogen, myeloperoxidase (MPO), and especially for lipoprotein-associated phospholipase A2 (LpPLA2) (Table 2). Hs-CRP values were significantly higher in prediabetic and diabetic men (+20.0%; +90.0%) and women (+83.3%; +200.0%) versus controls. These differences for adiponectin in men were −8.5% and −20.2%, respectively, and in women −17.1% and −33.6%, respectively; and for fibrinogen in men +6.5% and +20.2% and in women +11.5% and +24.4%, versus controls. MPO values in prediabetic and diabetic men (+2.0%; +17.2%) and in women (+9.0%; +28.1%) were higher than in controls. For LpPLA2, the differences between prediabetic and diabetic men (−1.6%; −9.3%) and women (−1.6%; −9.7%) versus controls were quite modest.
The 10-year risk of diabetes, using the biochemical model developed from the Framingham Offspring Study, in healthy men was 0.6% and in women was 0.3%. For prediabetic men, this value was 11.6-fold higher at 7.0%, while for prediabetic women, it was 14-fold higher at 4.2% as compared to healthy subjects (Table 1).
3.4. Comparison of Healthy, Prediabetic, and Diabetic Subjects—Lipid Parameters
Table 3 shows that only very modest differences between prediabetic and diabetic subjects, versus controls, were observed for LDL-C, apolipoprotein (apo) B, and apoA-I, in contrast to fasting triglycerides, small dense low-density lipoprotein cholesterol (sdLDL-C), and HDL-C. Direct LDL-C values in prediabetic and diabetic men were −3.4% and −12.8%, respectively, and in women +5.4% and −2.6%, respectively, versus controls. For apoB, these differences in men were 0.0% and −2.1% and in women +5.4% and +7.5%, while the differences for apoA-I were in men −0.8% and −6.1% and in women −3.2% and −9.3%. In contrast, fasting triglyceride values in prediabetic and diabetic men (+12.6% and +44.7%, respectively) and women (+29.2% and +69.7%) were significantly higher than in controls, as were sdLDL-C values in men (+7.7; +19.2%) and women (+17.4%; +34.8) as compared to controls. HDL-C values in prediabetic and diabetic men were −6.1% and −18.4%, and in prediabetic and diabetic women, −10.9% and −23.4%, as compared to controls.
3.5. Effects of Lifestyle Modification in Prediabetes
In terms of the effects of lifestyle modification, as shown in Table 4, the biggest difference observed in prediabetic subjects was for predicted diabetes risk (−45.6% versus −1.6% in controls, p < 0.001), as determined by our model. Other significant effects of lifestyle change in this group were increases in adiponectin (+16.8% versus +10.7% in controls, p < 0.001) and decreases in triglycerides (−10.6% versus −6.3% in controls, p < 0.001), LDL-C (−11.1% versus 6.3% in controls, p < 0.011), and HOMAIR (−11.8% versus −8.6% in controls, p = 0.023).
3.6. Effects of Lifestyle Modification in Diabetes
As compared with the control group, the effects of lifestyle modification were somewhat greater in diabetic subjects (Table 5) than in prediabetic subjects. The most important effect was that the percentage that were not diabetic was decreased by 8.2% in the lifestyle group versus 3.4% in the control group (p = 0.012), accompanied by weight reduction of 3.2% in the lifestyle group versus 0.9% in the control group (p = 0.004). Other significant effects were decreases in HbA1c (−5.6% versus −4.2% in controls, p = 0.031), GSP (−12.7% versus −8.0% in controls, p < 0.001), fasting glucose (−10.5% versus −7.8% in controls, p = 0.028), HOMAIR (−29.5% versus −22.9% in controls, p = 0.048), hs-CRP (−30.8% versus −13.0% in controls, p < 0.001), LDL-C (−19.4% versus −7.1% in controls, p = 0.037), and fasting triglycerides (−15.6% versus −8.7%, p = 0.041). In both the prediabetic and the diabetic groups, the amount of weight loss in the lifestyle group correlated strongly with the amount of insulin resistance in both the prediabetic group (r = 0.675, p < 0.001) and in the diabetic group (r = 0.789, p < 0.001).
4. Discussion
It has long been known that subjects with diabetes are much more likely than healthy subjects to have obesity, elevated triglycerides, sdLDL-C, and hs-CRP levels, and decreased HDL-C levels, all important ASCVD risk factors [15,16,17,23,24]. Such differences are generally greater in older subjects versus younger subjects and more pronounced in women with diabetes than in men with diabetes. Moreover, since the development of insulin immunoassays, it has become clear that type 2 diabetes is often associated with both insulin resistance and decreased insulin production [25,26,27,28,29]. These findings have been facilitated by the development of equations for the calculation of HOMAIR and HOMAβ [22,29]. In addition, the measurement of fasting insulin and the calculation of HOMAIR and HOMAβ are essential for identifying subjects with diabetes requiring insulin therapy because of low insulin production, as seen in our population. Moreover, the most insulin-resistant subjects based on HOMAIR values had the greatest weight loss, and the degree of weight loss (although modest) correlated well with the level of insulin resistance at baseline. In our view, it is in diabetic subjects that the measurement of insulin or C-peptide has the greatest value. Recently the measurement of C-peptide has been recommended in subjects that have diabetes for more than 3 years, and it is known that some patients develop a relative insulin deficiency for genetic reasons [30,31,32].
The efficacy of lifestyle modification in reducing the risk of developing diabetes by promoting exercise and weight loss has been shown to be in excess of 50% in prediabetic subjects in many studies, but generally very intensive counseling has been required [1,4,5,6]. In our large population study, our web-based personalized lifestyle program with a one-session review with a dietitian reduced diabetes risk in prediabetics by about 45% and in diabetics increased the remission rate from about 3% to about 8%. However, only 10–12% of subjects chose to voluntarily participate in this free program.
Digital health lifestyle interventions may be more practical than individual in-person visits with a dietitian. A recent meta-analysis on the effectiveness of digital interventions in prediabetics, with regard to glucose homeostasis and body weight, assessed 33 studies, consisting of 14,398 subjects, with study duration ranging from 3 to 60 months and study designs consisting of in-person meetings, telephone calls, or fully digital interventions. The investigators found that overall weight loss was −1.74 kg or 3.8 pounds in the intervention group, and some improvements in glucose homeostasis were seen. They concluded that digital health lifestyle interventions can result in a statistically significant change in body weight and other secondary outcomes among people with prediabetes [33].
In another recent meta-analysis, the effectiveness of different intervention modes—digital health, face-to-face, and blended interventions—in reducing diabetes and facilitating the reversion to non-diabetic status was assessed as compared to usual care [34]. The interventions demonstrated a significant 46% conversion to non-diabetic status compared to the usual care control group. Using only digital health interventions was associated with a 12% conversion to non-diabetic status. Interventions combining digital and face-to-face interventions were associated with an 87% increase in conversion to non-diabetic status. No significant effect on the reversal of prediabetes to normoglycemia was observed with digital health interventions. The investigators concluded that face-to-face interventions have consistently demonstrated promising effectiveness in reductions in diabetes incidence and in reversion to normoglycemia in adults with prediabetes, but this was not the case for digital interventions [34]. They recommended blended approaches, which is what we did, but only had one interaction between patients and dietitians.
