Abstract
Objectives
This study aimed to evaluate the longitudinal association between overbasalization and glycemic control.
Methods
Retrospective cohort study including subjects with T2DM diagnosis, insulin users and treated in the Endocrinology Service of EsSalud through the National Telemedicine Center, part of the Social Security (EsSalud) in Lima, Peru. Data was taken from January 2023 to December 2024, and only those with at least two assessments were considered. Hospitalized participants, chronic users of corticosteroids, those with transfusions, and pregnant women during the study period were excluded. The dependent variable was glycemic control, defined using glycated hemoglobin (HbA1c < 53 mmol/mol [< 7%] or ≥ 53mmol/mol [≥ 7%]), while the independent variable was overbasalization, defined as a basal insulin dose > 0.5 IU/Kg/day. The association of interest was evaluated cross-sectionally and longitudinally.
Results
A total of 398 records, mean age of 63.8 (SD: 10.4) years and 50.8% women, was analyzed. The mean duration of diabetes was 15.8 (SD: 8.7) years, whereas the median duration of insulin use was 3.3 (interquartile range: 6) years. The prevalence of overbasalization at the cohort baseline was 43.4%, whereas only 14.5% had appropriate glycemic control. In cross-sectional analysis, no association of overbasalization with glycemic control was found; however, in the longitudinal analysis (mean: 9.6 [SD: 3.5] months of follow-up), the group with overbasalization had a 42% reduction (95% CI: 20% − 58%) in the probability of having appropriate glycemic control.
Conclusions
Among insulin users, overbasalization was observed to be associated with reduced glycemic control and higher HbA1c levels. More than 40% of insulin users had overbasalization at the cohort baseline. Our study suggests that overbasalization is frequent among people using insulin and should be evaluated to optimize glycemic control.
Supplementary Information
The online version contains supplementary material available at 10.1007/s40200-025-01756-9.
Keywords: Diabetes mellitus type 2, Insulin, Glycemic control, Glycated hemoglobin
Background
According to the International Diabetes Federation (IDF), there is an estimation of 537 million people with the diagnosis of type 2 diabetes mellitus (T2DM) worldwide, and such numbers are predicted to reach 643 million by 2030 and 783 million by 2045 [1]. The prevalence of T2DM in the Americas region is around 8% [2], whereas in Peru, a systematic review reported that the T2DM prevalence has increased over time (from 5% to 7% between 2005 and 2012), and also two new cases per 100 people per year are diagnosed [3].
T2DM is associated with a series of micro- and macrovascular complications, which are perpetuated in people without appropriate glycemic control [4, 5]. Usually, the criteria established by the American Diabetes Association (ADA) are used to determine if a person has appropriate glycemic control, i.e., glycated hemoglobin (HbA1c) < 53mmol/mol (7%), fasting glucose between 80 and 130 mg/dl, or postprandial glucose < 180 mg/dl [6].
Some people cannot reach appropriate glycemic control by using oral antidiabetics and lifestyle changes alone. As a result, in these people it is necessary to intensify treatment, including the use of basal insulin as an option [7, 8]. The initiation of basal insulin is often late, and it starts when chronic complications of T2DM appear. However, according to literature, less than 30% of people treated with insulin achieve HbA1c < 53mmol/mol (7%) [5, 9–11]. In Peru, previous literature has shown that 32% [12], 44% [13], and 48% of people with diabetes achieved appropriate glycemic control [14], with heterogeneous results depending on type of participants, medication use, and study site (i.e., hospital-based vs. population-based studies).
Given the poor glycemic control among people with T2DM using insulin [9, 11], examining whether the basal insulin dose is appropriate to achieve the target HbA1c is needed. In this regard, ADA recommendations indicate to be careful with overbasalization when using insulin. Overbasalization is defined as a basal insulin dose greater than 0.5 IU/kg/day [15], and can be identified in the following clinical scenarios: basal insulin dose >0.5 units/kg/day, postprandial blood glucose >180 mg/dL or when HbA1c is not in appropriate levels despite fasting blood glucose values being within adequate range and the bedtime-to-morning differential (BeAM) is >50 mg/dL [16, 17]. A study at the Department of Family Medicine at the University of South Florida reported that the prevalence of overbasalization among people with T2DM who were insulin users was close to 40% [18]. Since the concept of overbasalization has been scantly evaluated in resource-constrained settings like Peru, this becomes a relevant topic to be studied. Previous studies, most of them cross-sectional or post hoc analyses, have evaluated that the reduction in fasting glucose and HbA1c levels become minimal as basal insulin increases [19, 20], whereas another study identified that overbasalization is a barrier to achieving appropriate glycemic control in T2DM [21].
