A total of 460 subjects with T2DM who met the inclusion/exclusion criteria were included in the BASAL-BALI study, and 434 completed the 6-month follow-up. No violation of inclusion criteria or withdrawal of consent was reported, so all 460 subjects who entered the study were included in the analysis. The CONSORT diagram (Fig. 1) presents participant flow in the study.
Baseline and Disease Characteristics in the Study Cohort
Most subjects were aged 55–65 years (mean age was 59.8 ± 10.2 years). Male subjects included in the study were significantly younger (58.26 ± 10.3 years), than female subjects (61.01 ± 10.0 years, p < 0.01). Abdominal obesity was present in 78.3% of subjects, based on ethnicity-specific waist circumference of at least 80 cm in women and at least 94 cm in men, as per 2006 IDF guidelines . At enrollment, the mean duration of diabetes was 11.6 ± 6.6 years, with a median of 10.3 years.
Information on previous insulin therapy was collected for all study participants. Short-acting insulins had been used by 136 subjects (29.6%), intermediate- and long-acting insulins by 298 (64.8%), and premixes by 163 (35.4%). As short- and intermediate- or long-acting insulins may be combined in a single regimen to improve glycemic control, the total number of subjects previously treated with the three insulin types (597) exceeded the number of individuals enrolled in the study (460). Overall, mean duration of previous insulin therapy was 2.7 ± 3.3 years. Considering different insulin types, the average duration of therapy was 1.1 ± 2.5, 2.7 ± 3.3, and 1.5 ± 3.1 years for short-acting, long-acting insulins, and premix insulins, respectively. Daily doses of prior insulin therapy are available in the Supplementary Material, as are doses of basal insulin analogues administered while on study.
At baseline, mean systolic/diastolic blood pressure of study participants was 136.8 ± 15.8 mmHg/83.0 ± 9.1 mmHg. Chronic complications of diabetes mellitus were prevalent in this study cohort, with microvascular complications such as neuropathy (57.6%) and retinopathy (38.9%) being the most common. In addition, the majority of participants had significant comorbidities, including hypertension (72.1%) and dyslipidemia (72.6%).
Overall, 347 (75.4%) subjects received concomitant OADs, with biguanides being the most common (in 71.3%, 66.5%, and 65.6% of subjects at baseline, 3-month, and 6-month visits, respectively), followed by sulfonylureas (9.1%, 7.8%, and 7.6% of subjects), and other OADs (2.2%, 2.4%, and 2.4% of subjects). The type of OAD treatment did not significantly affect the incidence of hypoglycemic events (p = 0.23). Table 1 provides detailed characteristics of the study cohort at inclusion.
Change in HbA1c Level During the Study Period (Primary Objective)
HbA1c levels decreased significantly following basal insulin analogue initiation (Fig. 2), which was already evident at the 3-month visit (7.7 ± 0.7%, p < 0.01) compared with baseline (8.9 ± 1.3%), with a further reduction observed at 6 months (7.1 ± 0.7%, p < 0.01 vs. 3 months). Thus, the average reduction in HbA1c observed during the 6-month treatment period was 1.8% (p < 0.01 vs. baseline).
There was a significant reduction in FPG levels throughout the study duration (Fig. 3), with mean FPG levels reduced from 9.8 ± 2.6 mmol/L at the time of basal insulin analogue initiation to 7.5 ± 1.7 mmol/L at the 3-month visit (p < 0.01) and further to 6.7 ± 1.4 mmol/L at the 6-month visit (p < 0.01 vs. 3-month visit). At baseline, 25 subjects (5.4%) were already at the target FPG level. This number increased to 144 subjects (31.3%) at 3 months, and to 245 subjects (53.3%) at 6 months, the latter change being statistically significant (p < 0.01 vs. baseline and vs. the previous visit).
