ABSTRACT
Aim
This Finnish nationwide retrospective 'real‐world' study evaluated the effects of oral semaglutide on glycaemic control, body weight, lipid profile, and liver enzymes in adults living with type 2 diabetes (T2D) and naïve to any GLP‐1 receptor agonists in Finland.
Materials and Methods
Adults living with T2D (n = 7249) who initiated oral semaglutide between April 1, 2021, and December 31, 2023, were identified from national registers and followed until June 30, 2024. Primary endpoints were changes in HbA1c and body weight from baseline to 182.5 and 365 days. Secondary endpoints included changes in LDL‐cholesterol, HDL‐cholesterol, total cholesterol, triglycerides, HDL/triglycerides ratio, and ALT.
Results
Of those initiating oral semaglutide, 65.8% remained on treatment at 180 days and 55.2% at 365 days. Among patients included in longitudinal trajectory modelling, mean HbA1c (n = 2784) decreased from 7.9% to 6.8% at 182.5 days (95% CI: −1.20 to −1.04; p < 0.001) and to 6.9% at 365 days (95% CI: −1.18 to −0.95; p < 0.001). Mean body weight (n = 832) declined from 106 to 100 kg at 182.5 days (95% CI: −7 to −5; p < 0.001) and to 98 kg at 365 days (95% CI: −10 to −6; p < 0.001). LDL‐cholesterol (n = 1402), and triglycerides (n = 975) were significantly reduced at 182.5 days with little further change at 365 days. Mean ALT (n = 1762) declined from 50 to 42 U/L at 182.5 days (95% CI: −10 to −6, p < 0.001) and to 37 U/L at 365 days (; 95% CI: −16 to −10; p < 0.001). While weight at baseline was lower at patients aged at least 75 years, the changes in HbA1c and weight were similar compared to patients under 75 years of age.
Conclusion
Oral semaglutide significantly improves glycaemic and lipid control, body weight, and liver enzymes in adults living with T2D treated in clinical practice.
Keywords: glycaemic control, liver, real‐world evidence, semaglutide, type 2 diabetes, weight management
1. Introduction
Oral semaglutide belongs to the class of glucagon‐like peptide‐1 receptor agonists (GLP‐1RA) and has been shown to reduce glycated haemoglobin (HbA1c), support weight loss, and decrease the risk of major cardiovascular events in persons living with type 2 diabetes (T2D) [1, 2, 3, 4, 5, 6]. In Finland, oral semaglutide was granted reimbursement in April 2021 for persons living with T2D with a body mass index (BMI) of at least 30 who have not achieved glycaemic targets despite prior treatment with at least two antidiabetic medications. In March 2022, the criteria were revised to include patients who had inadequate glycaemic control following one antidiabetic medication.
The efficacy of oral semaglutide has been examined in the global PIONEER phase III trial programme. This programme included various patient groups: those with early T2D managed by diet and exercise, individuals with more established but uncontrolled T2D on one to three oral antidiabetic drugs, and patients with advanced disease treated with insulin. Oral semaglutide lowered HbA1c by 1.0%–1.4%‐points at 26 weeks and reduced body weight by 2.5–4.7 kg at 52 weeks, being superior to Dipeptidyl Peptidase‐4 inhibitor, sitagliptin, and Sodium‐Glucose Co‐Transporter 2 inhibitor, empagliflozin [1, 2, 3, 4, 5]. Additionally, a clinical trial showed that oral semaglutide reduced major cardiovascular event risk more than placebo in people with T2D and atherosclerotic cardiovascular disease, chronic kidney disease, or both [6].
Several prospective, multicentre, non‐interventional studies from the PIONEER REAL programme have demonstrated significant 'real‐world' benefits of oral semaglutide in reducing both HbA1c and body weight. Across these studies, oral semaglutide reduced HbA1c levels by 0.9%–1.2%‐points and body weight by 4.4–7.2 kg [7, 8, 9, 10, 11]. In addition, a few retrospective 'real‐world' studies have reported consistent findings [12, 13, 14]. However, since many of these studies had limited patient numbers for repeated HbA1c and body weight measurements or were conducted with selected clinics or databases, the findings of these studies may not be fully generalizable to the broader T2D population.
T2D is more prevalent in older adults, with prevalence rates peaking after the age of 80 [15]. Consequently, this ageing and potentially frail population constitutes a substantial proportion of the overall T2D population. However, evidence on the treatment outcomes of oral semaglutide in the older T2D population remains scarce. In many of the prospective PIONEER REAL studies and retrospective analyses, the average age of participants has been below 60 years [9, 10, 13]. An Italian 'real‐world' study (n = 101) demonstrated that oral semaglutide significantly reduced both HbA1c and body weight after 6 months of treatment in adults aged over 65 years with T2D. However, the small sample size and short follow‐up period of this study limit the ability to assess the long‐term benefits of treatment.
The aim of this study was to evaluate the 'real‐world' effectiveness of oral semaglutide using a nationwide cohort of GLP‐1RA–naïve adults living with T2D in Finland. The study assessed changes in HbA1c, body weight, lipid profile, and liver enzymes, and separately analysed changes in HbA1c and body weight among patients aged over 75 years.
2. Materials and Methods
2.1. Study Design
This nationwide retrospective observational study utilised individual‐level data from national registers in Finland. The study population included adult persons living with T2D who had a reimbursement for oral semaglutide and were GLP‐1RA–naïve. Reimbursement criteria included inadequate glycaemic control following one antidiabetic medication and BMI over 30. The index date was defined as the date of the first recorded purchase of oral semaglutide between April 1, 2021, and December 31, 2023. GLP‐1RA–naïve status was defined as having no GLP‐1RA purchases for at least 1 year prior to the index date (from April 1, 2020, onward). This time frame was selected based on the reimbursement effective date of oral semaglutide in Finland, which was April 1, 2021. During the study period, the approved and available doses of oral semaglutide in Finland were 3, 7 and 14 mg once daily.
The pre‐index period, defined as the 365 days prior to the index date, was used to capture patients' medication history and comorbidities. Data on the first recorded T2D diagnosis and reimbursed T2D medication purchase, used to determine T2D duration, were collected from 1.1.2000 onwards.