Therefore, overbasalization could be highly prevalent in our context and may impact on HbA1c levels and, for instance, on glycemic control. Thus, this study aimed to evaluate the longitudinal association between overbasalization and glycemic control. Our hypothesis is that those people with T2DM, insulin users, with overbasalization will have a lower probability of achieving appropriate glycemic control.
Methodology
Study design
A retrospective cohort study was built using data from Electronic Health Records (EHR) from the Smart Health Service (EsSi) of the Peruvian Social Security (EsSalud). EsSalud, is a Peruvian public institution responsible for ensuring comprehensive healthcare services to insured individuals and their dependents, encompassing prevention, promotion, recovery, rehabilitation, and social and economic benefits related to health. EsSalud operates a network of hospitals, polyclinics, and specialized health facilities across Peru. The EsSi, implemented since 2019, is the official computer platform for the registration of all health beneficiaries in all the EsSalud institutions [22].
Study participants
The data of adult individuals (18+ years), with T2DM diagnosis, insulin users (i.e., new or continuers), and with complete HbA1c data (i.e., to assess glycemic control) and weight (i.e., to calculate basal insulin dose and overbasalization) were selected for the study. Patients who received care between January 2023 and December 2024, and with at least two assessments, one considered as baseline and at least one subsequent check-up in the following 12 months, were considered. If a participant had more check-ups during the study period, all these evaluations were included for analysis. Hospitalized participants, chronic users of corticosteroids, those with transfusions, and pregnant women during the study period were excluded.
Records were reviewed between January and May 2025. Data of these participants was taken from the registration system of the Endocrinology Service of the National Center for Telemedicine (CENATE in Spanish) which included level I to III hospitals of the three EsSalud healthcare networks in Lima, the capital of Peru: three from the Almenara Hospital Network, three from the Rebagliati Hospital Network and one from the Sabogal Hospital Network.
Sample size
The sample size was calculated using Epidat version 4.2 [23]. Assuming a 95% confidence level and 80% power, and a 15% incidence difference to be detected (i.e., 30% appropriate glycemic control in those with overbasalization [24] and 45% in those without overbasalization [19]) over approximately 6 or more months of follow-up, a total of 369 participants would be required (123 in the overbasalization group and 246 in the non-overbasalization group). Assuming a 5% loss to data and follow-up, the minimum sample size required was 388 participants (129 in the overbasalization group and 259 in the non-overbasalization group).
Variables
Outcome
Glycemic control, assessed using HbA1c levels (in mmol/mol and %) at baseline and defined in two categories: appropriate glycemic control (< 53 mmol/mol [7%]) and poor glycemic control (≥ 53mmol/mol [7%]) [15]. The same definition was used for all the follow-ups. As a sensitivity analysis, we also utilized the HbA1c levels in its continuous form (i.e., numerical) as outcome.
Exposure
Overbasalization, defined using only basal insulin dose, either NPH insulin or Glargine U100, >0.5 IU/kg/day [16] and categorized as present or absent (yes vs. no). This definition was made at the beginning of the cohort and was re-evaluated at each follow-up considering the participant’s weight at the time of the respective evaluation. When overbasalization was detected, the treating physician, usually an endocrinologist, indicated reduction of the insulin dose.
Covariates
Other variables included in the analysis as potential confounders were: sex (male vs. female); age (in years) and categorized into < 60, 60–69 and ≥ 70 years; duration of diabetes in years; time of insulin use (in years); type of insulin used (NPH vs. Glargine U100); whether the participant was a new insulin user (yes vs. no); body mass index (BMI) in categories according to international standards (normal [< 25 kg/m2], overweight [≥ 25 and < 30 kg/m2] and obesity [≥ 30 kg/m2]) [25] and history of arterial hypertension according to clinical records (yes vs. no).
In addition, other data were also collected for descriptive purposes, including insulin dose used (in IU/kg/day), prandial insulin use (yes vs. no), fasting glucose (mg/dL), oral antidiabetic used (metformin vs. glibenclamide), and HbA1c (in %); comorbidities, including hypothyroidism (yes vs. no), arrhythmia (yes vs. no), coronary artery disease (yes vs. no), cerebrovascular disease (yes vs. no), and neoplasia (yes vs. no); chronic complications of T2DM including neuropathy (yes vs. no), retinopathy (yes vs. no), and renal disease (yes vs. no); and biochemical markers such as total cholesterol (mg/dL), LDL cholesterol (mg/dL), triglycerides (mg/dL), creatinine (mg/dL), microalbuminuria (mg/24 hours), and hemoglobin (g/dL).