Diabetes is a major risk factor for ASCVD [17]. The effectiveness of lifestyle interventions to reduce ASCVD risk was assessed in a meta-analysis of 29 studies consisting of 5490 adults with no CVD at baseline; 15 of these studies were randomized controlled trials (3605 subjects). The following lifestyle interventions were implemented: diet, physical activity, motivational interviewing, problem-solving, psychological counseling, cardiovascular risk assessment and feedback, health self-management education, and peer support. Various ASCVD risk assessment tools were used, including Framingham, SCORE, Heart Health Risk Assessment Score, Dundee, ASSIGN, and the UK Prospective Diabetes Study risk score. In the 15 randomized trials, lifestyle intervention reduced absolute ASCVD risk significantly [35]. The investigators concluded that lifestyle modification had a favorable impact on the absolute ASCVD risk score in adult populations without ASCVD at baseline.
Weight loss may not benefit all subjects with diabetes, especially those that have decreased insulin production, as more commonly seen in Asian populations. Such subjects may require insulin therapy [36,37]. Nevertheless, motivated individuals who follow a regular diet and exercise plan (daily 1 h exercise) can achieve and maintain significant weight loss and markedly improve their ASCVD risk [38]. Diet intervention can have a major impact on ASCVD risk reduction, especially if animal fat is substantially reduced along with sugar [39]. It has been clearly documented that compliance and follow-up with dietitians on a regular basis is key in lifestyle intervention weight loss programs [40]. However, many prediabetic and diabetic subjects require medications to achieve glycemic control [41,42]. More recently, glucagon-like peptide 1 receptor agonists (GLP-1RA) for type 2 diabetes and obesity have been developed and are now being widely used [43]. These medications have been shown to be very effective in promoting weight loss, diabetes reversal, and ASCVD risk reduction. They also may have efficacy for metabolic liver disease, peripheral artery disease, Parkinson disease, and Alzheimer’s disease. These medications can have significant side effects, and the long-term effects of these medications are not known. Moreover, once patients go off these medications, there is often substantial weight regain. Therefore, the cornerstone of therapy for prediabetes and diabetes remains lifestyle modification with diet and exercise.
It should be emphasized that the American Diabetes Association has stressed the importance of using fasting plasma glucose for diagnosing prediabetes and diabetes instead of fasting serum glucose [1]. Concern has been raised that serum values may be significantly lower than plasma values for glucose [44]. The fact that we used fasting serum glucose measurements may be a significant shortcoming and limitation of our studies. We have specified to our healthcare providers and phlebotomists the importance of the use of gel separator tubes, only allowing tubes to clot for 30 min, and centrifugation and placement of serum into transfer vials. In our own studies under such conditions, we have not noted significant differences between plasma and serum glucose values, and the Roche package insert also indicates that either serum or plasma can be used with their automated glucose assay. Nevertheless, this remains a potential significant limitation of our studies.
5. Conclusions
In conclusion, our studies are consistent with the following concepts: (1) diabetes risk prediction is useful in subjects with prediabetes, (2) a web-based personalized lifestyle program with professional review can be effective in reducing diabetes risk, and (3) the measurement of fasting glucose and insulin and the calculation of HOMAIR and HOMAβ are useful for identifying diabetic subjects with insulin deficiency and low production who may require insulin therapy. The strength of our study is the large sample size we used. Potential weaknesses are that this was a retrospective analysis carried out only with information made available to the laboratory on the patient’s requisition and the fact that serum glucose was used instead of plasma glucose measurements.
Author Contributions
Conceptualization, M.L.D.; methodology, M.L.D. and J.A.G.; validation, B.F.A. and J.M.; formal analysis, M.L.D., B.F.A., J.M. and M.R.D.; writing—original draft preparation, M.L.D.; writing—review and editing, M.L.D., J.A.G. and M.R.D.; visualization, M.R.D.; supervision, M.L.D. All authors have read and agreed to the published version of the manuscript.
Funding
This study was supported by Boston Heart Diagnostics Corporation, Framingham, MA, USA. The research received no external funding.
Institutional Review Board Statement
Although this study entailed the analysis and interpretation of data obtained from human subjects, it had no subject enrollment, and it required all laboratory data to be de-identified prior to analysis. The study was determined on 25 September 2020 to be exempted from institutional review board approval under 45 CFR46.104(d) by the Advarra Institutional Review Board (Columbia, MD, USA). The dataset used in this study consists of patent test results that were extracted from medical records without names or identification numbers and were analyzed as anonymized data. This type of research is exempted from the requirement for human institutional review board approval as per exemption 4, as listed at https://grants.nih.gov/policy/humansubjects.htm (accessed on 9 September 2020) and at the open education resource website for research involving human subjects. Exemption 4 “involves the collection or study of data or specimens if publicly available or recorded such that subjects cannot be identified”.
Informed Consent Statement
Study participant consent was not applicable for this study.
Data Availability Statement
The data presented in this study may be available on request from the corresponding authors. Restrictions may apply due to legal reasons.
Acknowledgments
The authors thank the laboratory staff of Boston Heart Diagnostics, Framingham, MA, USA and the company’s lifestyle coaches for their technical expertise and implementation of the lifestyle intervention program. The authors also thank Ernst J. Schaefer of Boston Heart Diagnostics for data acquisition and editing of the manuscript. This research was presented in part at the annual meetings of the National Lipid Association, Atlanta, GA, USA, 3 June 2023, and at the Cardiometabolic Health Congress, Boston, MA, USA, 19–20 October 2023.
Conflicts of Interest
M.R.D. is an employee of Boston Heart Diagnostics, Framingham, MA, USA. All other authors are former employees of Boston Heart Diagnostics but have no competing interests to declare.
Abbreviations
| ASCVD | atherosclerotic cardiovascular disease |
| apo | apolipoprotein |
| BMI | body mass index |
| GLP-1RA | glucagon-like peptide 1 receptor agonist |
| GSP | glycated serum protein |
| HbA1c | glycosylated hemoglobin |
| HDL-C | high-density lipoprotein cholesterol |
| HOMAβ | homeostasis model of insulin production |
| HOMAIR | homeostasis model of insulin resistance |
| HOMAs | homeostasis model of insulin sensitivity |
| hs-CRP | high sensitivity C-reactive protein |
| LDL-C | low-density lipoprotein cholesterol |
| LpPLA2 | lipoprotein-associated phospholipase A2 |
| MPO | myeloperoxidase |
| sdLDL-C | small dense low-density lipoprotein cholesterol |
References
- American Diabetes Association Professional Practice Committee. 2. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes-2025. Diabetes Care 2025, 48 (Suppl. S1), S27–S49. [Google Scholar] [CrossRef] [PubMed]
- Centers for Disease Control and Prevention. Fact Sheet on Diabetes and Prediabetes. Available online: https://www.cdc.gov/diabetes/index.html (accessed on 12 January 2025).
