β | 95% CI | P value | ||
Age* | ||||
Calendar year | −0.107 | −1.429 to 1.215 | 0.874 | <0.001† |
Calendar year2 | −0.111 | −0.248 to 0.027 | 0.114 | |
Age <60 years | −8.066 | −10.411 to −5.723 | <0.001 | |
Age 60–69 years | −4.115 | −6.369 to −1.861 | <0.001 | |
Age 70–79 years | −1.168 | −3.407 to 1.072 | 0.307 | |
Age ≥80 years | Reference group | |||
Interaction year* age | None are significant | |||
Frailty‡ | ||||
Calendar year | 0.247 | −1.100 to 1.593 | 0.719 | <0.001† |
Calendar year2 | −0.159 | −0.299 to 0.018 | 0.027 | |
Frailty 0–0.03 | −0.060 | −1.734 to 1.614 | 0.944 | |
Frailty 0.06–0.09 | 0.127 | −1.519 to 1.772 | 0.880 | |
Frailty 0.11–0.40 | Reference group | |||
Interaction year* frailty | None are significant |
The intraclass correlation coefficient calculated from the empty model was 0.032.
*The age model was adjusted for sex, duration of diabetes, number of chronic medication at initiation, number and/or type of glucose lowering therapy, lipid lowering therapy, presence of albuminuria, presence of dyslipidaemia, haemoglobin A1C (HbA1c), history of cardiovascular events, estimated glomerular filtration rate and body mass index.
†Joined significance of calendar year and calendar year2 using the Wald test.
‡The frailty model was adjusted for sex, duration of diabetes, number and/or type of glucose lowering therapy, lipid lowering therapy and HbA1c.