Statistical analysis
The information obtained was coded in an electronic template of Microsoft Excel 2010 for Windows using a double-entry table. This database was then transferred to STATA 16 for Windows (StataCorp, College Station, TX, USA) for analysis, considering a value of p < 0.05 as statistically significant. Initially, the numerical variables were described using means and standard deviation (SD) if normal distribution was verified, or median and interquartile range (IQR) if not, whereas absolute and relative frequencies were used in the case of categorical variables. In addition, the frequency and 95% confidence intervals (95% CI) of the variables of interest were reported. Then, the study population was described according to overbasalization groups and glycemic control (according to both definitions) and comparisons between variables were conducted using the Chi-square test for categorical variables and the Student t test for numerical variables.
The association between overbasalization and glycemic control at baseline was based on a complete-case analysis approach and evaluated using crude and adjusted Poisson regression models, including robust variance, and reporting prevalence ratios (PR) and 95% CI. The adjusted model included sex, age, years with diabetes, years of insulin use, insulin type, new insulin user, body mass index, and hypertension as potential confounders. These confounders were defined a priori according to previous studies reported in the literature [7–10]. On the other hand, to assess the association between overbasalization and glycemic control during follow-up, crude and adjusted Poisson regression models were similarly used, including robust variance and an indicator of follow-up time for each participant (in months) in the analysis. Thus, relative risks (RR) and 95% CI were reported. The list of confounders (sex, age, years with diabetes, years of insulin use, insulin type, new insulin user, body mass index, and hypertension) was considered a priori as described above.
As sensitivity analysis, the association between overbasalization and HbA1c levels (i.e., as numerical variable) over time was performed using linear mixed regression models. Both crude and adjusted models were created including an interaction term between overbasalization (exposure) and time (in months) to define the change in HbA1c levels (dependent variable). This hierarchical model had two levels: the first one defined by the HbA1c measurements over time, whereas the second level was defined by the subject (i.e., patient). An unstructured variance-covariance structure was utilized to avoid any assumption made by the investigators, and robust standard errors were used to correct for any model misspecification. Similarly, the list of confounders was considered a priori, as described above. Coefficients and 95% CIs are reported.
Ethics
This study adheres to the Declaration of Helsinki. The study protocol was reviewed and approved by the Institutional Ethics Committee of the Universidad Peruana Cayetano Heredia (code: 213434), the Institutional Ethics Committee of the Edgardo Rebagliati Martins Hospital, and authorized by the National Center for Telemedicine (CENATE).
Results
Description of the study population
A total of 895 records of T2DM patients receiving insulin therapy were identified during 2023. Of them, 416 patients were excluded for insufficient follow-up (only one assessment in a year), hospitalization, death during the study period, glucocorticoids use, exclusive use of prandial insulin or mixtures, or pregnancy. Thus, 479 patients met the inclusion criteria. After that, 81 records were subsequently discarded due to incomplete data on the variables of interest (i.e., HbA1c, initial insulin dose and weight), resulting in a final cohort of 398 patients included in the analysis, with a mean age of 63.8 (SD: 10.4) years and 202 (50.8%) women. The mean duration of diabetes was 15.8 (SD: 8.7) years, whereas the median duration of insulin use was 3.3 (interquartile range: 1–7) years.
On the other hand, arterial hypertension was the most frequent comorbidity present in 54.0% of the participants, whereas neuropathy (95.7%) was the most common complication.
Description of the population according to overbasalization
At baseline, 43.4% (95% CI: 38.5% – 48.5%) of subjects had overbasalization. In the bivariable analysis (Table 1and Supplementary Table 1), sex (p = 0.02), age (p < 0.001), type of insulin (p < 0.001), years of insulin use (p < 0.001), type of insulin used (p < 0.001) and being a new insulin user (p < 0.001) were associated with overbasalization status. In addition, in the overbasalization group, the mean insulin dose was 0.80 (SD: 0.2) IU/kg/day, but in the non-overbasalization group it was 0.3 (SD: 0.1) IU/kg/day (p < 0.001). Finally, the mean of HbA1c (80mmol/mol (9.5%) vs. 74mmol/mol (8.9%), p = 0.008) and triglycerides (219.2 mg/dL vs. 180.4 mg/dL, p = 0.03) were higher in those with overbasalization compared to those in the non-overbasalization group.
Table 1.
