Table 4.

Physical (PCS) scores and the relationship with multiple comorbid chronic conditions among type 2 diabetes patients

Chronic condition	Each comorbidity independentlya			Multiple analysisb
	B	(95 % CIs)	P value	B	(95 % CIs)	P value
Retinopathy	−3.65	(−5.19 to −2.11)	<0.001	−3.88	(−5.41 to −2.35)	<0.001
Cerebral diseases	−2.88	(−5.49 to −0.29)	0.029	–
Myocardial infarction	−0.26	(−3.30 to 2.78)	0.867	–
Heart diseases	−3.96	(−5.74 to −2.19)	<0.001	−4.25	(−5.97 to −2.52)	<0.001
Cancer/malignancies	0.43	(−1.89 to 2.74)	0.719	–
Migraine/severe headache	−0.93	(−2.48 to 0.62)	0.238	–
High blood pressure	−0.19	(−1.12 to 0.74)	0.695	–
Atherosclerosis abdomen/legs	−4.42	(−5.99 to −2.84)	<0.001	−4.51	(−6.07 to −2.94)	<0.001
Lung diseases	−3.57	(−4.95 to −2.19)	<0.001	−3.66	(−5.03 to −2.28)	<0.001
Intestine disorders	−0.54	(−2.25 to 1.16)	0.536	–
Incontinence	−3.08	(−4.29 to −1.87)	<0.001	−3.32	(−4.52 to −2.12)	<0.001
Back disorder	−4.55	(−5.82 to −3.29)	<0.001	−4.76	(−6.01 to −3.51)	<0.001
Osteoarthritis	−4.71	(−5.75 to −3.66)	<0.001	−4.74	(−5.78 to −3.70)	<0.001
Chronic rheumatoid arthritis	−2.74	(−4.20 to −1.28)	<0.001	−2.96	(−4.38 to −1.54)	<0.001
Neck, shoulder disorder	−2.77	(−4.06 to −1.49)	<0.001	−3.03	(−4.26 to −1.78)	<0.001
Elbow, wrist, hand disorder	−0.49	(−1.91 to 0.92)	0.495	–

^aLinear regression analysis was used to calculate unstandardized coefficients B and 95 % confidence interval (CIs) for each comorbid chronic condition independently

^bA prediction model was calculated using a backwards stepwise regression procedure starting with all comorbid chronic conditions and then eliminating all variables which did not contribute (P > 0.1) to the model