Group-based trajectory analysis identifies varying diabetes-related cost trajectories among type 2 diabetes patients in Texas: an empirical study using commercial insurance

Table 4 Univariable and multivariable cumulative logistic regression analyses of low stable vs. medium stable vs. high stable cost over 4 years. Low stable cost is the reference level. (N = 260,984)

Variable	Odds Ratio	95% Confidence Interval	P-value
Univariable analysis
Age (continuous)	1.040	1.039–1.041	< 0.0001
Sex (female vs. male)	1.049	1.034–1.064	< 0.0001
Baseline A1C (continuous)	1.260	1.253–1.267	< 0.0001
Complication (short vs. none)	2.502	2.457–2.547	< 0.0001
Complication (long vs. none)	3.018	2.953–3.085	< 0.0001
Complication (Both short and long vs. none)	8.021	7.821–8.227	< 0.0001
Metro status (non-metro vs. metro)	0.978	0.958–0.999	0.036
Charlson Comorbidity Index (Continuous)	1.668	1.656–1.680	< 0.0001
Charlson Comorbidity Index (> 2 vs. < = 2)	4.441	4.358–4.525	< 0.0001
Multivariable analysis
Age (continuous)	1.040	1.039–1.041	< 0.0001
Sex (female vs. male)	1.103	1.081–1.126	< 0.0001
Baseline A1C (continuous)	1.178	1.171–1.185	< 0.0001
Complication (short vs. none)	2.461	2.401–2.524	< 0.0001
Complication (long vs. none)	2.084	2.953–3.085	< 0.0001
Complication (Both short and long vs. none)	4.646	4.479–4.819	< 0.0001
Metro status (non-metro vs. metro)	1.021	0.989–1.054	0.198
Charlson Comorbidity Index (Continuous)	1.372	1.358–1.385	< 0.0001

ISSN: 1472-6963