Description of the population according to baseline overbasalization
| Total | Overbasalization | p-value | ||
|---|---|---|---|---|
| (n = 398) | Yes (n = 173) | No (n = 225) | ||
| Sex [n, (%)] | 0.02 | |||
| Female | 202 (50.8) | 99 (57.2) | 103 (45.8) | |
| Age [n, (%)] | < 0.001 | |||
| < 60 years | 125 (31.4) | 72 (41.6) | 53 (23.6) | |
| 60–69 years | 151 (37.9) | 61 (35.3) | 90 (40.0) | |
| ≥ 70 years | 122 (30.7) | 40 (23.1) | 82 (36.4) | |
| Duration of T2DM, years [mean (SD)] | 15.8 (8.7) | 16.7 (8.5) | 15.1 (8.8) | 0.06 |
| Time of insulin use, years [median (IQR)] | 3.3 (6.0) | 5.0 (6.7) | 2.3 (4.6) | < 0.001 |
| Type of insulin used [n, (%)] | < 0.001 | |||
| NPH | 241 (60.6) | 143 (82.7) | 98 (43.6) | |
| Glargine U100 | 157 (39.4) | 30 (17.3) | 137 (56.4) | |
| Prandial insulin use [n, (%)] | 144 (36.4) | 88 (50.9) | 56 (25.1) | |
| New insulin user [n, (%)] | 39 (9.9) | 4 (2.3) | 35 (15.7) | < 0.001 |
| Insulin dose used (UI/Kg/day) [mean (SD)] | 0.5 (0.3) | 0.8 (0.2) | 0.3 (0.1) | < 0.001 |
| Fasting glucose (mg/dl) [mean (SD)] | 166.0 (76.9) | 171.8 (84.8) | 161.6 (70.2) | 0.20 |
| HbA1c in % [mean (SD)] | 9.2 (2.1) | 9.5 (2.1) | 8.9 (2.1) | 0.008 |
| Body mass index [n, (%)] | 0.36 | |||
| Normal | 95 (23.9) | 43 (24.9) | 52 (23.1) | |
| Overweight | 179 (44.9) | 71 (41.0) | 108 (48.0) | |
| Obesity | 124 (31.2) | 59 (34.1) | 65 (28.9) | |
| Oral antidiabetic used [n, (%)] | ||||
| Metformin | 267 (67.1) | 118 (68.2) | 149 (66.2) | 0.68 |
| Glibenclamide | 9 (2.3) | 3 (1.7) | 6 (2.7) | 0.54 |
| Comorbidities [n, (%)] | ||||
| Arterial hypertension | 215 (54.0) | 122 (54.2) | 93 (53.8) | 0.93 |
| Hypothyroidism | 58 (14.6) | 26 (15.0) | 32 (14.2) | 0.82 |
| Arrhythmia | 12 (3.0) | 5 (2.9) | 7 (3.1) | 0.90 |
| Coronary disease | 24 (6.0) | 6 (3.5) | 18 (8.0) | 0.06 |
| Stroke | 20 (5.0) | 6 (3.5) | 14 (6.2) | 0.21 |
| Neoplasm | 24 (6.0) | 8 (4.6) | 16 (7.1) | 0.30 |
| Late complications [n, (%)] | ||||
| Neuropathy | 381 (95.7) | 165 (95.4) | 216 (96.0) | 0.76 |
| Renal disease | 190 (47.7) | 86 (49.7) | 104 (46.2) | 0.49 |
| Retinopathy | 55 (13.8) | 28 (16.2) | 27 (12.0) | 0.35 |
Description of the population according to glycemic control
Only 14.5% (95% CI: 11.3% − 18.4%) had adequate glycemic control at the start of the study. In the bivariable analysis (Table 2and Supplementary Table 2), glycemic control was associated with age (p < 0.001), years of insulin use (p < 0.001), Hb1Ac levels (p < 0.001), fasting glucose (p < 0.001), history of arterial hypertension (p < 0.001), neoplasia (p < 0.001) and kidney disease (p = 0.02).
Table 2.