The baseline period—defined as the 90 days before the index date—was used to assess changes in clinical outcomes relative to the follow‐up period. Follow‐up assessments were conducted at 182.5 and 365 days post‐index.
2.2. Data Sources
Nationwide data on patients' medical records from secondary and primary care visits, as well as inpatient admissions, were collected from the HILMO (secondary care) and AvoHilmo (primary care) registers maintained by the Finnish Institute for Health and Welfare (THL). Information on whether a patient had received reimbursement for medication was obtained from the Medication Reimbursement Register of the Social Insurance Institution of Finland (Kela). Data on medication purchases were collected from Kela's Prescription Centre register. Body weight measurements and laboratory results from both primary and secondary care were obtained from the Kanta patient records register maintained by Kela.
Finnish Social and Health Data Permit Authority Findata has given a permit for this research (THL/2367/14.02.00/2024). Prior to analysis, Finnish Social and Health Data Permit Authority Findata combined data from different registers using a unique identifier, pseudonymised data and ensured the anonymity of the results under the Act on the Secondary Use of Health and Social Data (552/2019). All analyses were carried out within Findata's secure platform for processing individual‐level data.
2.3. Study Population
This study included adult persons (≥ 18 years) living with T2D (ICD code: E11) who received reimbursement for and had at least one purchase of oral semaglutide between 1.4.2021 and 31.12.2023 without prior use of any other GLP‐1RA medication. Patients with any purchase of a GLP‐1 receptor agonist during the pre‐index period (365 days prior to the index date) were excluded. Individuals diagnosed with T1D (E10) or gestational diabetes (ICD‐10 code: O24.4) were also excluded. All patients were followed until 30.6.2024 to ensure a minimum follow‐up period of 180 days.
2.4. Outcomes
Primary outcomes were changes in HbA1c and body weight from baseline to 182.5 and 365 days. Secondary outcomes included changes in lipid parameters (LDL‐cholesterol, HDL‐cholesterol, total cholesterol, triglycerides, HDL/triglycerides ratio) and alanine aminotransferase (ALT). Longitudinal changes were analysed using linear mixed‐effects models with random intercepts to account for within‐subject correlation. Patients with at least one baseline measurement (−90 to 0 days) and one follow‐up measurement (0–12 months) were included. Maximum one (last) measurement was included per patient per 90‐day interval. Time since treatment initiation was modelled using natural cubic splines to account for potential non‐linear trends during follow‐up. To address potential selection bias from differential outcome availability, stabilised inverse probability weights were estimated using logistic regression including age and sex and T2DM duration. Weight distributions and covariate balance were assessed, and weighted models were used as the primary analysis. Sensitivity analyses included models with time specified as a categorical variable in 90‐day intervals and unweighted models. Analyses were restricted to patients who maintained oral semaglutide treatment during the respective follow‐up period. The oral semaglutide treatment was considered to be discontinued if a patient has been without an oral semaglutide purchase for over 4 months. HbA1c and body weight changes were separately analysed in participants aged 75 or older as well as for women and men separately. In addition, a sensitive analysis was performed using intention‐to‐treat approach for HbA1c and weight trajectories. Additionally, time to treatment discontinuation for oral semaglutide and GLP‐1RA therapy was compared between individuals younger than 75 years and those aged 75 years or older. LDL‐cholesterol changes were also measured separately for patients on lipid‐lowering medication at baseline and those who were not.
2.5. Statistical Analysis
Descriptive statistics are reported as means with 95% confidence intervals (CIs) for continuous variables and patient counts with percentages for categorical variables. Changes in clinical outcomes from baseline were calculated as the difference between model‐based estimates and the observed mean baseline HbA1c. Statistical significance of changes from baseline was assessed using the two‐sided Wald test. Locally weighted smoothing (LOESS) was applied to assess non‐linear temporal trends in clinical outcomes. For analysis, measurements of the outcomes were aggregated into 90‐day intervals covering the period from 180 days prior to 365 days after the index date. To avoid overrepresentation of single individuals with more frequent testing, only one measurement (the latest) per patient was retained within each window. For the triglyceride‐to‐HDL ratio, only measurements taken on the same day were included. Trends were visualised as smoothed curves with 95% CIs. All analyses were conducted in R (R Foundation for Statistical Computing, Vienna, Austria).
3. Results
3.1. Study Cohort and Proportion of Patients With Measurements Over 12 Months Follow‐Up
A total of 7249 GLP‐1RA–naïve patients with T2D who initiated oral semaglutide treatment between April 1, 2021, and December 31, 2023, were included in the study and followed until the end of the study period on June 30, 2024. Nearly two thirds of the patients (65.8%, n = 4768) received oral semaglutide continuously for 180 days follow‐up, and at 365 days of follow‐up, the continuation rate was 55.2%. The proportion of patients who had at least one HbA1c, body weight, lipids, and ALT measurement or had measurement both at baseline and follow‐up and were included in longitudinal trajectory modelling is shown in Figure 1. The number of patients with each outcome measurement in each 3‐month window over 12 months follow‐up period is shown in Table S1. Standardised mean differences for baseline covariates—including age, sex, and T2D duration—in the HbA1c and weight cohorts, before and after inverse probability weighting, are shown in Table S2.
FIGURE 1.
Flowchart for patient selection for the study and proportion of patients with at least one measurement and one at baseline and follow‐up period for HbA1c, body weight, lipids and ALT.
3.2. Patient Characteristics
Baseline characteristics are presented in Table 1. The mean age was 61.4 years with 12.7% aged 75 or older. Half (53%) of the patients were male and 37.0% of patients had a T2D duration of less than 5 years. The mean HbA1c level was 7.9%, and the mean body weight was 105 kg. The majority of patients were receiving antihypertensive medication (84.3%) and/or lipid‐lowering therapy (73.3%). The most commonly used T2D medication prior to initiating oral semaglutide was an SGLT2 inhibitor ±metformin (29.4%), followed by a combination of a DPP‐4 inhibitor and an SGLT2 inhibitor ± metformin (22.0%).
TABLE 1.