- American Diabetes Association Professional Practice Committee. 9. Pharmacologic Approaches to Glycemic Treatment: Standards of Care in Diabetes-2024. Diabetes Care 2024, 47 (Suppl. S1), S158–S178, Erratum in Diabetes Care 2024, 47, 1238. [Google Scholar] [CrossRef]
- Knowler, W.C.; Barrett-Connor, E.; Fowler, S.E.; Hamman, R.F.; Lachin, J.M.; Walker, E.A.; Nathan, D.M.; Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N. Engl. J. Med. 2002, 346, 393–403. [Google Scholar] [PubMed]
- Steven, S.; Hollingsworth, K.G.; Al-Mrabeh, A.; Avery, L.; Aribisala, B.; Caslake, M.; Taylor, R. Very Low-Calorie Diet and 6 Months of Weight Stability in Type 2 Diabetes: Pathophysiological Changes in Responders and Nonresponders. Diabetes Care 2016, 39, 808–815, Erratum in Diabetes Care 2018, 41, 1321. [Google Scholar] [CrossRef]
- Hemmingsen, B.; Gimenez-Perez, G.; Mauricio, D.; Roqué I Figuls, M.; Metzendorf, M.I.; Richter, B. Diet, physical activity or both for prevention or delay of type 2 diabetes mellitus and its associated complications in people at increased risk of developing type 2 diabetes mellitus. Cochrane Database Syst. Rev. 2017, 12, CD003054. [Google Scholar] [CrossRef]
- Stern, M.P.; Williams, K.; Haffner, S.M. Identification of persons at high risk for type 2 diabetes mellitus: Do we need the oral glucose tolerance test? Ann. Intern. Med. 2002, 136, 575–581. [Google Scholar] [CrossRef]
- Wilson, P.W.; Meigs, J.B.; Sullivan, L.; Fox, C.S.; Nathan, D.M.; D’Agostino, R.B., Sr. Prediction of incident diabetes mellitus in middle-aged adults: The Framingham Offspring Study. Arch. Intern. Med. 2007, 167, 1068–1074. [Google Scholar] [CrossRef]
- Noble, D.; Mathur, R.; Dent, T.; Meads, C.; Greenhalgh, T. Risk models and scores for type 2 diabetes: Systematic review. BMJ 2011, 343, 1–31. [Google Scholar] [CrossRef]
- Bellou, V.; Belbasis, L.; Tzoulaki, I.; Evangelou, E. Risk factors for type 2 diabetes mellitus: An exposure-wide umbrella review of meta-analyses. PLoS ONE 2018, 13, e0194127. [Google Scholar] [CrossRef]
- Richter, B.; Hemmingsen, B.; Metzendorf, M.I.; Takwoingi, Y. Development of type 2 diabetes mellitus in people with intermediate hyperglycaemia. Cochrane Database Syst. Rev. 2018, 10, CD012661. [Google Scholar] [CrossRef]
- Liu, Y.; Yu, S.; Feng, W.; Mo, H.; Hua, Y.; Zhang, M.; Zhu, Z.; Zhang, X.; Wu, Z.; Zheng, L.; et al. A meta-analysis of diabetes risk prediction models applied to prediabetes screening. Diabetes Obes. Metab. 2024, 26, 1593–1604. [Google Scholar] [CrossRef] [PubMed]
- Ai, M.; Otokozawa, S.; Asztalos, B.F.; Maddalena, J.; Diffenderfer, M.R.; Russo, G.; Thongtang, N.; Liu, C.-T.; Dansinger, M.L. Diabetes mellitus risk prediction in the Framingham Offspring Study and large population analysis. Nutrients 2025, 17, 1117. [Google Scholar] [CrossRef]
- Gill, R.; Superko, H.R.; McCarthy, M.M.; Jack, K.; Jones, B.; Ghosh, D.; Richards, S.; Gleason, J.A.; Williams, P.T.; Dansinger, M. Cardiovascular risk factor reduction in first responders resulting from an individualized lifestyle and blood test program: A randomized controlled trial. J. Occup. Environ. Med. 2019, 61, 183–189. [Google Scholar] [CrossRef]
- Ikezaki, H.; Fisher, V.A.; Lim, E.; Ai, M.; Liu, C.-T.; Cupples, L.A.; Nakajima, K.; Asztalos, B.F.; Furusyo, N.; Schaefer, E.J. Direct versus calculated LDL cholesterol and C-reactive protein in cardiovascular disease risk assessment in the Framingham Offspring Study. Clin. Chem. 2019, 65, 1102–1114. [Google Scholar] [CrossRef] [PubMed]
- Ikezaki, H.; Lim, E.; Cupples, L.A.; Liu, C.-T.; Asztalos, B.F.; Schaefer, E.J. Small dense low-density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study. J. Am. Heart Assoc. 2021, 10, e019140. [Google Scholar] [CrossRef] [PubMed]
- Schaefer, E.J.; Ikezaki, H.; Diffenderfer, M.R.; Lim, E.; Liu, C.-T.; Hoogeveen, R.C.; Guan, W.; Tsai, M.Y.; Ballantyne, C.M. Atherosclerotic cardiovascular disease risk and small dense low-density lipoprotein cholesterol in men, women, African Americans and non-African Americans: The pooling project. Atherosclerosis 2023, 367, 15–23. [Google Scholar] [CrossRef]
- Schaefer, E.J.; Tsunoda, F.; Diffenderfer, M.; Polisecki, E.; Thai, N.; Asztalos, B. The Measurement of lipids, lipoproteins, apolipoproteins, fatty acids, and sterols, and next generation sequencing for the diagnosis and treatment of lipid disorders. In Endotext [Internet]; Feingold, K.R., Anawalt, B., Blackman, M.R., Boyce, A., Chrousos, G., Corpas, E., de Herder, W.W., Dhatariya, K., Dungan, K., Hofland, J., et al., Eds.; MDText.com, Inc.: South Dartmouth, MA, USA, 2000. [Google Scholar]
- Ai, M.; Otokozawa, S.; Asztalos, B.F.; White, C.C.; Cupples, L.A.; Nakajima, K.; Lamon-Fava, S.; Wilson, P.W.; Matsuzawa, Y.; Schaefer, E.J. Adiponectin: An independent risk factor for coronary heart disease in men in the Framingham Offspring Study. Atherosclerosis 2011, 217, 543–548. [Google Scholar] [CrossRef]
- Ai, M.; Otokozawa, S.; Schaefer, E.J.; Asztalos, B.F.; Nakajima, K.; Shrader, P.; Kathiresan, S.; Meigs, J.B.; Williams, G.; Nathan, D.M. Glycated albumin and direct low density lipoprotein cholesterol levels in type 2 diabetes mellitus. Clin. Chim. Acta 2009, 406, 71–74. [Google Scholar] [CrossRef]
- Rutter, M.K.; Meigs, J.B.; Sullivan, L.M.; D’Agostino, R.B., Sr.; Wilson, P.W. C-reactive protein, the metabolic syndrome, and prediction of cardiovascular events in the Framingham Offspring Study. Circulation 2004, 110, 380–385. [Google Scholar] [CrossRef]
- Matthews, D.R.; Hosker, J.P.; Rudenski, A.S.; Naylor, B.A.; Treacher, D.F.; Turner, R.C. Homeostasis model assessment: Insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985, 28, 412–419. [Google Scholar] [CrossRef]
