Umeå University's logo

Aim: We assessed the performance of non-invasive risk prediction models to predict 10-year incident type 2 diabetes in a large Swedish cohort.

Methods: Using the Västerbotten Intervention Programme (VIP) cohort, which includes serial oral glucose tolerance tests (OGTTs), we assessed discrimination (concordance (c)-statistic) and calibration (expected-to-observed probability ratio, integrated calibration index, calibration slope and plots) before and after recalibration in twelve non-invasive models. Incident diabetes cases were determined by an OGTT at a 10-year follow-up visit or through previously validated register-based cases.

Results: Among 91708 VIP participants, the 10-year diabetes incidence was 2.8 %. Most models had acceptable discrimination (c-statistic ≥0.70 and < 0.80). Discrimination was better in women and persons <50 years old. Eight models overestimated and four models underestimated mean absolute risk. Recalibration improved miscalibration in all models. Overall, the Framingham Personal and QDScore models’ predictions were most accurate but the Framingham model included more easily obtainable variables. Most models overestimated risk in older people while no consistent pattern was observed across sexes.

Conclusion: All models required recalibration to improve prediction accuracy. The Framingham personal model is recommended for risk predictions and will be the easiest to implement.

Background: Limited evidence exists for effect modification of genetic characteristics on the associations of food consumption and incident type 2 diabetes (T2D).

Objectives: We aimed to investigate whether the food-T2D association would vary by genetic susceptibility to metabolic traits.

Methods: We analyzed data from 9542 incident T2D cases and a subcohort of 12,477 participants nested within the 340,234-participant cohort recruited in 1991–1998 and followed up for 10.9 y on average in 8 European countries. Polygenic risk scores (PRSs) for higher body mass index, insulin resistance, and T2D were constructed. Fifteen dietary variables potentially associated with T2D, obtained with cohort-specific self-reported dietary assessment, were examined: fruits, green leafy vegetables, root vegetables, wholegrains, rice, legumes, nuts and seeds, fermented dairy, red meat, processed meat, fish, eggs and egg products, sugar-sweetened beverages, coffee, and tea. A cross-product term between each PRS and each food/beverage was evaluated by genotyping chip and country with Prentice-weighted Cox regression for incident T2D, and stratum-specific estimates were meta analyzed, followed by Benjamini–Yekutieli multiple-testing correction.

Results: Accounting for multiple tests of 3 PRSs × 15 dietary items, no evidence of statistical interaction was evident on either a multiplicative or additive scale, with exp(β for a multiplicative interaction) (95% confidence interval) ranging from 0.84 (0.64, 1.10) (root vegetables and PRS for T2D) to 1.45 (0.78–2.76) (fish and PRS for T2D).

Conclusions: Genetic susceptibility to high-risk metabolic traits did not modify the diet-T2D associations in European populations. Acknowledging the limitations of current PRS-based methods to detect gene–diet interactions, research should continue into the potential for precision nutrition and tailored food-based dietary guidance for T2D prevention.

Aims: Metformin has hitherto not been proven superior to other type 2 diabetes (T2D) medications for the prevention of organ complications. The aim of this study is to report baseline data and blinded interim analyses in the register-based randomised clinical trial (RRCT) SMARTEST, which compares metformin and the SGLT2 inhibitor dapagliflozin in early T2D. We also present learnings from the novel decentralised methodology of this first RRCT in diabetes care.

Materials and Methods: Participants with T2D since <4 years and without major cardiovascular or renal disease were included at 36 centres across Sweden between 2019 and 2023 at on-site visits or via remote inclusion using digital informed consent. Participants were randomised 1:1 to open-label dapagliflozin 10 mg/day or metformin at an individualised dose, and they are followed for 2–6 years, with blinding of researchers to endpoints per treatment arm. The composite primary endpoint is time to first event of: myocardial infarction, stroke, heart failure (MACE), appearance or progression of microvascular complications or all-cause death. These events are collected from the Swedish National Diabetes Register and the National Patient Register using automated extraction.

Results: A total of 2072 patients, mean age 61.2 years, 39% women, entered randomised treatment. Signs of nephropathy, retinopathy and foot-at-risk were found in 6.1%, 13.2%, and 5.7%, respectively. Hypertension was present in 64.4%, and dyslipidaemia in 57.1%. In blinded interim analyses at a mean follow-up time of 19.0 months, the preliminary event rate of the primary composite endpoint was 11.7/100 patient-years (py) in the whole study population, mainly driven by microvascular complications. In contrast, rates of cardiovascular events and all-cause death were 0.6 and 0.3/100 py, respectively.

Conclusions: This decentralised RRCT in newly onset T2D demonstrates a highly feasible option for large-scale trials in the primary care setting, enabling representative participant recruitment. Blinded interim analyses showed a low risk of MACE or death, but unexpectedly high rates of microvascular complications. Study completion is event-driven and is expected by January 2026. The study will challenge or reinforce the current metformin paradigm in early T2D. (EUDRA-CT number 2019-001046-17; EU number 2024-516228-33-00; ClinicalTrials.gov Identifier: NCT03982381).