Baseline characteristics of adults living with T2D treated with oral semaglutide.
| Characteristics | Patients n = 7249 |
|---|---|
| Sex, n (%) | |
| Female | 3384 (46.7%) |
| Male | 3865 (53.3%) |
| Mean age in years (SD) | 61.4 (11.8) |
| Age groups, n (%) | |
| 18–39 | 350 (4.8%) |
| 40–64 | 3827 (52.8%) |
| 65–74 | 2152 (29.7%) |
| 75 | 920 (12.7%) |
| Duration of T2D in years, n (%) | |
| < 5 | 2680 (37.0%) |
| 5–10 | 1981 (27.3%) |
| 10–15 | 1630 (22.5%) |
| > 15 | 958 (13.2%) |
| HbA1c in % | |
| Mean (SD) | 7.9 (1.3) |
| Missing, n (%) | 2937 (40.5%) |
| HbA1c level, n (%) | |
| > 8 | 1548 (35.9%) |
| 7.1–8 | 1762 (40.9%) |
| 6.6–7 | 606 (14.1%) |
| ≤ 6.5 | 396 (9.2%) |
| Body weight in kg | |
| Mean (SD) | 105 (17) |
| Missing, n (%) | 5083 (70.1%) |
| LDL cholesterol in mmol/L | |
| Mean (SD) | 2.4 (1.0) |
| Missing, n (%) | 4138 (57.1%) |
| Total cholesterol in mmol/L | |
| Mean (SD) | 3.8 (1.6) |
| Missing, n (%) | 4705 (64.9%) |
| HDL cholesterol in mmol/L | |
| Mean (SD) | 1.1 (0.3) |
| Missing, n (%) | 4764 (65.7%) |
| Triglycerides in mmol/L | |
| Mean (SD) | 2.5 (2.1) |
| Missing, n (%) | 4768 (65.8%) |
| ALT in U/L | |
| Mean (SD) v | 45.7 (32.0) |
| Missing, n (%) | 3893 (53.7%) |
| Medication at baseline, n (%) | |
| Anti‐diabetic | |
| Metformin monotherapy | 1146 (15.8%) |
| DPP4 inhibitors ± metformin | 853 (11.8%) |
| SGLT2 inhibitors ± metformin | 2130 (29.4%) |
| DPP4 inhibitor and SGLT2 inhibitor ± metformin | 1594 (22.0%) |
| Basal insulin ± oral T2D medication | 1393 (19.2%) |
| None | 106 (1.5%) |
| Other | 27 (0.4%) |
| Other | |
| Antihypertensive medication | 6108 (84.3%) |
| Lipid lowering medication | 5312 (73.3%) |
| Oral steroids | 388 (5.4%) |
| Concomitant anti‐diabetic medication during oral semaglutide treatment, n (%) | |
| Metformin monotherapy | 1358 (18.7%) |
| DPP4 inhibitors ± metformin | 364 (5.0%) |
| SGLT2 inhibitors ± metformin | 2546 (35.1%) |
| DPP4 inhibitor and SGLT2 inhibitor ± metformin | 463 (6.4%) |
| Basal insulin ± oral T2D medication | 1222 (16.9%) |
| None | 1037 (14.3%) |
| Other | 259 (3.6%) |
| Prescribed starting dose of oral semaglutide, n (%) | |
| 3 mg | 6826 (94.2%) |
| 7 mg | 328 (4.5%) |
| 14 mg | 95 (1.3%) |
Abbreviations: ALT, alanine aminotransferase; DPP4, dipeptidyl peptidase‐4; HbA1c, glycated haemoglobin; HDL, high density lipoprotein; LDL, low density lipoprotein; SD, standard deviation; SGLT2, Sodium‐glucose cotransporter‐2; T2D, type 2 diabetes.
3.3. Oral Semaglutide Treatment
The most common concomitant anti‐diabetic medication during oral semaglutide use was an SGLT2i ± metformin (35.1%), followed by metformin background therapy (18.7%), and 14.3% of patients received oral semaglutide without concomitant anti‐diabetic medication. Few patients received a DPP4i (5.0%) as a background therapy with oral semaglutide (Table 1). Almost all patients initiated oral semaglutide at a dose of 3 mg (94.2%). The most commonly used dose during follow‐up was 14 mg, used by 56% of patients The median duration of continuous oral semaglutide use was 18.5 months (95% CI: 16.9–20.2). After 365 days of treatment, the continuation rate was 55.2%. Among those who discontinued oral semaglutide, one‐fourth switched to injectable semaglutide, and among these switchers, the discontinuation rate for semaglutide treatment was 23% at 365 days.
3.4. Glycaemic and Body Weight Trajectories From Baseline to 182.5 and 365 Days
The estimated changes in HbA1c and body weight following oral semaglutide treatment are presented for all adult patients with measurements in Figure 2A,B and for under and at least 75 years old separately in Figure 2C,D. For all adult patients mean HbA1c (n = 2784) significantly decreased from a baseline of 7.9% to 6.8% at 182.5 days and to 6.9% at 365 days following oral semaglutide treatment. The estimated change was −1.1%‐points (95% CI: −1.20 to −1.04, p < 0.001) at 182.5 days and −1.1%‐points (95% CI: −1.18 to −0.95, p < 0.001) at 365 days (Figure 2A). At baseline, the mean body weight of patients (n = 832) was 106 kg following oral semaglutide treatment, weight was significantly reduced to 100 kg, with an estimated change of −6 kg (95% CI: −7 to −5, p < 0.001) at 182.5 days, and to 98 kg, with a mean change of −8 kg (95% CI: −10 to −6, p < 0.001) at 365 days (Figure 2B). Sensitivity analyses were performed, including models with time specified as a categorical variable in 90‐day intervals and unweighted models. Results were consistent across both approaches. Model assumptions were evaluated using residual diagnostics, indicating approximate normality with minor tail deviations and a few outliers that did not affect the results. Additionally, a sensitivity analysis using an intention‐to‐treat approach was conducted to assess HbA1c (n = 3714) and weight (n = 1190) trajectories following oral semaglutide treatment. The results showed similarly significant decreases in both outcomes, although reductions were greater among those who continued oral semaglutide use during the follow‐up period (Figure S1).
FIGURE 2.
HbA1c and body weight trajectories from baseline to 182.5 and 365 days in patients with type 2 diabetes using oral semaglutide for adults (A, B) and for those under and at least 75 years old separately (C, D). *Baseline is observed, follow‐up measurements predicted.