- Buckley, D.I.; Fu, R.; Freeman, M.; Rogers, K.; Helfand, M. C-reactive protein as a risk factor for coronary heart disease: A systematic review and meta-analyses for the U.S. Preventive Services Task Force. Ann. Intern. Med. 2009, 151, 483–495. [Google Scholar] [CrossRef] [PubMed]
- Hoogeveen, R.C.; Gaubatz, J.W.; Sun, W.; Dodge, R.C.; Crosby, J.R.; Jiang, J.; Couper, D.; Virani, S.S.; Kathiresan, S.; Boerwinkle, E.; et al. Small dense low-density lipoprotein-cholesterol concentrations predict risk for coronary heart disease: The Atherosclerosis Risk In Communities (ARIC) study. Arterioscler. Thromb. Vasc. Biol. 2014, 34, 1069–1077. [Google Scholar] [CrossRef] [PubMed]
- Yalow, R.S.; Berson, S.A. Immunoassay of endogenous plasma insulin in man. J. Clin. Investig. 1960, 39, 1157–1175. [Google Scholar] [CrossRef]
- Turner, R.C.; Holman, R.R.; Matthews, D.; Hockaday, T.D.; Peto, J. Insulin deficiency and insulin resistance interaction in diabetes: Estimation of their relative contribution by feedback analysis from basal plasma insulin and glucose concentrations. Metabolism 1979, 28, 1086–1096. [Google Scholar] [CrossRef]
- Rudenski, A.S.; Hadden, D.R.; Atkinson, A.B.; Kennedy, L.; Matthews, D.R.; Merrett, J.D.; Pockaj, B.; Turner, R.C. Natural history of pancreatic islet B-cell function in type 2 diabetes mellitus studied over six years by homeostasis model assessment. Diabet. Med. 1988, 5, 36–41. [Google Scholar] [CrossRef]
- Rudenski, A.S.; Matthews, D.R.; Levy, J.C.; Turner, R.C. Understanding “insulin resistance”: Both glucose resistance and insulin resistance are required to model human diabetes. Metabolism 1991, 40, 908–917. [Google Scholar] [CrossRef]
- Levy, J.C.; Matthews, D.R.; Hermans, M.P. Correct homeostasis model assessment (HOMA) evaluation uses the computer program. Diabetes Care 1998, 21, 2191–2192. [Google Scholar] [CrossRef]
- Sacks, D.B.; Arnold, M.; Bakris, G.L.; Bruns, D.E.; Horvath, A.R.; Lernmark, Å.; Metzger, B.E.; Nathan, D.M.; Kirkman, M.S. Guidelines and recommendations for laboratory analysis in the diagnosis and management of diabetes mellitus. Diabetes Care 2023, 46, e151–e199. [Google Scholar] [CrossRef]
- Taylor, R. Understanding the cause of type 2 diabetes. Lancet Diabetes Endocrinol. 2024, 12, 664–673, Erratum in Lancet Diabetes Endocrinol. 2024, 12, e18. [Google Scholar] [CrossRef]
- Horikawa, Y.; Hosomichi, K.; Yabe, D. Monogenic diabetes. Diabetol. Int. 2024, 15, 679–687. [Google Scholar] [CrossRef]
- Fredensborg Holm, T.; Udsen, F.W.; Giese, I.E.; Færch, K.; Jensen, M.H.; von Scholten, B.J.; Hangaard, S. The effectiveness of digital health lifestyle interventions on weight loss in people with prediabetes: A systematic review, meta-analysis, and meta-regression. J. Diabetes Sci. Technol. 2024, 19322968241292646. [Google Scholar] [CrossRef] [PubMed]
- Wang, Y.; Chai, X.; Wang, Y.; Yin, X.; Huang, X.; Gong, Q.; Zhang, J.; Shao, R.; Li, G. Effectiveness of different intervention modes in lifestyle intervention for the prevention of type 2 diabetes and the reversion to normoglycemia in adults with prediabetes: Systematic review and meta-analysis of randomized controlled trials. J. Med. Internet. Res. 2025, 27, e63975. [Google Scholar] [CrossRef] [PubMed]
- Kariuki, J.K.; Imes, C.C.; Engberg, S.J.; Scott, P.W.; Klem, M.L.; Cortes, Y.I. Impact of lifestyle-based interventions on absolute cardiovascular disease risk: A systematic review and meta-analysis. JBI Evid. Synth. 2024, 22, 4–65. [Google Scholar] [CrossRef] [PubMed]
- Ikezaki, H.; Ai, M.; Schaefer, E.J.; Otokozawa, S.; Asztalos, B.F.; Nakajima, K.; Zhou, Y.; Liu, C.-T.; Jacques, P.F.; Cupples, L.A.; et al. Ethnic differences in glucose homeostasis markers between the Kyushu-Okinawa Population Study and the Framingham Offspring Study. Sci. Rep. 2016, 6, 36725. [Google Scholar] [CrossRef]
- Yoshinari, M.; Hirakawa, Y.; Hata, J.; Higashioka, M.; Honda, T.; Yoshida, D.; Mukai, N.; Nakamura, U.; Kitazono, T.; Ninomiya, T. Comparison of the contributions of impaired beta cell function and insulin resistance to the development of type 2 diabetes in a Japanese community: The Hisayama Study. Diabetologia 2021, 64, 1775–1784. [Google Scholar] [CrossRef]
- Wing, R.R.; Phelan, S. Long-term weight loss maintenance. Am. J. Clin. Nutr. 2005, 82 (Suppl. S1), 222S–225S. [Google Scholar] [CrossRef]
- Schaefer, E.J. Nutrition in the prevention of coronary heart disease and the management of lipoprotein disorders. In Modern Nutrition in Health and Disease, 12th ed.; Tucker, K.L., Duggan, C.P., Jensen, G.L., Peterson, K.E., Ross, C.A., Touger-Decker, R., Eds.; Jones & Bartlett Learning: Burlington, MA, USA, 2024; pp. 941–956. [Google Scholar]
- Dansinger, M.L.; Gleason, J.A.; Griffith, J.L.; Selker, H.P.; Schaefer, E.J. Comparison of the Atkins, Ornish, Weight Watchers, and Zone diets for weight loss and heart disease risk reduction: A randomized trial. JAMA 2005, 293, 43–53. [Google Scholar] [CrossRef]
- Turner, R.C.; Cull, C.A.; Frighi, V.; Holman, R.R.; the UK Prospective Diabetes Study (UKPDS) Group. Glycemic control with diet, sulfonylurea, metformin, or insulin in patients with type 2 diabetes mellitus: Progressive requirement for multiple therapies (UKPDS 49). JAMA 1999, 281, 2005–2012. [Google Scholar] [CrossRef]
- Echouffo-Tcheugul, J.B.; Perreault, L.; Dagogo-Jack, S. Diagnosis and management of prediabetes. A review. JAMA 2023, 329, 1206–1216. [Google Scholar] [CrossRef]
- Drucker, D.J. Efficacy and safety of GLP-1 medicines for type 2 diabetes and obesity. Diabetes Care 2024, 47, 1873–1888. [Google Scholar] [CrossRef]
- Kim, H.S. Blood Glucose Measurement: Is Serum Equal to Plasma? Diabetes Metab. J. 2016, 40, 365–366. [Google Scholar] [CrossRef]
Figure 1.