Description of the population according to glycemic control at the study baseline
| Total | Glycemic control | p-value | ||
|---|---|---|---|---|
| (n = 398) | Yes (n = 58) | No (n = 340) | ||
| Sex [n, (%)] | 0.47 | |||
| Female | 202 (50.8) | 32 (55.2) | 170 (50.0) | |
| Age [n, (%)] | < 0.001 | |||
| < 60 years | 125 (31.4) | 9 (15.5) | 116 (34.1) | |
| 60–69 years | 151 (37.9) | 18 (31.0) | 133 (39.1) | |
| ≥ 70 years | 122 (30.7) | 31 (53.5) | 91 (26.8) | |
| Duration of T2DM, in years [mean (SD)] | 15.8 (8.7) | 15.7 (10.7) | 15.8 (8.4) | 0.91 |
| Time of insulin use, years [median (IQR)] | 3.3 (6.0) | 3.0 (5.3) | 5.0 (9.3) | 0.02 |
| Type of insulin used [n, (%)] | 0.08 | |||
| NPH | 241 (60.6) | 29 (50.0) | 212 (62.4) | |
| Glargine U100 | 157 (39.5) | 29 (50.0) | 128 (37.7) | |
| Prandial insulin use [n, (%)] | 144 (36.4) | 20 (34.5) | 124 (36.7) | 0.75 |
| New insulin user [n, (%)] | 39 (9.9) | 6 (10.5) | 33 (9.8) | 0.86 |
| Insulin dose used (UI/Kg/day) [mean (SD)] | 0.5 (0.3) | 0.4 (0.2) | 0.5 (0.3) | 0.06 |
| Fasting glucose (mg/dl) [mean (SD)] | 166.0 (76.9) | 123.5 (56.0) | 173.3 (77.7) | < 0.001 |
| HbA1c in % [mean (SD)] | 9.2 (2.1) | 6.4 (0.4) | 9.6 (1.9) | < 0.001 |
| Body mass index [n, (%)] | 0.67 | |||
| Normal | 95 (23.9) | 13 (22.4) | 82 (24.1) | |
| Overweight | 179 (44.9) | 24 (41.4) | 155 (45.6) | |
| Obesity | 124 (31.2) | 21 (36.2) | 103 (30.3) | |
| Oral antidiabetic used [n, (%)] | ||||
| Metformin | 267 (67.1) | 36 (62.1) | 231 (67.9) | 0.38 |
| Glibenclamide | 9 (2.3) | 0 (0.0) | 9 (2.7) | 0.21 |
| Comorbidities [n, (%)] | ||||
| Arterial hypertension | 215 (54.0) | 43 (74.1) | 172 (50.6) | < 0.001 |
| Hypothyroidism | 58 (14.6) | 9 (15.5) | 49 (14.4) | 0.83 |
| Arrhythmia | 12 (3.0) | 1 (1.7) | 11 (3.2) | 0.53 |
| Coronary disease | 24 (6.0) | 3 (5.2) | 21 (6.2) | 0.77 |
| Stroke | 20 (5.0) | 3 (5.2) | 17 (5.0) | 0.96 |
| Neoplasm | 24 (6.0) | 9 (15.5) | 15 (4.4) | < 0.001 |
| Late complications [n, (%)] | ||||
| Neuropathy | 381 (95.7) | 56 (96.6) | 325 (95.6) | 0.74 |
| Renal disease | 190 (47.7) | 36 (62.1) | 154 (45.3) | 0.02 |
| Retinopathy | 55 (13.8) | 5 (8.6) | 50 (14.7) | 0.10 |
Association between overbasalization and glycemic control
At baseline, in cross-sectional analysis, overbasalization was not associated with glycemic control in multivariate analysis (PR = 0.75; 95% CI: 0.43; 1.30, Table 3). In contrast, during the 9.6 (SD: 3.5) months of follow-up and in multivariate analysis, those with overbasalization had a 42% (95% CI: 20% − 58%) reduction in the probability of having appropriate glycemic control (Table 4). In sensitivity analysis, using HbA1c levels as the dependent variable (Supplementary Table 3), we found: (a) at baseline, patients with overbasalization had lower HbA1c levels than without overbasalization, but this finding was not significant (β = −0.62; 95% CI: −1.45; 0.20); (b) over time, those without overbasalization at baseline showed an average 0.05% reduction in Hb1Ac levels per month of follow-up, but this change was not significant (β = −0.05; 95% CI: −0.11; 0.01); and (c) those with overbasalization at baseline showed a 0.09% increase in Hb1Ac levels per month of follow-up (β = 0.09%; 95% CI: 0.01; 0.17).
Table 3.
Association between overbasalization and baseline glycemic control: crude and adjusted models
| Crude model | Adjusted model* | |
|---|---|---|
| Overbasalization | PR (95% CI) | PR (95% CI) |
| No | 1 (Reference) | 1 (Reference) |
| Yes | 0.63 (0.38–1.05) | 0.75 (0.43–1.30) |
PR = Prevalence ratio
* Adjusted for sex, age, years with diabetes, years of insulin use, insulin type, new insulin user, body mass index, and high blood pressure
Table 4.