The estimated changes in HbA1c and body weight were also significant for patients at least 75 years old. The mean baseline HbA1c was 8.0% (n = 349, 38% of total 920) and the estimated change in HbA1c was −0.9%‐points (95% CI: −1.1 to −0.7, p < 0.001) at 182.5 days and −0.8%‐points (95% CI: −1.0 to −0.6, p < 0.001) at 365 days. Mean body weight was 96 kg (n = 104, 11.3% of total) and estimated change in body weight was −4 kg (95% CI: −8 to −1, p < 0.001) in 182.5 days, and −8 kg (95% CI: −12 to −4, p < 0.001) in 365 days (Figure 2C,D; Table S3). No significant difference in time to treatment discontinuation was observed between individuals below 75 (n = 6329) and those aged 75 or older (n = 920) for oral semaglutide treatment or for those who initiated oral semaglutide and may have switched to other GLP‐1RA medications (Figure S2). The HbA1c and body weight changes were also analysed separately for women and men. The results showed that changes in HbA1c from baseline were similar between women (n = 1219) and men (n = 1565) (Figure S3). For body weight, women (n = 386) experienced greater reductions at both 182.5 days (−7 vs. −5 kg in men) and 365 days (−9 vs. −7 kg in men) compared to men (n = 446), despite having a lower baseline weight (99 vs. 112 kg), highlighting greater relative weight loss in women (Figure S3).
3.5. Lipid and Liver Profile Changes From Baseline to 182.5 and 365 Days
The estimated changes in lipid and ALT profiles following oral semaglutide treatment are presented in Table 2. LDL‐cholesterol (n = 1402) and triglyceride (n = 975) levels were significantly reduced at 182.5 days, with no further reduction observed at 365 days; however, the significant change compared to baseline was maintained. For total cholesterol (n = 999) the estimated significant change was larger at 182.5 days compared to 365 days. HDL‐cholesterol (n = 961) levels showed a significant increase only at 365 days, with no significant change observed yet at 182.5 days. HDL/Triglycerides ratio (n = 941) changed significantly from baseline to 182.5 and 365 days, with grater change 365 days.
TABLE 2.
LDL, total cholesterol, HDL, triglyceride, HDL/Triglyceride ratio and ALT trajectories over months 0, 6 and 12 months.
| Timepoint | Estimate (95% CI) | Change from baseline (95% CI) | p |
|---|---|---|---|
| LDL (mmol/L) at months 0, 6 and 12, n = 1402 (19.3%) | |||
| Baseline a | 2.5 (2.5, 2.6) | Reference | |
| Day 182.5 | 2.0 (2.0, 2.0) | −0.5 (−0.6, −0.5) | < 0.001 |
| Day 365 | 2.0 (2.0, 2.1) | −0.5 (−0.6, −0.4) | < 0.001 |
| Total cholesterol (mmol/L) at months 0, 6 and 12, n = 999 (13.8%) | |||
| Baseline a | 3.8 (3.7, 3.9) | Reference | |
| Day 182.5 | 3.3 (3.3, 3.4) | −0.5 (−0.7, −0.3) | < 0.001 |
| Day 365 | 3.6 (3.4, 3.7) | −0.3 (−0.5, −0.1) | < 0.001 |
| HDL (mmol/L) at months 0, 6 and 12, n = 961 (13.3%) | |||
| Baseline a | 1.1 (1.11, 1.14) | Reference | |
| Day 182.5 | 1.1 (1.12, 1.14) | 0.01 (−0.02, 0.03) | 0.088 |
| Day 365 | 1.2 (1.16, 1.19) | 0.05 (0.02, 0.08) | < 0.001 |
| Triglycerides (mmol/L) at months 0, 6 and 12, n = 975 (13.5%) | |||
| Baseline a | 2.9 (2.70, 3.00) | Reference | |
| Day 182.5 | 2.1 (2.04, 2.17) | −0.74 (−0.95, −0.52) | < 0.001 |
| Day 365 | 2.1 (1.93, 2.19) | −0.79 (−1.07, −0.51) | < 0.001 |
| HDL/Triglycerides ratio at months 0, 6 and 12 months, n = 941 (13.0%) | |||
| Baseline a | 3.0 (2.70, 3.20) | Reference | NA |
| Day 182.5 | 2.1 (1.95, 2.14) | −0.90 (−1.25, −0.56) | < 0.001 |
| Day 365 | 1.9 (1.69, 2.06) | −1.07 (−1.51, −0.63) | < 0.001 |
| ALT (U/L) at months 0, 6 and 12, n = 1762 (24.3%) | |||
| Baseline a | 50 (49, 52) | Reference | |
| Day 182.5 | 42 (41, 43) | −8 (−10, −6) | < 0.001 |
| Day 365 | 37 (36, 39) | −13 (−16, −10) | < 0.001 |
Baseline is observed, follow‐up measurements predicted.
ALT levels significantly reduced both at 182.5 and 365 days (n = 1762), with the greater decrease observed at 365 days (estimated change −13.0; 95% CI: −16 to −10, p < 0.001) as compared to 182.5 days (estimated change −8; 95% CI: −10 to −6, p < 0.001).
Given that nearly two thirds of patients (73.3%) were receiving lipid‐lowering therapy at baseline, LDL‐cholesterol levels were stratified by baseline lipid‐lowering treatment status. At baseline, the mean LDL‐cholesterol level was 2.3 mmol/L in patients on lipid‐lowering therapy (n = 1053) and 3.3 mmol/L in those not receiving such treatment (n = 348). At 182.5‐day follow‐up, estimated LDL‐cholesterol levels were 1.8 mmol/L (mean change: −0.5 mmol/L; 95% CI: −0.6 to −0.4, p < 0.001) in the treated group and 2.2 mmol/L (mean change: −1.2 mmol/L; 95% CI: −1.3 to −1.0, p < 0.001) in the untreated group. For patients not receiving lipid‐lowering medication, LDL cholesterol decreased further to 2.0 mmol/L, whereas no additional decline at 365 days was observed in patients receiving lipid‐lowering treatment (Table S4).
