Table 2.

Mixed-effects linear regression model assessing HbA1c variation over time according to SES quartiles.

Variable	Coefficient ( β )	95% confidence interval	p value
Time (post vs. pre)	−1.04	–1.34 to –0.74	< 0.001
SES Quartile 2	−0.82	–1.52 to –0.13	0.020
SES Quartile 3	−0.59	–1.29 to 0.12	0.102
SES Quartile 4	−1.02	–1.73 to –0.30	0.005
Interaction: time × Quartile 2	0.30	–0.13 to 0.72	0.169
Interaction: time × Quartile 3	0.20	–0.23 to 0.62	0.361
Interaction: time × Quartile 4	0.47	0.03 to 0.91	0.037
Age	−0.012	–0.023 to –0.001	0.034
Sex (female)	0.19	–0.04 to 0.42	0.104
Smoking	0.42	0.12 to 0.73	0.007
BMI	0.007	–0.019 to 0.032	0.607
Other antidiabetics agents	0.061	–0.049 to 0.170	0.276
Basal insulin (kg)	0.39	0.03 to 0.75	0.033

Note: Results of the multivariable mixed-effects linear regression model evaluating changes in HbA1c before and after isCGM initiation, incorporating SES quartiles, time point (pre/post), and their interaction as fixed effects. The model is also adjusted for age, sex, smoking status, BMI, number of oral antidiabetic agents, and baseline insulin dose per kilogram. A random intercept at the patient level was included to account for repeated measures.

Abbreviations: BMI, body mass index; HbA1c; glycated hemoglobin; SES, socioeconomic status.