Insulin production and insulin sensitivity in healthy, prediabetic, and diabetic subjects. In this figure we have plotted data for the entire population of 133,764 subjects (56.3% healthy, 36.2% prediabetic, and 7.5% diabetic). Homeostasis assessment model assessment of insulin production, or HOMAβ, was calculated as equal to [360 × fasting insulin (µU/mL)]/[fasting plasma glucose (mg/dL) − 63] as previously described and plotted on the horizontal axis (22). The homeostasis model of insulin resistance, or HOMAIR, was calculated as equal to [fasting insulin (µU/mL)] × [fasting plasma glucose (mg/dL)]/405 as previously described [22]. We then plotted the reciprocal of this value multiplied by 100, or as [(1/HOMAIR) × 100], for the same subjects as a measure of insulin sensitivity (HOMAS). What can be clearly seen on the graph is that diabetic subjects not infrequently have HOMAβ of <60 (the 25th percentile value in healthy subjects), as well as decreased insulin sensitivity as compared to healthy and prediabetic subjects, with clear lines of demarcation between diabetic, prediabetic, and healthy subjects.
Figure 1.
Insulin production and insulin sensitivity in healthy, prediabetic, and diabetic subjects. In this figure we have plotted data for the entire population of 133,764 subjects (56.3% healthy, 36.2% prediabetic, and 7.5% diabetic). Homeostasis assessment model assessment of insulin production, or HOMAβ, was calculated as equal to [360 × fasting insulin (µU/mL)]/[fasting plasma glucose (mg/dL) − 63] as previously described and plotted on the horizontal axis (22). The homeostasis model of insulin resistance, or HOMAIR, was calculated as equal to [fasting insulin (µU/mL)] × [fasting plasma glucose (mg/dL)]/405 as previously described [22]. We then plotted the reciprocal of this value multiplied by 100, or as [(1/HOMAIR) × 100], for the same subjects as a measure of insulin sensitivity (HOMAS). What can be clearly seen on the graph is that diabetic subjects not infrequently have HOMAβ of <60 (the 25th percentile value in healthy subjects), as well as decreased insulin sensitivity as compared to healthy and prediabetic subjects, with clear lines of demarcation between diabetic, prediabetic, and healthy subjects.
Table 1.
Characteristics of prediabetic and diabetic subjects compared with healthy subjects.
| Parameter | Healthy n = 75,271 (56.3%) | Prediabetic n = 48,455 (36.2%) | Diabetic n = 10,038 (7.5%) | % Difference, vs. Healthy Subjects | |||
|---|---|---|---|---|---|---|---|
| N | Median (IQR) | N | Median (IQR) | N | Median (IQR) | ||
| Demographics | |||||||
| Age (years) | 75,271 | 52.0 (22.0) | 48,455 | 59.0 (17.0) | 10,038 | 60.0 (17.0) * | +15.4% |
| Females | 48,476 | 52.0 (22.0) | 22,977 | 59.0 (17.0) | 4175 | 60.0 (18.0) * | +15.4% |
| Males | 26,795 | 52.0 (23.0) | 25,478 | 58.0 (18.0) | 5863 | 60.0 (17.0) * | +15.4% |
| Sex | |||||||
| Females | 48,476 | 64.4% | 22,977 | 47.4% | 4175 | 41.6% * | −35.4% |
| Males | 26,795 | 35.6% | 25,478 | 52.6% | 5863 | 58.4% * | +64.0% |
| BMI (kg/m2) | 12,794 | 27.0 (7.0) | 11,758 | 30.0 (8.0) | 801 | 32.0 (9.0) * | +18.5% |
| Females | 8575 | 26.0 (8.0) | 5605 | 30.0 (10.0) | 331 | 34.0 (10.0) * | +30.8% |
| Males | 4219 | 28.0 (7.0) | 6153 | 30.0 (6.0) | 470 | 31.0 (7.0) * | +10.7% |
| Weight (pounds) | 34,298 | 169.0 (58.0) | 22,078 | 191.0 (59.0) | 4432 | 204.0 (65.0) * | +20.7% |
| Females | 22,311 | 153.0 (49.0) | 10,394 | 172.0 (57.0) | 1850 | 188.0 (63.0) * | +22.9% |
| Males | 11,987 | 195.0 (49.0) | 11,684 | 205.0 (51.0) | 2582 | 214.0 (60.0) * | +9.7% |
| Metabolism | |||||||
| Glucose (mg/dL) | 75,271 | 90.0 (10.0) | 48,455 | 106.0 (10.0) | 10,038 | 154.0 (57.0) * | +71.1% |
| Females | 48,476 | 90.0 (9.0) | 22,977 | 105.0 (9.0) | 4175 | 153.0 (56.0) * | +70.0% |
| Males | 26,795 | 92.0 (9.0) | 25,478 | 106.0 (10.0) | 5863 | 155.0 (57.0) * | +68.5% |
| Adiponectin (µg/mL) | 75,271 | 12.6 (9.3) | 48,455 | 10.1 (7.6) | 10,038 | 8.3 (6.3) * | −34.1% |
| Females | 48,476 | 14.6 (9.6) | 22,977 | 12.1 (8.6) | 4175 | 9.7 (7.4) * | −33.6% |
| Males | 26,795 | 9.4 (6.4) | 25,478 | 8.6 (5.9) | 5863 | 7.5 (5.4) * | −20.2% |
| GSP (µmol/L) | 75,269 | 199 (53) | 48,454 | 205 (59) | 10,038 | 299 (62) * | +50.3% |
| Females | 48,476 | 198 (55) | 22,976 | 200 (60) | 4175 | 285 (98) * | +43.9% |
| Males | 26,793 | 202 (52) | 25,478 | 210 (58) | 5863 | 308 (63) * | +52.5% |
| 10 yr Diabetes Risk (%) | 75,271 | 0.4 (0.6) | 48,455 | 5.5 (12.1) | 10,038 | 100.0 | +250% |
| Females | 48,476 | 0.3 (0.5) | 22,977 | 4.2 (8.9) | 4175 | 100.0 | +333% |
| Males | 26,795 | 0.6 (1.0) | 25,478 | 7.0 (14.7) | 5863 | 100.0 | +167% |
* p < 0.001 based on the non-parametric Kruskal–Wallis test. The top row is for all subjects. Percentage difference is between diabetic and healthy subjects.