3.6. Trend Analysis of HbA1c, Body Weight, Lipid and Liver Profiles
HbA1c, LDL‐cholesterol, and total cholesterol levels declined sharply during the first 90 days after treatment initiation, with only modest additional reductions observed between 180 and 365 days. In contrast, body weight, triglycerides, triglycerides/HDL‐cholesterol ratio and ALT levels showed a slower but continuous decrease throughout the 365‐day follow‐up period. HDL‐cholesterol levels initially declined but gradually increased, reaching values slightly above baseline by day 360 post‐treatment (Figure 3).
FIGURE 3.
Progression of HbA1c (A), weight (B), LDL (C), HDL (D), total cholesterol (E), triglycerides (F), ALT (G), and triglycerides/HDL ratio (H) relative to the index date. Changes in measurements are presented as smoothed linear trends with 95% confidence intervals, using a 90‐day time window.
4. Discussion
In this retrospective Finnish nationwide 'real‐world' study, the effects of oral semaglutide on glycaemic control, body weight, lipid profile, and liver enzymes were evaluated in people living with T2D. The main outcomes, HbA1c and body weight, showed significant reductions after 6 months of treatment. While body weight continued to decline through 12 months, HbA1c levels did not decrease further beyond 6 months. Among lipid markers, LDL‐cholesterol and triglycerides were significantly reduced to 6 months, with no further decline by 12 months. In contrast, HDL cholesterol increased significantly first after 12 months of oral semaglutide treatment. Liver enzyme (ALT) levels were significantly reduced to 6 months, with a further decline at 12 months.
In our study, we evaluated the effects of oral semaglutide on trajectories of metabolic parameters over a 12 months period. Estimated changes were assessed at two follow‐up time points: 6 and 12 months after treatment initiation. For trend analysis, we applied a 90‐day window. Because the data were derived from routine clinical practice rather than collected prospectively for research purposes, the proportion of patients with available measurements both at baseline and 12 months follow‐up was substantially lower than the total cohort size. Measurement coverage also varies across clinical parameters, as not all assessments are performed simultaneously in routine care. Nonetheless, the number of patients included in trajectory analysis—particularly for HbA1c—was higher in our study than in previously published studies. To avoid bias that patients with measurement would represent better monitored and adherent patients, inverse probability weighting based on age, sex, and diabetes duration was applied to account for differences in measurement availability.
The baseline HbA1c level of our patients was 7.9%, which was comparable to that reported in other studies. In our study, we observed an estimated reduction of 1.1%‐points in HbA1c after 6 months of oral semaglutide treatment, consistent with findings from other 'real‐world' studies and reaching the HbA1c target of < 7% [7, 8, 9, 10]. In the PIONEER clinical trials, the reduction in HbA1c was slightly greater than in 'real‐world' outcomes, ranging from 1.2% to 1.3%. However, in these trials baseline HbA1c levels were higher and background T2D medication was less heterogeneous due to differences in inclusion criteria compared with our study, which may explain differences [1, 2, 3, 4, 5, 6]. Additionally, patients in clinical trials are monitored more closely and receive more continuous care compared with routine practice, which may also contribute to better outcomes. In addition, in these clinical trials, oral semaglutide was administered at a 14 mg dose, whereas our study also included adults with T2D who used 3 mg and 7 mg doses. Our results showed that after 12 months of treatment, the reduction in HbA1c was 1.1%‐points, indicating that the greatest decline occurred within the first 6 months. Consistently, our trend analysis demonstrated that the most substantial reduction in HbA1c took place within the first 90 days after the initiation of oral semaglutide.
The mean body weight of our study cohort was 106 kg, with an estimated reduction of 6 kg at 6 months and 8 kg at 12 months suggesting that weight loss may continue beyond 12 months. In comparison, previous 'real‐world' studies have reported an average weight reduction of around 3 kg over 6–12 months of follow‐up, which is lower than in our study [16]. Similarly, clinical trials have shown smaller reductions by 12 months, ranging from 2.6 to 4.4 kg [1, 17, 18]. However, it is important to consider the baseline weight of study populations. In both 'real‐world' studies [7, 8, 9] and clinical trials [1, 17, 18] where weight reductions were smaller, the mean initial body weight was approximately 10 kg or more lower than in our cohort. By contrast, studies with similar baseline weights demonstrated comparable results. For example, in the PIONEER REAL Netherlands study (mean body weight 102.6 kg, n = 187), the mean reduction at 38 weeks was 5.8 kg, closely resembling our findings [10]. We also found that women experienced greater weight loss than men, indicating sex differences in treatment response. This is consistent with previous studies on GLP‐1 receptor agonists, which have also reported sex differences in treatment outcomes [19].
A significant proportion of people living with T2D are 75 years of age or older. However, the 'real world' outcomes of oral semaglutide in the older population remain limited. A retrospective study from Italy (n = 101) found that oral semaglutide led to a reduction of 0.4%‐points in HbA1c and 3.1 kg in weight after 6 months of treatment among adults living with T2D aged over 65 [20]. In our cohort, 920 patients were aged 75 years or older. We confirmed previous evidence showing that oral semaglutide was associated with a 0.9%‐point reduction in HbA1c among these patients (n = 349, 38% of the total n = 920) and a 4 kg decrease in body weight (n = 104, 11% of the total n = 920) after 6 months of treatment. At 12 months, the HbA1c reduction was 0.8%‐point and body weight loss 8 kg. These results are consistent with those of the overall study cohort, indicating that oral semaglutide treatment outcomes are similar in the older population. Additionally, in our study treatment duration did not differ between patients below 75 and those aged 75.
Our study showed that oral semaglutide treatment improved the lipid profile. LDL cholesterol and triglycerides were significantly reduced at 6 months after treatment initiation, with no further decline observed up to 12 months. These findings are consistent with results from clinical trials with 12‐month follow‐up [6, 21]. HDL‐cholesterol showed a significant increase after 12 months of treatment. Triglycerides to HDL‐cholesterol ratio that has been reported to be associated with insulin resistance, obesity and risk for cardiovascular diseases [22] decreased in our study from 3.0 to 1.9 after 12 months of oral semaglutide treatment. Additionally, oral semaglutide significantly reduced LDL cholesterol levels in patients both with and without concomitant lipid‐lowering therapy. Given that the effects of weight reduction on LDL‐cholesterol levels vary and are not consistent [23], our findings cannot rule out the possibility that the LDL‐cholesterol–lowering effect of oral semaglutide is independent of weight reduction and may involve mechanisms different from those of other lipid‐lowering medications; however, this requires further research.