Association between overbasalization and glycemic control during follow-up: crude and adjusted models
| Crude model | Adjusted model* | |
|---|---|---|
| Overbasalization | RR (95% CI) | RR (95% CI) |
| No | 1 (Reference) | 1 (Reference) |
| Yes | 0.54 (0.40–0.72) | 0.58 (0.42–0.80) |
Estimates in bold are significant (p < 0.05); RR = Relative risk
* Adjusted for sex, age, years with diabetes, years of insulin use, insulin type, new insulin user, body mass index, and high blood pressure
Discussion
Main findings
In our study, more than 40% of T2DM patients using insulin had overbasalization, a finding similar to that reported in the literature [18]. In multivariable analysis, overbasalization was associated with a reduction in glycemic control in the longitudinal but not in the cross-sectional analysis. Similarly, our sensitivity analysis showed consistent results with differences in HbA1c levels during follow-up.
Comparison with previous studies
According to the literature, overbasalization is documented as a barrier to obtaining appropriate glycemic control [10]. The ADA included this term since 2018 and a later study such as that of Umpierrez et al showed a greater and linear glycemic effect with doses ≤ 0.3 IU/kg/day, a non-linear decreasing response with doses between 0.3 and 0.5 IU/kg/day, and a plateau effect with doses >0.5 IU/kg/day [20]. Another study, reanalyzing the data of 63 clinical trials evaluating insulin glargine between 1997 and 2007, reported that the adjusted mean HbA1C after 24 weeks of follow-up was 7.3% (56mmol/mol), 7.3% (56mmol/mol) and 7.5% (59mmol/mol) in those participants who exceeded the basal insulin dose cut-off points of 0.5, 0.7 and 1.0 IU/kg/day, respectively [19]. These estimates were significantly higher than those of HbA1c found in participants who were below the basal insulin cut-off points: 7.16% (55mmol/mol), 7.18% (55mmol/mol) and 7.20% (55mmol/mol), respectively (all p < 0.05). Thus, these results correlate with our findings where the outcome was measured as glycemic control and as change in HbA1c levels.
There are few studies in the literature evaluating the association between overbasalization and glycemic control. In this regard, the study by Stewart-Lynch et al. reported that no participant with doses >0.5 IU/kg/day achieved Hb1Ac levels < 53mmol/mol (7%); however, the result was not significant when compared with those without overbasalization [21]. This same study, despite its cross-sectional nature, also reported that achieving glycemic control was less likely with basal insulin doses >0.3 IU/kg/day. Although it is the most similar study to ours, it had a smaller sample size. In our study, the cross-sectional analysis was also not significant with a first definition of glycemic control (HbA1c < 53mmol/mol [< 7%]), but it was when using a second definition of glycemic control (HbA1c < 53mmol/mol [< 7%] in < 60 years and HbA1c < 64mmol/mol [< 8%] in those ≥ 60 years, data not shown). Finally, another study reanalyzing the results of a clinical trial evaluated the change in Hb1A1c from baseline to 12 months in five groups according to insulin use (non-insulin user, basal insulin alone – without overbasalization, basal insulin alone – with overbasalization, basal/bolus insulin – without overbasalization, and basal/bolus insulin – without overbasalization). The reduction in HbA1c levels was consistently lower in the groups with overbasalization [26], findings that go in the same direction as our results.
Public health relevance and implications
The present study provides new data on overbasalization, a term recently included in the American Diabetes guidelines. Many of the available reports are cross-sectional in nature or reanalyze results of clinical trials using HbA1c levels as a continuous outcome, not specifically designed to look for the association of interest (i.e., overbasalization and glycemic control). The present study allows us to know that in our country we also have a considerable percentage of people who use insulin with overbasalization (insulin dose > 0.5 IU/kg/day), a clear barrier to achieve glycemic control.
The results of the present study demonstrate the need for close monitoring from the start of basal insulin use to avoid overbasalization, but also to ensure appropriate glycemic control. Accordingly, we can propose strategies such as checking for the presence of postprandial hyperglycemia once we reach values close to 0.5 IU/kg/day and intensifying treatment with prandial insulin [16]. The use of additional medication may be another option, including glucagon-like peptide-1 (GLP-1) receptor agonists or sodium-glucose transport protein 2 (SGLT2) inhibitors. However, these medications are not always available due to the restrictive costs.