Table 2.
Characteristics of prediabetic and diabetic subjects compared with healthy subjects (metabolism and inflammation).
Table 2.
Characteristics of prediabetic and diabetic subjects compared with healthy subjects (metabolism and inflammation).
| Parameter | Healthy n = 75,271 (56.3%) | Prediabetic n = 48,455 (36.2%) | Diabetic n = 10,038 (7.5%) | % Difference, vs. Healthy Subjects | |||
|---|---|---|---|---|---|---|---|
| N | Median (IQR) | N | Median (IQR) | N | Median (IQR) | ||
| Metabolism | |||||||
| HbA1c (%) | 72,980 | 5.5 (0.5) | 45,176 | 5.7 (0.5) | 9599 | 7.2 (1.9) * | +30.9% |
| Females | 47,206 | 5.4 (0.5) | 21,540 | 5.7 (0.5) | 3989 | 7.2 (1.8) * | +33.3% |
| Males | 25,774 | 5.5 (0.4) | 23,636 | 5.7 (0.6) | 5610 | 7.2 (1.9) * | +30.9% |
| Insulin (µU/mL) | 73,624 | 8.0 (8.0) | 45,176 | 13.0 (12.0) | 9940 | 17.0 (18.0) * | +112.5% |
| Females | 47,420 | 8.0 (7.0) | 21,477 | 13.0 (12.0) | 4130 | 18.0 (18.0) * | +125.0% |
| Males | 26,204 | 9.0 (8.0) | 23,699 | 13.0 (12.0) | 5810 | 17.0 (19.0) * | +88.9% |
| HOMAIR | 73,460 | 1.8 (1.8) | 45,069 | 3.5 (3.3) | 9781 | 7.3 (8.1) * | +305.6% |
| Females | 47,331 | 1.7 (1.6) | 21,421 | 3.6 (3.3) | 4042 | 7.6 (8.0) * | +347.1% |
| Males | 26,129 | 2.0 (2.0) | 23,648 | 3.5 (3.3) | 5739 | 7.2 (8.2) * | +260.0% |
| HOMAβ | 73,572 | 111 (102) | 45,176 | 108 (95) | 9940 | 64 (81) * | −42.1% |
| Females | 47,390 | 109 (98) | 21,477 | 110 (96) | 4130 | 68 (83) * | −37.9% |
| Males | 26,182 | 115 (111) | 23,699 | 107 (96) | 5810 | 62 (80) * | −46.2% |
| C-Peptide (ng/mL) | 10,322 | 2.1 (1.3) | 4897 | 3.2 (1.8) | 1394 | 3.8 (2.4) * | +81.0% |
| Females | 6597 | 2.0 (1.1) | 2310 | 3.2 (1.8) | 531 | 3.8 (2.5) * | +90.0% |
| Males | 3725 | 2.3 (1.5) | 2587 | 3.2 (1.8) | 863 | 3.8 (2.3) * | +65.2% |
| Inflammation | |||||||
| hs-CRP (mg/L) | 72,717 | 1.1 (2.4) | 45,569 | 1.6 (3.1) | 9765 | 2.5 (4.4) * | +127.3% |
| Females | 46,937 | 1.2 (2.7) | 21,723 | 2.2 (4.1) | 4052 | 3.6 (6.0) * | +200.0% |
| Males | 25,780 | 1.0 (1.9) | 23,846 | 1.2 (2.3) | 5713 | 1.9 (3.3) * | +90.0% |
| Fibrinogen (mg/dL) | 56,241 | 356 (108) | 29,643 | 386 (113) | 7621 | 433 (139) * | +21.6% |
| Females | 36,635 | 356 (107) | 13,976 | 397 (112) | 3187 | 443 (132) * | +24.4% |
| Males | 19,606 | 354 (108) | 15,667 | 377 (113) | 4434 | 425 (141) * | +20.2% |
| MPO (pmol/L) | 58,752 | 263 (146) | 34,592 | 274.0 (151) | 8538 | 317 (180) * | +20.5% |
| Females | 37,830 | 267 (149) | 16,407 | 291 (160) | 3547 | 342 (189) * | +28.2% |
| Males | 20,922 | 256 (139) | 18,185 | 261 (140) | 4991 | 300 (170) * | +17.2% |
| LpPLA2 (nmol/min/mL) | 71,783 | 184 (58) | 44,587 | 180 (59) | 9489 | 166 (61) * | −9.8% |
| Females | 46,307 | 185 (57) | 21,280 | 182 (60) | 3946 | 167 (60) * | −9.7% |
| Males | 25,476 | 182 (60) | 23,307 | 179 (60) | 5543 | 165 (63) * | −9.3% |
* p < 0.001 based on the non-parametric Kruskal–Wallis test.
Table 3.
Lipid and apolipoprotein concentrations of prediabetic and diabetic subjects compared with healthy subjects.
Table 3.
Lipid and apolipoprotein concentrations of prediabetic and diabetic subjects compared with healthy subjects.