Current 'real‐world' evidence regarding the impact of oral semaglutide on liver enzymes particularly within European populations with T2D remains scarce, with limited patient numbers [24]. ALT is a biomarker used to detect chronic liver diseases, particularly in high‐risk populations such as individuals with T2D and obesity. Our results showed that oral semaglutide treatment significantly lowered ALT levels at 6 months, with a further decline observed at 12 months. Other liver‐related variables, including the FIB‐4 score and diagnosis of metabolic dysfunction‐associated steatotic liver disease (MASLD), were not available in the national register data. Therefore, the 'real‐world' outcomes of oral semaglutide use on reducing risk for MASLD require further research with a larger patient cohort.
In our cohort, the risk of discontinuing oral semaglutide after 1 year was 45%. For once‐weekly injectable semaglutide, a discontinuation rate of 33% among adults with T2D was reported in a U.S. 'real‐world' study [25]. In our study among those who switched from oral to injectable semaglutide, the discontinuation rate for semaglutide was 23% within the same period. This is consistent with the Danish 'real‐word' study reporting 21.2% discontinuation rate at 1 year with GLP‐1RA users with T2D [26]. Although this rate is lower than that reported in a similar study from the United States, where approximately half of the patients discontinued treatment after 1 year [27], one in five patients still represents a substantial proportion. Reasons for discontinuation or for switching from oral to injectable semaglutide could not be addressed herein, but persistence in the treatment need to be addressed in order to tackle this one of the major problems of GLP‐1RA treatment.
Strength of our study is that we utilised national registry data from Finland, offering comprehensive nationwide coverage of adults living with T2D treated with oral semaglutide, thereby representing the average T2D population. Previous research, in contrast, has utilised data from prospective studies, selected healthcare centres, or claims‐based databases [7, 8, 9, 10, 11, 13, 14]. To our knowledge, this is also the first study to include liver enzyme (ALT) measurements alongside HbA1c, body weight, and lipid profile, with a follow‐up of up to 12 months. One of the strengths of our study is also that we performed a subgroup analysis for adults aged 75 years and older, a population that is underrepresented in clinical trials as well as in other ‘real world’ studies.
Our study has limitations. As a retrospective ‘real‐world’ study, the data were based on routine clinical measurements, which have not been performed consistently over 12‐month periods. Consequently, metabolic parameters at baseline and follow‐up (12 months) were available for only 11.5%–38.4% of patients, depending on the laboratory test and time point. In addition, not all clinical parameters, particularly body weight, were available for all the patients, either because they were not measured regularly in healthcare or not documented in a structured format transferable to national registers. Because the study did not include a control group and potential confounding variables, including concomitant medication changes and lifestyle factors, were not accounted for in the analysis, definitive causal conclusions cannot be drawn. Due to the local reimbursement requirements including inadequate therapeutic balance with at least one T2D medication and BMI over 30, only adults living with T2D having BMI over 30 were included in the study, which may limit the generalizability of the findings to T2D patients with lower BMI. Also, other liver‐related parameters such as FIB‐4 score or diagnosis of MASLD were not available in the dataset reflecting their low usage in clinical practice. Finally, during the study period, supply‐related availability issues with semaglutide products may have affected the number of patients initiating and switching oral semaglutide.
This nationwide Finnish study presents ‘real‐world’, population‐based outcomes for oral semaglutide in T2D. The results confirm the previous clinical trials and ‘real‐world’ results, demonstrating that oral semaglutide significantly lowers HbA1c, body weight, LDL‐cholesterol, total cholesterol, and triglycerides. The study highlights ongoing knowledge gaps regarding ‘real‐world’ effects on liver enzymes (ALT levels) and confirms the effectiveness of oral semaglutide in patients aged 75 years and older with T2D.
Author Contributions
All authors contributed to design, interpretation of results, and writing manuscript. Lalli Nurmi and Mari Lahelma conducted the data analysis.
Funding
This work was supported by Novo Nordisk Pharma.
Conflicts of Interest
Leo Niskanen has received speaker honoraria from Amgen, Lilly, GSK, Boehringer Ingelheim, Novo Nordisk, Sanofi, MSD, and Astra and has participated in the scientific advisory boards of Amgen, Lilly, Boehringer Ingelheim, Astra, MSD, and Novo Nordisk.
Atte Vaden: Speakers fees from Amgen, AstraZeneca, Boehringer, Eli Lilly Novo Nordisk and Orion. Advisory board fees from Bayer and Novo Nordisk. Congress expenses AstraZeneca, Nordic Infucare, and Novo Nordisk.
Mari Lahelma, Lalli Nurmi, Eija Heikkilä: Employees of NHG.
Tero Saukkonen, Hannu Vessari and Hanne Luoto: Employees and shareholders of Novo Nordisk.
Supporting information
Figure S1: HbA1c and body weight trajectories from baseline to 182.5 and 365 days in patients with type 2 diabetes with intention to treat with oral semaglutide. *Baseline is observed, follow‐up measurements predicted.
Figure S2: Time to treatment discontinuation (TTD) in patients with type 2 diabetes aged below 75 or 75 years and older with (A) oral semaglutide and (B) oral semaglutide treatment including also switches to any other GLP‐1RA treatment.
Figure S3: HbA1c and body weight trajectories from baseline to 182.5 and 365 days in patients with type 2 diabetes using oral semaglutide for women (A, C) and men (B, D) *Baseline is observed, follow‐up measurements predicted.
Table S1: Number of patients with outcome measurements in each 3‐month window over follow‐up period for each measurement.
Table S2: Standardised mean differences (SMD) for baseline covariates in the HbA1c and weight cohorts before and after inverse probability weighting (IPW).
Table S3: HbA1c (%) and weight (kg) trajectories from baseline to 182.5 and 365 days for under and at least 75 years of age.
Table S4: LDL‐C trajectories from baseline to 182.5 and 365 days for those patients who received and did not receive lipid‐lowering medication at baseline.