Strengths and limitations
The report is based on a longitudinal study that allowed us to estimate the association of interest, covering information taken from different hospitals in Lima. Nevertheless, this study has some limitations that deserve discussion. First, due to the observational and retrospective nature of the study, only association can be assessed, even though the study design allows us to rule out reverse causality. Besides, some clinical variables were not available (treatment adherence, prandial insulin use, insulin application technique, among others), which may be sources of residual confounding. Second, there may be a selection bias because a convenience sample of participants beneficiaries of EsSalud, using telemedicine system (CENATE), was used, and for instance, be not representative of the general population of Peru, which may affect the external validity of our findings. Moreover, patients included in this study may have longer duration of disease, had co-morbidities, and were insulin users, which can explain the high rate of complications, including neuropathy. Third, our sample size did not consider healthcare facility clustering, which may affect our power to detect differences as proposed in the study. Nevertheless, our results are in line with other manuscripts in literature. Finally, not all confounders were considered in the analysis. For example, not all types of basal insulin were evaluated, since only insulin glargine U100 and NPH insulin were included. Moreover, we did not use other criteria to define overbasalization as information about bedtime-to-morning differential or postprandial glucose was not available. In addition, indicators of diet, physical activity, or others that could affect the association of interest were not available. Despite that, our results are in line with previous studies, especially those derived from clinical trials.
Conclusions
The results of our study show that overbasalization at baseline was observed to be associated with poor glycemic control and increases of HbA1c levels. In addition, more than 40% of patients with T2DM using insulin had overbasalization. Our study suggests that overbasalization is frequent among people using insulin and should be evaluated to optimize glycemic control.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
Not applicable.
Author contributions
SAV-S and AB-O conceived the idea of the study. SAV-S conducted data collection and wrote the first version of the manuscript. AB-O performed statistical analysis and provided support and comments on the different versions of the manuscript. Both authors approved the version submitted for publication.
Funding
None to declare.
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethics approval
The study protocol was reviewed and approved by the Institutional Ethics Committee of the Universidad Peruana Cayetano Heredia (code: 213434), the Institutional Ethics Committee of the Edgardo Rebagliati Martins Hospital, and authorized by the National Center for Telemedicine (CENATE). This study adheres to the Declaration of Helsinki.
Consent to participate
Not applicable. For this analysis, we used deidentified clinical records. The data from clinical records were de-identified to ensure the anonymity of the participants.
Conflict of interest
None to declare.
Competing interests
The authors declare no competing interests.
Footnotes
The original online version of this article was revised: In this article, the author's name Antonio Bernabé-Ortiz was inadvertently listed twice.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Change history
11/1/2025
The original online version of this article was revised: In this article, the author's name Antonio Bernabé-Ortiz was inadvertently listed twice.
Change history
11/7/2025
A Correction to this paper has been published: 10.1007/s40200-025-01786-3
References
- 1.International Diabetes Federation. Diabetes Atlas 10th edition. Brussels, Belgium: IDF; 2021. [Google Scholar]
- 2.Organización Panamericana de la Salud. Panorama de La diabetes En La Región de Las Américas. Washington DC, US: OPS; 2023. [Google Scholar]
- 3.Carrillo-Larco RM, Bernabé-Ortiz A. [Type 2 diabetes mellitus in peru: A systematic review of prevalence and incidence in the general population]. Rev Peru Med Exp Salud Publica. 2019;36(1):26–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dal Canto E, Ceriello A, Rydén L, Ferrini M, Hansen TB, Schnell O, et al. Diabetes as a cardiovascular risk factor: an overview of global trends of macro and micro vascular complications. Eur J Prev Cardiol. 2019;26(2suppl):25–32. [DOI] [PubMed] [Google Scholar]
- 5.Mauricio D, Meneghini L, Seufert J, Liao L, Wang H, Tong L, et al. Glycaemic control and hypoglycaemia burden in patients with type 2 diabetes initiating basal insulin in Europe and the USA. Diabetes Obes Metab. 2017;19(8):1155–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.American Diabetes Association Professional Practice Committee. 6. Glycemic Goals and Hypoglycemia: Standards of Care in Diabetes-2025. Diabetes Care. 2025;48(Supplement_1):S128-s45. [DOI] [PMC free article] [PubMed]