Body weight, waist circumference, and BMI were all significantly lower at the end of the study compared with the start of basal analogue treatment. Body weight was reduced from a mean of 82.4 ± 14.3 kg at baseline to 81.5 ± 13.5 kg at 6 months (p < 0.01). Over the same period, average waist circumference decreased from 94.9 ± 14.0 to 93.4 ± 13.5 cm (p < 0.01) and mean BMI decreased from 28.6 ± 4.7 to 28.4 ± 4.8 kg/m2 (p < 0.01).
The mean individualized HbA1c target set for each subject by the physicians at study entry was 7.0 ± 0.4%. Most subjects (50.5%) had HbA1c targets of 7.0% to below 7.5%, 31.3% had targets of 6.5% to below 7.0%, and 13.4% had targets of 7.5% to below 8.0%. At the 3-month visit, 53 (11.5%) subjects reached their individualized HbA1c target, and this number increased significantly to 226 (49.1%, p < 0.01) at the 6-month visit. Although the individualized treatment targets were generally close to 7%, subjects who achieved a reduction in HbA1c of at least 0.5% were considered as having a satisfactory response to treatment and there was a significant increase in the proportion of these subjects, from 61.3% at 3 months to 81.3% at 6 months (p < 0.01). Furthermore, a substantial proportion of subjects achieved decreases in HbA1c of more than 1.0%—again, significantly more subjects did so at 6 months (67.2%) than at 3 months (40.4%, p < 0.01).
When both HbA1c and FPG (mean target of 6.5 ± 0.7 mmol/L with a median of 6.5 mmol/L) were considered as a composite target, the proportion of subjects achieving individualized treatment goals also increased significantly from the 3-month (8.3%) to the 6-month visit (42.0%, p < 0.01). Another composite measure of interest was the proportion of individuals achieving target HbA1c values below 7% without experiencing hypoglycemia or weight gain (Fig. 4). Among individuals with available data, this increased significantly from the 3-month (n = 36, 8.4%) to the 6-month visit (n = 140, 32.9%, p < 0.001).
Factors Affecting Subjects’ Ability to Reach HbA1c Targets
Unsurprisingly, adherence to treatment played a substantial part in achieving HbA1c targets. Among subjects reaching HbA1c below 7.0%, significantly more followed the physician’s instructions on insulin titration (n = 89, 19.3%) compared with subjects who did not titrate insulin as instructed (n = 77, 16.7%, p < 0.05).
Subjects with late complications of diabetes (n = 310, 67.4%) appeared less likely to reach HbA1c levels below 7% than those without such complications—only 114 (39.7%) of subjects with late complications achieved HbA1c below 7% compared with 75 (54.3%) of those without such complications (p < 0.01). However, when individualized HbA1c targets were evaluated, there was no significant difference in the proportion of subjects reaching their targets according to the presence of late complications (57.9% without late complications of diabetes vs. 50.8% with late complications, p = 0.17).
The presence of comorbidities, affecting 392 (85.2%) subjects, had a similar adverse effect on subjects’ ability to achieve lower HbA1c levels—the proportion of subjects with no comorbidities reaching HbA1c below 7.0% (n = 42, 66.7%) was significantly higher than the proportion of those with comorbidities who reached this target (n = 147, 40.6%, p < 0.01). The presence of comorbidities had a similar negative impact on reaching individualized HbA1c targets at the end of the follow-up. At the 6-month visit significantly more subjects without comorbidities reached their individual HbA1c target (68.2%) compared with subjects who had comorbidities (50.5%, p < 0.01).
Safety of Basal Insulin Analogues
During the 6 months prior to initiating basal insulin analogues, 443 subjects (96.3%) reported hypoglycemic events (Table 2). After 6 months of basal insulin analogue treatment, the incidence of symptomatic hypoglycemia was substantially lower, affecting 71 subjects (15.4%; p < 0.001 vs. 6 months before insulin analogue treatment). Similarly, the incidence of serious hypoglycemic episodes and nocturnal hypoglycemia was lower during the 6-month study period than over the same time period preceding basal insulin analogue treatment (p < 0.001 vs. 6 months before insulin analogue treatment).