Table 4. Results of the multiple linear regression analysis with several independent baseline variables predicting PCS at five years later.

Variable	B*	SE†	β‡	p-value
Age	−0.158	0.046	−0.096	<0.001
Sex	0.863	0.611	0.042	0.158
Nationality	2.433	1.947	0.033	0.212
Unmarried	2.174	1.327	0.043	0.102
Divorced/Widowed	0.777	0.715	0.030	0.276
Duration of school education (years)
0–8 years	0.611	1.329	0.012	0.646
11–12 years	2.310	1.088	0.056	0.034
>12 years	−0.681	1.372	−0.013	0.620
Smoking status	−0.958	0.403	−0.066	0.018
Insulin (yes)	−0.048	0.732	−0.002	0.948
Diabetes-related complications	−1.335	0.327	−0.115	<0.001
Comorbidities	−0.621	0.356	−0.046	0.081
BMI (kg/m2)	−0.253	0.064	−0.108	<0.001
Self-reported depression	−0.712	0.869	−0.023	0.413
MCS at baseline	0.106	0.029	0.105	<0.001
PCS at baseline	0.501	0.030	0.481	<0.001
HbA1c (%)	−0.103	0.205	−0.014	0.617

Coding of categorical variables is mentioned in the footnote.

*unstandardized regression coefficient; † standard error; ‡ standardized regression coefficient; statistically significant associations are printed in bold

Sex: male and female (referent); Nationality: German nationality (referent), other nationalities. Marital status: “married” (referent), “divorced/widowed” and “unmarried”. Duration of school education: ≤8, 9–10 (referent), 11–12, and > 2 years. Smoking status: variable ranging from 0 to 2, classified into “never”, “former” or “current”. Treatment with insulin: “yes” and “no” (referent). Diabetes-related complications: variable ranging from 0 (no complications) to 4 (four complications). Comorbid diseases: composite variable ranging from 0 to 4. Depression: “no” (referent) and “yes” (depression episode has been diagnosed).