Acknowledgment
Open access publishing facilitated by Ita‐Suomen yliopisto, as part of the Wiley ‐ FinELib agreement.
Vaden A., Lahelma M., Nurmi L., et al., “Nationwide Real‐World Retrospective Study of Oral Semaglutide Use in Adults Living With Type 2 Diabetes in Finland,” Diabetes, Obesity and Metabolism 28, no. 9 (2026): 7814–7824, 10.1111/dom.70920.
Handling Editor: Johan Jendle
Data Availability Statement
Data is available upon research permit and data request from national health data access body.
References
- 1. Rodbard H. W., Rosenstock J., Canani L. H., et al., “Oral Semaglutide Versus Empagliflozin in Patients With Type 2 Diabetes Uncontrolled on Metformin: The PIONEER 2 Trial,” Diabetes Care 42, no. 12 (2019): 2272–2281, 10.2337/dc19-0883. [DOI] [PubMed] [Google Scholar]
- 2. Aroda V. R., Bain S. C., Cariou B., et al., “Efficacy and Safety of Once‐Weekly Semaglutide Versus Once‐Daily Insulin Glargine as Add‐On to Metformin (With or Without Sulfonylureas) in Insulin‐Naive Patients With Type 2 Diabetes (SUSTAIN 4): A Randomised, Open‐Label, Parallel‐Group, Multicentre, Multinational, Phase 3a Trial,” Lancet Diabetes and Endocrinology 5, no. 5 (2017): 355–366, 10.1016/S2213-8587(17)30085-2. [DOI] [PubMed] [Google Scholar]
- 3. Rosenstock J., Allison D., Birkenfeld A. L., et al., “Effect of Additional Oral Semaglutide vs Sitagliptin on Glycated Hemoglobin in Adults With Type 2 Diabetes Uncontrolled With Metformin Alone or With Sulfonylurea: The PIONEER 3 Randomized Clinical Trial,” Journal of the American Medical Association 321, no. 15 (2019): 1466–1480, 10.1001/jama.2019.2942. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Mosenzon O., Blicher T. M., Rosenlund S., et al., “Efficacy and Safety of Oral Semaglutide in Patients With Type 2 Diabetes and Moderate Renal Impairment (PIONEER 5): A Placebo‐Controlled, Randomised, Phase 3a Trial,” Lancet Diabetes and Endocrinology 7, no. 7 (2019): 515–527, 10.1016/S2213-8587(19)30192-5. [DOI] [PubMed] [Google Scholar]
- 5. Pieber T. R., Bode B., Mertens A., et al., “Efficacy and Safety of Oral Semaglutide With Flexible Dose Adjustment Versus Sitagliptin in Type 2 Diabetes (PIONEER 7): A Multicentre, Open‐Label, Randomised, Phase 3a Trial,” Lancet Diabetes and Endocrinology 7, no. 7 (2019): 528–539, 10.1016/S2213-8587(19)30194-9. [DOI] [PubMed] [Google Scholar]
- 6. McGuire D. K., Marx N., Mulvagh S. L., et al., “Oral Semaglutide and Cardiovascular Outcomes in High‐Risk Type 2 Diabetes,” New England Journal of Medicine 392, no. 20 (2025): 2001–2012, 10.1056/NEJMoa2501006. [DOI] [PubMed] [Google Scholar]
- 7. Kick A., M'Rabet‐Bensalah K., Acquistapace F., et al., “Real‐World Use of Oral Semaglutide in Adults With Type 2 Diabetes: The PIONEER REAL Switzerland Multicentre, Prospective, Observational Study,” Diabetes Therapy 15, no. 3 (2024): 623–637, 10.1007/s13300-023-01525-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Catrina S. B., Amadid H., Braae U. C., et al., “PIONEER REAL Sweden: A Multicentre, Prospective, Real‐World Observational Study of Oral Semaglutide Use in Adults With Type 2 Diabetes in Swedish Clinical Practice,” Diabetes Therapy 15, no. 9 (2024): 2079–2095, 10.1007/s13300-024-01614-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Jain A. B., Reichert S. M., Amadid H., et al., “Use of Once‐Daily Oral Semaglutide and Associated Clinical Outcomes Among Adults With Type 2 Diabetes in Routine Clinical Practice in Canada: A Multicentre, Prospective Real‐World Study (PIONEER REAL Canada),” Diabetes, Obesity and Metabolism 26, no. 5 (2024): 1799–1807, 10.1111/dom.15493. [DOI] [PubMed] [Google Scholar]
- 10. van Houtum W., Schrömbges P., Amadid H., et al., “Real‐World Use of Oral Semaglutide in Adults With Type 2 Diabetes in the PIONEER REAL Netherlands Multicentre, Prospective, Observational Study,” Diabetes Therapy 15, no. 8 (2024): 1749–1768, 10.1007/s13300-024-01588-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. ElBadawi H., Albalkhi N., Al Kadhim I., et al., “PIONEER REAL Saudi Arabia: A Multicentre, Prospective, Real‐World Study of Once‐Daily Oral Semaglutide Use in Adults With Type 2 Diabetes in Saudi Arabia,” Diabetes, Obesity and Metabolism 28, no. 2 (2025): 1304–1314, 10.1111/dom.70317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Moreno‐Pérez O., Reyes‐Garcia R., Modrego‐Pardo I., et al., “Real‐World Effectiveness and Safety of Oral Semaglutide in People Living With Type 2 Diabetes: A Nationwide Multicentre Retrospective Observational Study (ENDO2 S‐RWD),” Diabetes, Obesity and Metabolism 26, no. 4 (2024): 1519–1523, 10.1111/dom.15431. [DOI] [PubMed] [Google Scholar]
- 13. Swift C., Frazer M., Gronroos N. N., et al., “Real‐World Glycated Haemoglobin Changes Among Type 2 Diabetes Mellitus Patients Treated With a Maintenance Dose of 7 mg or 14 mg of Oral Semaglutide,” Diabetes, Obesity & Metabolism 26, no. 6 (2024): 2111–2118, 10.1111/dom.15516. [DOI] [PubMed] [Google Scholar]