Key features: 

• The Northern Sweden Health and Disease Study (NSHDS) was initiated in the mid-1980s. The NSHDS is a population-based prospective longitudinal cohort comprising >140 000 participants in the two northernmost regions in Sweden, Norrbotten and Västerbotten, with >240 000 blood samples and 1.5 million person-years of follow-up.
• The NSHDS includes three sub-cohorts: the Västerbotten Intervention Programme (VIP), the expanded Northern Sweden Monitoring of Trends and Determinants of Cardiovascular Disease (MONICA) Study, and the Mammography Screening Project (MSP). The VIP is both a community-based cardiometabolic intervention programme encouraging healthy lifestyle (targeting individuals 40, 50, and 60 years of age), and a corresponding research cohort. The MONICA is an observational study focusing on cardiovascular disease and its associated risk factors, recruiting individuals aged 25–74 years. The MSP recruited women attending mammography during 1995–2006. The NSHDS median participation age is 50 years (53% women).
• Most participants contribute data on health, lifestyle, anthropometric measures, blood pressure, blood lipids, and glucose tolerance, along with research blood samples that are fractionated, frozen within an hour of collection, and stored at –80°C. Linkage to registries, clinical cohorts, and biological tissue archives facilitates studies of well-characterized participants (often combined with intervention studies).
• Collaborations are encouraged. Additional information can be found at: info.brs@umu.se; https://www.umu.se/en/biobank

Background: The majority of individuals in Sweden with type 2 diabetes have their sole health care provider in primary health care. Metformin treatment often causes gastrointestinal side-effects. Our aim was to construct and validate a questionnaire assessing gastrointestinal symptoms before and after starting metformin treatment for type 2 diabetes.

Methods: In the Interaction Between Metformin and Microbiota (MEMO) study, 54 participants rated six gastrointestinal symptoms at baseline and after 2 months of metformin treatment in a questionnaire (measured change, i.e. the difference between assessments at these two time points), as well as direct assessment of perceived change in symptoms after 2 months in a separate validation questionnaire (reported change, i.e. how participants themselves have perceived the change between the same two time points). Spearman’s ρ was calculated and reported with its 95% CI.

Results: The agreement between reported and measured change of symptoms, measured as Spearman’s ρ, was above 0.4 for 4 out of 6 symptoms (poor appetite 0.60 [95% CI 0.39–0.75], loose stool or diarrhea 0.58 [95% CI 0.37–0.74], flatulence 0.45 [95% CI 0.21–0.64], and abdominal pain 0.45 [95% CI 0.20–0.65]). The agreement was lower for nausea and vomiting, although these were numerically above 75% in agreement, likely due to few symptomatic participants overall.

Conclusion: For common side-effect symptoms associated with metformin treatment, our study shows that symptom change measured as the difference between assessments at two different time-points was in overall agreement, validating the usability of the constructed questionnaire for metformin side-effects.

Aims: Plasminogen activator inhibitor-1 (PAI-1) is the main inhibitor of the fibrinolytic system and is mainly secreted from adipose tissue. It is associated with cardiovascular disease and has also been considered a possible early risk marker for type 2 diabetes. Here, we present the results of a large prospective study investigating PAI-1 levels in relation to incident type 2 diabetes mellitus.

Methods: We conducted a prospective incident case-referent study within the Västerbotten Intervention Programme (VIP). Data on cardiovascular risk factors, fasting plasma glucose (FPG) and 2-hour plasma glucose (2-hPG) were collected at baseline health examination 1990–2005. Blood samples were collected and stored for future analyses. Participants were followed and 484 cases developed type 2 diabetes. Referents without type 2 diabetes were matched for sex, age, and year of participation, n = 484. Baseline plasma samples were analysed for PAI-1. Subgroup analysis was performed for 201 cases and 201 matched referents with normal baseline glucose levels (FPG < 6.1 and 2hPG < 8.9 mmol/L).

Results: Elevated baseline levels of PAI-1 were associated with incident type 2 diabetes after adjustments for BMI, family history of diabetes, smoking status, hypertension, FPG and 2hPG (PAI-1; OR = 1.87, 95% CI: 1.06–3.29). A similar result was shown in the subgroup analysis with 201 participants who had normal glucose levels at time of the health examination (PAI-1; OR = 2.68, 95% CI: 1.03–6.95).

Conclusions: Elevated PAI-1 levels in non-diabetic persons precede the manifestation of type 2 diabetes and can be detected before an elevation of FPG or 2-hPG is observed.