| Parameter | Healthy n = 75,271 (56.3%) | Prediabetic n = 48,455 (36.2%) | Diabetic n = 10,038 (7.5%) | % Difference, vs. Healthy Subjects | |||
|---|---|---|---|---|---|---|---|
| N | Median (IQR) | N | vs. Healthy Subjects | N | Median (IQR) | ||
| LDL-C (mg/dL) | 73,824 | 117.0 (51.0) | 46,550 | 117.0 (54.0) | 9966 | 107.0 (57.0) * | −8.5% |
| Females | 47,539 | 117.0 (49.0) | 22,127 | 121.0 (53.0) | 4144 | 114.0 (59.0) * | −2.6% |
| Males | 26,285 | 117.0 (54.0) | 24,423 | 113.0 (55.0) | 5822 | 102.0 (55.0) * | −12.8% |
| sdLDL-C (mg/dL) | 73,572 | 24.0 (15.0) | 46,258 | 27.0 (20.0) | 9938 | 31.0 (25.0) * | +29.2% |
| Females | 47,431 | 23.0 (14.0) | 22,036 | 27.0 (18.0) | 4135 | 31.0 (25.0) * | +34.8% |
| Males | 26,141 | 26.0 (19.0) | 24,222 | 28.0 (21.0) | 5803 | 31.0 (25.0) * | +19.2% |
| HDL-C (mg/dL) | 74,272 | 58.0 (24.0) | 46,978 | 51.0 (22.0) | 10,017 | 43.0 (17.0) * | −25.9% |
| Females | 47,793 | 64.0 (25.0) | 22,306 | 57.0 (22.0) | 4161 | 49.0 (19.0) * | −23.4% |
| Males | 26,479 | 49.0 (19.0) | 24,672 | 46.0 (18.0) | 5856 | 40.0 (15.0) * | −18.4% |
| Triglycerides (mg/dL) | 75,271 | 93.0 (67.0) | 48,455 | 116.0 (80.0) | 10,038 | 150 (114) * | +61.3% |
| Females | 48,476 | 89.0 (60.0) | 22,977 | 115.0 (77.0) | 4175 | 151 (107) * | +69.7% |
| Males | 26,795 | 103.0 (75.0) | 25,478 | 116.0 (83.0) | 5863 | 149 (118) * | +44.7% |
| ApoA-I (mg/dL) | 74,089 | 160.6 (44.4) | 46,701 | 152.7 (40.6) | 9994 | 142 (36.5) * | −11.8% |
| Females | 47,705 | 170.5 (43.0) | 22,166 | 165.0 (41.6) | 4156 | 155 (38.8) * | −9.1% |
| Males | 26,384 | 144.0 (34.7) | 24,535 | 142.9 (33.8) | 5838 | 135 (31.1) * | −6.3% |
| ApoB (mg/dL) | 72,409 | 94.0 (36.0) | 44,843 | 97.0 (38.0) | 9715 | 96.0 (42.0) * | +2.1% |
| Females | 46,616 | 93.0 (35.0) | 21,338 | 98.0 (37.0) | 4030 | 100.0 (44.0) * | +7.5% |
| Males | 25,793 | 95.0 (38.0) | 23,505 | 95.0 (39.0) | 5685 | 93.0 (41.0) * | −2.1% |
* p < 0.001 based on the non-parametric Kruskal–Wallis test.
Table 4.
Effect of lifestyle intervention in prediabetic subjects.
| Parameter * | Testing Only (N = 8559) | Testing and Life Plan (N = 1179) | p Value § | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| N | Baseline | Most Recent | % Change † | p Value ‡ | N | Baseline | Most Recent | % Change † | p Value ‡ | ||
| Age (years) | 61.0 (17.0) | 60.0 (16.5) | |||||||||
| Females, % | 3988 (46.6%) | 646 (54.8%) | |||||||||
| Weight (pounds) | 2988 | 190.0 (60.0) | 188.0 (60.0) | −1.7% | 0.004 | 569 | 189.0 (58.0) | 187.0 (57.0) | −2.0% | <0.001 | 0.592 |
| HbA1c (%) | 7909 | 5.8 (0.5) | 5.7 (0.5) | −0.0% | <0.001 | 1140 | 5.8 (0.5) | 5.7 (0.5) | −0.1% | <0.001 | <0.001 |
| Glucose (mg/dL) | 8559 | 106.0 (10.0) | 104.0 (14.0) | −1.9% | <0.001 | 1179 | 107.0 (10.0) | 104.0 (12.0) | −2.7% | <0.001 | 0.067 |
| Insulin (µU/mL) | 8098 | 13.0 (11.0) | 12.0 (11.0) | −0.5% | <0.001 | 1150 | 13.0 (11.0) | 12.0 (11.0) | −1.4% | <0.001 | 0.014 |
| HOMA-IR | 8081 | 3.5 (3.2) | 3.2 (3.2) | −0.1% | 0.004 | 1150 | 3.4 (3.1) | 3.0 (3.0) | −0.4% | <0.001 | 0.023 |
| HOMA-B | 8095 | 106.4 (90.6) | 110.0 (96.0) | +5.8% | <0.001 | 1150 | 102.9 (88.6) | 105.4 (88.9) | +2.5% | 0.542 | 0.252 |
| hs-CRP (mg/L) | 7941 | 1.5 (2.9) | 1.4 (2.9) | −0.1% | 0.145 | 1171 | 1.5 (3.0) | 1.4 (2.7) | −0.2% | 0.249 | 0.171 |
| Adiponectin (µg/mL) | 8559 | 9.8 (7.6) | 10.7 (8.5) | +0.9% | <0.001 | 1179 | 9.5 (7.9) | 11.1 (9.1) | +1.4% | <0.001 | <0.001 |
| GSP (µmol/L) | 8559 | 211.0 (59.0) | 202.0 (58.0) | −7.7% | <0.001 | 1179 | 207.0 (59.5) | 199.0 (57.0) | −7.2% | <0.001 | 0.433 |
| Fibrinogen (mg/dL) | 5548 | 390.0 (109.2) | 385.0 (113.0) | −2.6% | 0.019 | 826 | 388.0 (110.0) | 380.0 (109.0) | −7.6% | 0.004 | 0.041 |
| MPO (pmol/L) | 5775 | 272.0 (144.5) | 272.0 (150.0) | +2.2% | 0.401 | 908 | 271.0 (149.0) | 261.0 (144.2) | −11.0% | 0.041 | 0.043 |
| LpPLA2 nmol/min/mL | 7758 | 179.0 (60.0) | 172.0 (59.0) | −6.5% | <0.001 | 1110 | 183.0 (59.0) | 170.5 (62.8) | −10.8% | <0.001 | 0.002 |
| 10 yr Diabetes Risk | 8559 | 6.3 (14.1) | 6.2 (12.1) | −1.6% | 0.945 | 1179 | 5.7 (13.1) | 3.1 (8.7) | −45.6% | <0.001 | <0.001 |
| LDL-C (mg/dL) | 8157 | 112.0 (56.0) | 105.0 (55.0) | −6.2% | <0.001 | 1178 | 117.0 (55.0) | 107.0 (57.0) | −11.1% | <0.001 | <0.001 |
| sdLDL-C (mg/dL) | 8071 | 27.0 (19.0) | 25.0 (17.0) | −3.0% | <0.001 | 1176 | 28.0 (21.0) | 24.0 (16.0) | −5.4% | <0.001 | <0.001 |
| HDL-C (mg/dL) | 8316 | 51.0 (22.0) | 51.0 (22.0) | +0.6% | <0.001 | 1178 | 52.0 (21.8) | 53.0 (23.0) | +1.0% | <0.001 | 0.066 |
| Triglycerides (mg/dL) | 8559 | 116.0 (80.0) | 109.0 (74.0) | −7.7% | <0.001 | 1179 | 113.0 (82.0) | 101.0 (70.0) | −15.4% | <0.001 | <0.001 |
| ApoA-1 (mg/dL) | 8286 | 153.9 (41.1) | 152.2 (41.5) | −2.1% | <0.001 | 1172 | 157.1 (42.3) | 153.0 (40.8) | −4.5% | <0.001 | 0.003 |
| ApoB (mg/dL) | 7895 | 94.0 (39.0) | 90.0 (37.0) | −4.0% | <0.001 | 1120 | 97.0 (40.0) | 89.0 (37.0) | −8.1% | <0.001 | <0.001 |
* Parameters are expressed as median (IQR). † Percent change. ‡ p-value based on paired t-test, baseline compared with most recent test results. § p-value based on the effect of the Life Plan on the most recent test results after controlling for baseline values.