- 14. Postema A., Gaspersz J., Baak B. N., et al., “Real‐World Evaluation of Clinical Outcomes in Dutch Patients With Type 2 Diabetes Treated With Oral Semaglutide: A Retrospective, Observational Cohort Study Using the PHARMO Data Network,” Diabetes, Obesity and Metabolism 28, no. 2 (2025): 895–905, 10.1111/dom.70251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Khan M. A. B., Hashim M. J., King J. K., Govender R. D., Mustafa H., and Al Kaabi J., “Epidemiology of Type 2 Diabetes ‐ Global Burden of Disease and Forecasted Trends,” Journal of Epidemiology and Global Health 10, no. 1 (2020): 107–111, 10.2991/jegh.k.191028.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Marassi M. and Fadini G. P., “Real‐World Evidence on Oral Semaglutide for the Management of Type 2 Diabetes. A Narrative Review for Clinical Practice,” Clinical Therapeutics 47, no. 1 (2025): 102–110, 10.1016/j.clinthera.2024.11.005. [DOI] [PubMed] [Google Scholar]
- 17. Pratley R., Amod A., Hoff S. T., et al., “Oral Semaglutide Versus Subcutaneous Liraglutide and Placebo in Type 2 Diabetes (PIONEER 4): A Randomised, Double‐Blind, Phase 3a Trial,” Lancet 394, no. 10192 (2019): 39–50, 10.1016/S0140-6736(19)31271-1. [DOI] [PubMed] [Google Scholar]
- 18. Zinman B., Aroda V. R., Buse J. B., et al., “Efficacy, Safety, and Tolerability of Oral Semaglutide Versus Placebo Added to Insulin With or Without Metformin in Patients With Type 2 Diabetes: The PIONEER 8 Trial,” Diabetes Care 42, no. 12 (2019): 2262–2271, 10.2337/dc19-0898. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Yang Y., He L., Han S., et al., “Sex Differences in the Efficacy of Glucagon‐Like Peptide‐1 Receptor Agonists for Weight Reduction: A Systematic Review and Meta‐Analysis,” Journal of Diabetes 17, no. 3 (2025): e70063, 10.1111/1753-0407.70063. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Fiore V., Carbotta G., Barraco S., Falasca P., Aricò C. N., and Barucca A., “Real‐World Retrospective Study in Elderly Patients Aged 65 Years and Older With Type 2 Diabetes Mellitus Treated With Daily Oral Semaglutide (SEMA‐Elderly),” Diabetes, Obesity & Metabolism 27, no. 4 (2025): 1805–1814, 10.1111/dom.16174. [DOI] [PubMed] [Google Scholar]
- 21. Husain M., Birkenfeld A. L., Donsmark M., et al., “Oral Semaglutide and Cardiovascular Outcomes in Patients With Type 2 Diabetes,” New England Journal of Medicine 381, no. 9 (2019): 841–851, 10.1056/NEJMoa1901118. [DOI] [PubMed] [Google Scholar]
- 22. Kosmas C. E., Rodriguez Polanco S., Bousvarou M. D., et al., “The Triglyceride/High‐Density Lipoprotein Cholesterol (TG/HDL‐C) Ratio as a Risk Marker for Metabolic Syndrome and Cardiovascular Disease,” Diagnostics (Basel, Switzerland) 13, no. 5 (2023): 929, 10.3390/diagnostics13050929. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Bays H. E., Kirkpatrick C. F., Maki K. C., et al., “Obesity, Dyslipidemia, and Cardiovascular Disease: A Joint Expert Review From the Obesity Medicine Association and the National Lipid Association 2024,” Journal of Clinical Lipidology 18, no. 3 (2024): e320–e350, 10.1016/j.jacl.2024.04.001. [DOI] [PubMed] [Google Scholar]
- 24. Volpe S., Lisco G., Fanelli M., et al., “Oral Semaglutide Improves Body Composition and Preserves Lean Mass in Patients With Type 2 Diabetes: A 26‐Week Prospective Real‐Life Study,” Frontiers in Endocrinology 14 (2023): 1240263, 10.3389/fendo.2023.1240263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Uzoigwe C., Liang Y., Whitmire S., and Paprocki Y., “Semaglutide Once‐Weekly Persistence and Adherence Versus Other GLP‐1 RAs in Patients With Type 2 Diabetes in a US Real‐World Setting,” Diabetes Therapy 12, no. 5 (2021): 1475–1489, 10.1007/s13300-021-01053-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Lassen M. C. H., Johansen N. D., Modin D., et al., “Adherence to Glucagon‐Like Peptide‐1 Receptor Agonist Treatment in Type 2 Diabetes Mellitus: A Nationwide Registry Study,” Diabetes, Obesity & Metabolism 26, no. 11 (2024): 5239–5250, 10.1111/dom.15872. [DOI] [PubMed] [Google Scholar]
- 27. Rodriguez P. J., Zhang V., Gratzl S., et al., “Discontinuation and Reinitiation of Dual‐Labeled GLP‐1 Receptor Agonists Among US Adults With Overweight or Obesity,” JAMA Network Open 8, no. 1 (2025): e2457349, 10.1001/jamanetworkopen.2024.57349. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1: HbA1c and body weight trajectories from baseline to 182.5 and 365 days in patients with type 2 diabetes with intention to treat with oral semaglutide. *Baseline is observed, follow‐up measurements predicted.
Figure S2: Time to treatment discontinuation (TTD) in patients with type 2 diabetes aged below 75 or 75 years and older with (A) oral semaglutide and (B) oral semaglutide treatment including also switches to any other GLP‐1RA treatment.
Figure S3: HbA1c and body weight trajectories from baseline to 182.5 and 365 days in patients with type 2 diabetes using oral semaglutide for women (A, C) and men (B, D) *Baseline is observed, follow‐up measurements predicted.
Table S1: Number of patients with outcome measurements in each 3‐month window over follow‐up period for each measurement.
Table S2: Standardised mean differences (SMD) for baseline covariates in the HbA1c and weight cohorts before and after inverse probability weighting (IPW).
Table S3: HbA1c (%) and weight (kg) trajectories from baseline to 182.5 and 365 days for under and at least 75 years of age.
Table S4: LDL‐C trajectories from baseline to 182.5 and 365 days for those patients who received and did not receive lipid‐lowering medication at baseline.
Data Availability Statement
Data is available upon research permit and data request from national health data access body.