Our aim was to provide new data on the incidence, prevalence, and secular trend of type 2 diabetes (T2D) in Sweden, specifically early-onset T2D. We followed the Swedish population 2006 to 2021 and calculated age-standardized incidence (per 100 000) and prevalence (%) of T2D (overall) and early-onset T2D (age 23-39 years) stratified by sex, region of birth, and educational level. We projected the future prevalence of early-onset T2D by combining observed trends with population projections. From 2006 to 2021, the prevalence of T2D rose from 4.87% to 7.50%, and incidence from 477 [95% confidence interval (CI) 471-482] to 574 (CI 568-579). Early-onset T2D incidence increased from 54 to 107 (4.7% annual rise; CI 3.7%-5.7%) during this period. Incidence of early-onset T2D was higher in individuals born outside Europe (211, CI 195-226 vs 89, CI 84-93 in 2021) or low education (204, CI 185-223 vs 71, CI 65-77 in 2021), but a rise in incidence was seen irrespective of educational level, region of origin, and sex. If the incidence of early-onset T2D continues to increase at the same pace, its prevalence is projected to increase from 0.64% in 2021 to 3.2% in 2050. While T2D incidence rose marginally in Sweden 2006 to 2021, there was a significant rise in early-onset T2D, seen across different socioeconomic characteristics, with prevalence more than doubling and incidence nearly doubling. This development calls for targeted preventive efforts.

Background: Leisure-time physical activity decreases the risk of type 2 diabetes. Whether occupational physical activity affects the risk of type 2 diabetes is still not fully understood. The primary aim of this study was to investigate the association between occupational physical activity and 10-year diabetes incidence in a general adult population in Northern Sweden. The secondary aim was to explore the moderating role of BMI on this association.

Methods: This population-based, longitudinal cohort study included 16,282 diabetes-free individuals aged 28–52 years who participated in a cardiovascular intervention programme in Northern Sweden, and who reported the same occupational physical activity level at baseline and at 10-year follow-up. Incident type 2 diabetes was diagnosed based on oral glucose tolerance testing or a register-based diagnosis. Occupational physical activity was self-reported and categorized as: a) Low: ‘Sedentary or standing’ or ‘Light but partly physically active’, b) Moderate: ‘Light and physically active’, or c) High: Sometimes physically strenuous or ‘Physically strenuous most of the time’. Odds ratios (OR) and 95% confidence intervals (CI) for incident diabetes were calculated using multivariable logistic regression analysis, adjusting for age, sex, smoking, education level, family history of diabetes, country of birth, intake of fruits and vegetables, leisure-time physical activity, prediabetes and BMI. Potential interactions between BMI category and T2D were tested using interaction terms in the multivariable model.

Results: Six hundred twenty-four individuals developed type 2 diabetes in the 10 years between the first visit and the follow-up. A significant moderation effect of BMI on occupational physical activity was found (p = 0.01). Having a low level of occupational physical activity, compared with a moderate level of occupational physical activity, was associated with an increased risk of incident type 2 diabetes in overweight and obese individuals (OR 1.46, 95% CI 1.09–1.96), but not in those with normal weight (OR 0.80, 95% CI 0.52–1.23). High level of occupational physical activity was not associated with type 2 diabetes (OR 1.12, 95% CI 0.82–1.54).

Conclusions: Low occupational physical activity was associated with incident type 2 diabetes in overweight and obese individuals. Public-health efforts may benefit from encouraging less sitting and standing and more light physical activity during the workday.

Background: Dietary diversity may affect type 2 diabetes (T2D) but no studies have examined protein diversity by source. We examined five diversity scores and the 10-year risk of T2D and effect modification.

Methods: A prospective study of 10 363 incident T2D cases and a representative sub-cohort of 13 937 individuals sampled from a cohort of 340 234 participants in eight European countries (1993-2007). Five diversity scores were derived from self-reported diet data (gr/day): diversity of food groups (range: 0-5); and diversity within subtype of vegetables (0-4); meat/alternatives (0-6); animal-protein (0-8); and plant-protein sources (0-5). Country-specific hazard ratios (HRs) and 95% confidence intervals (CIs) were obtained by using Prentice-weighted Cox regression and combined by using mixed-effects models. Models were stratified by sex (male/female) and obesity status (body mass index ≥ 30 kg/m2; waist circumference ≥ 88 cm for females and ≥102 cm for males).

Results: Daily intake of five food groups (versus up to three) was linked to lower T2D incidence overall [HR 0.86 (95% CI 0.75, 0.98)], in females [0.86 (0.77, 0.96)], and in people without central obesity [0.79 (0.70, 0.89)]. Three or more subtypes of plant protein were inversely associated with T2D overall [0.78 (0.65, 0.98)], in females [0.75 (0.62, 0.90)] and people without central obesity [0.82 (0.68, 1.00)]. Additionally, consuming three subtypes of vegetables was inversely associated with T2D overall [0.90 (0.83, 0.98)] and in males [0.85 (0.73, 0.99)].

Conclusion: Diabetes prevention may benefit not only from a diet consisting of five different food groups, but also from a diet that is diverse in plant-protein sources, with specific benefits for female Europeans and those without central obesity.