- 7.Johnson EL, Frias JP, Trujillo JM. Anticipatory guidance in type 2 diabetes to improve disease management; next steps after basal insulin. Postgrad Med. 2018;130(4):365–74. [DOI] [PubMed] [Google Scholar]
- 8.Meece J. Basal insulin intensification in patients with type 2 diabetes: a review. Diabetes Ther. 2018;9(3):877–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Blonde L, Meneghini L, Peng XV, Boss A, Rhee K, Shaunik A, et al. Probability of achieving glycemic control with basal insulin in patients with type 2 diabetes in real-world practice in the USA. Diabetes Ther. 2018;9(3):1347–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dalal MR, Grabner M, Bonine N, Stephenson JJ, DiGenio A, Bieszk N. Are patients on basal insulin attaining glycemic targets? Characteristics and goal achievement of patients with type 2 diabetes mellitus treated with basal insulin and physician-perceived barriers to achieving glycemic targets. Diabetes Res Clin Pract. 2016;121:17–26. [DOI] [PubMed] [Google Scholar]
- 11.Meneghini LF, Mauricio D, Orsi E, Lalic NM, Cali AMG, Westerbacka J, et al. The diabetes unmet need with basal insulin evaluation (DUNE) study in type 2 diabetes: achieving HbA1c targets with basal insulin in a real-world setting. Diabetes Obes Metab. 2019;21(6):1429–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Jasso-Huamán LE, Villena-Pacheco A, Guevara-Linares X. [Metabolic control in diabetic patients attended as outpatients in a general hospital]. Rev Med Hered. 2015;26(3):167–72. [Google Scholar]
- 13.Guevara-Tirado A. Niveles de control glicémico En Pacientes diabéticos a través de La hemoglobina glucada En Un área urbana de Villa El Salvador, Lima, Perú, 2020–2021. Rev Peru De Investig Salud. 2022;6(1):29–32. [Google Scholar]
- 14.Azañedo D, Bendezú-Quispe G, Lazo-Porras M, Cárdenas-Montero D, Beltrán-Ale G, Thomas NJ, et al. [Quality of metabolic control in outpatients with type 2 diabetes seen in a private clinic]. Acta Med Peru. 2017;34(2):106–13. [Google Scholar]
- 15.American Diabetes Association Professional Practice Committee. 9. Pharmacologic approaches to glycemic treatment: standards of care in Diabetes-2025. Diabetes Care. 2025;48(Supplement1):S181–206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cowart K, Overbasalization. Addressing hesitancy in treatment intensification beyond basal insulin. Clin Diabetes. 2020;38(3):304–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zisman A, Morales F, Stewart J, Stuhr A, Vlajnic A, Zhou R. Beam value: an indicator of the need to initiate and intensify prandial therapy in patients with type 2 diabetes mellitus receiving basal insulin. BMJ Open Diabetes Res Care. 2016;4(1):e000171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cowart K, Updike WH, Pathak R. Prevalence of and characteristics associated with overbasalization among patients with type 2 diabetes using basal insulin: a cross-sectional study. Clin Diabetes. 2021;39(2):173–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Reid T, Gao L, Gill J, Stuhr A, Traylor L, Vlajnic A, et al. How much is too much? Outcomes in patients using high-dose insulin glargine. Int J Clin Pract. 2016;70(1):56–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Umpierrez GE, Skolnik N, Dex T, Traylor L, Chao J, Shaefer C. When basal insulin is not enough: a dose-response relationship between insulin glargine 100 units/mL and glycaemic control. Diabetes Obes Metab. 2019;21(6):1305–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Stewart-Lynch A, Meyers R, Sidig D, McConville SO, Heiple L. Quantifying and characterizing the presence of insulin overbasalization in a family medicine practice. Clin Diabetes. 2024;42(2):266–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Seguro Social de Salud. Directiva para el acceso, registro y uso de la información de las prestaciones de salud en el Sistema Informático de Servicio de Salud Inteligente - ESSI del Seguro Social de Salud - EsSalud. Lima, Perú: EsSalud; 2023.
- 23.Consellería de Sanidade. Epidat: programa Para análisis epidemiológico de Datos. Galicia, España: OPS-OMS; 2016. [Google Scholar]
- 24.Goldman JD, Angueira-Serrano E, Gonzalez JS, Pang C, Tait J, Edelman S. Survey reveals patient and health care provider experiences and challenges with the use of high doses of basal insulin. Clin Diabetes. 2023;41(2):244–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.World Health Organization. Obesity: preventing and managing the global epidemic : report of a WHO consultation. Geneva, Switzerland: WHO; 2000. [PubMed] [Google Scholar]
- 26.Cowart K, Vascimini A, Kumar A, Tsalatsanis A, Saba Y, Carris NW. Impact of overbasalization on clinical outcomes in patients with type 2 diabetes: a post hoc analysis of a large randomized controlled trial. Clin Diabetes. 2023;41(2):147–53. [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
Data Availability Statement
No datasets were generated or analysed during the current study.