Table 5.
Effect of lifestyle intervention in diabetic subjects.
| Parameter | Testing Only (n = 2017) | Testing and Life Plan (n = 216) | p Value § | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| N | Baseline * | Most Recent * | % Change † | p Value ‡ | N | Baseline * | Most Recent * | % Change † | p Value ‡ | ||
| Age (years) | 63.0 (17.0) | 62.0 (16.0) | |||||||||
| Females, % | 773 (38.3%) | 99 (45.8%) | |||||||||
| Weight (pounds) | 719 | 205.0 (65.5) | 203.0 (65.0) | −2.0% | <0.001 | 110 | 204.5 (65.0) | 198.0 (60.5) | −6.2% | <0.001 | 0.004 |
| HbA1c (%) | 1867 | 7.2 (1.8) | 6.9 (1.7) | −0.3% | <0.001 | 212 | 7.1 (1.5) | 6.8 (1.4) | −0.5% | <0.001 | 0.001 |
| Glucose (mg/dL) | 2017 | 153.0 (54.0) | 141.0 (57.0) | −17.1% | <0.001 | 216 | 152.0 (50.0) | 136.5 (49.5) | −24.7% | <0.001 | 0.028 |
| Insulin (µU/mL) | 1997 | 17.0 (18.0) | 15.0 (15.0) | −4.9% | <0.001 | 215 | 18.0 (19.0) | 16.0 (13.0) | −4.3% | 0.278 | 0.139 |
| HOMA-IR | 1980 | 7.0 (7.3) | 5.4 (6.4) | −2.9% | <0.001 | 212 | 7.8 (8.2) | 5.5 (5.6) | −1.1% | 0.751 | 0.048 |
| HOMA-B | 1997 | 62.9 (79.1) | 68.2 (83.4) | +4.3% | 0.167 | 215 | 70.2 (78.1) | 73.8 (76.3) | −7.4% | 0.491 | 0.414 |
| hs-CRP (mg/L) | 1902 | 2.3 (4.1) | 2.0 (3.8) | −0.5% | 0.014 | 216 | 2.6 (4.1) | 1.8 (3.4) | −0.5% | 0.422 | 0.625 |
| Adiponectin (µg/mL) | 2017 | 8.0 (6.1) | 9.0 (6.8) | +0.9% | <0.001 | 216 | 7.2 (5.0) | 8.1 (6.0) | +1.1% | <0.001 | 0.425 |
| GSP (µmol/L) | 2017 | 301.0 (151.0) | 277.0 (144.0) | −27.5% | <0.001 | 216 | 292.0 (127.8) | 255.5 (110.5) | −50.6% | <0.001 | <0.001 |
| Fibrinogen (mg/dL) | 1546 | 435.5 (142.8) | 436.0 (134.0) | +2.8% | 0.301 | 172 | 433.5 (128.8) | 435.0 (129.0) | −0.5% | 0.952 | 0.623 |
| MPO (pmol/L) | 1567 | 313.0 (174.0) | 303.0 (170.5) | −6.6% | 0.205 | 185 | 303.0 (192.0) | 287.0 (178.0) | −20.9% | 0.010 | 0.285 |
| LpPLA2 nmol/min/mL | 1840 | 168.0 (60.0) | 157.0 (59.2) | −8.5% | <0.001 | 203 | 167.0 (64.5) | 154.0 (59.0) | −14.6% | <0.001 | 0.076 |
| Diabetes, % | 2017 | 100.0 | 96.6 | −3.4% | <0.001 | 216 | 100.0 | 91.8 | −8.2% | <0.001 | 0.012 |
| LDL-C (mg/dL) | 2000 | 103.0 (59.0) | 94.0 (53.0) | −7.1% | <0.001 | 216 | 108.0 (55.8) | 87.0 (56.5) | −12.4% | <0.001 | 0.037 |
| sdLDL-C (mg/dL) | 1989 | 31.0 (26.0) | 26.0 (21.0) | −4.3% | <0.001 | 216 | 32.0 (27.2) | 24.0 (18.0) | −8.0% | <0.001 | 0.005 |
| HDL-C (mg/dL) | 2016 | 43.0 (17.0) | 44.0 (18.0) | +0.6% | 0.001 | 216 | 43.0 (15.5) | 43.5 (18.2) | +0.9% | 0.110 | 0.755 |
| Triglycerides (mg/dL) | 2017 | 150.0 (108.0) | 137.0 (100.0) | −16.4% | <0.001 | 216 | 154.0 (114.2) | 130.5 (88.0) | −30.7% | 0.004 | 0.174 |
| ApoA-I (mg/dL) | 2007 | 143.0 (37.3) | 141.0 (36.9) | −2.7% | <0.001 | 216 | 144.3 (36.1) | 140.5 (34.1) | −6.3% | <0.001 | 0.008 |
| ApoB (mg/dL) | 1984 | 93.0 (43.0) | 87.0 (38.0) | −5.2% | <0.001 | 212 | 96.0 (42.0) | 81.5 (35.2) | −11.5% | <0.001 | 0.004 |
* Parameters are expressed as median (IQR). † Percent change. ‡ p-value based on paired t-test, baseline compared with most recent test results. § p-value based on the effect of the Life Plan on the most recent test results after controlling for baseline values.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
MDPI and ACS Style
Dansinger, M.L.; Gleason, J.A.; Maddalena, J.; Asztalos, B.F.; Diffenderfer, M.R. Lifestyle Modification in Prediabetes and Diabetes: A Large Population Analysis. Nutrients 2025, 17, 1333. https://doi.org/10.3390/nu17081333
AMA Style
Dansinger ML, Gleason JA, Maddalena J, Asztalos BF, Diffenderfer MR. Lifestyle Modification in Prediabetes and Diabetes: A Large Population Analysis. Nutrients. 2025; 17(8):1333. https://doi.org/10.3390/nu17081333
Chicago/Turabian StyleDansinger, Michael L., Joi A. Gleason, Julia Maddalena, Bela F. Asztalos, and Margaret R. Diffenderfer. 2025. "Lifestyle Modification in Prediabetes and Diabetes: A Large Population Analysis" Nutrients 17, no. 8: 1333. https://doi.org/10.3390/nu17081333
APA StyleDansinger, M. L., Gleason, J. A., Maddalena, J., Asztalos, B. F., & Diffenderfer, M. R. (2025). Lifestyle Modification in Prediabetes and Diabetes: A Large Population Analysis. Nutrients, 17(8), 1333. https://doi.org/10.3390/nu17081333
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.
