. 2014 Oct 24;14(1):112–121. doi: 10.1111/acel.12275

Table 4.

Variables cross-sectionally and longitudinally associated with plasma dihydrosphingomyelins in the Baltimore Longitudinal Study of Aging

	C16:0		C18:0		C22:0		C24:0
	Baseline	Time-Dependent	Baseline	Time-Dependent	Baseline	Time-Dependent	Baseline	Time-Dependent
Covariates	b (SE)	b (SE)	b (SE)	b (SE)	b (SE)	b (SE)	b (SE)	b (SE)
Age	101.3 (15.6)^c		2785.5 (281.8)^c		−3.0 (6.4)		3.2 (20.3)
Age 2	0.5 (0.7)		19.2 (15.6)		−1.6 (0.6)^b		−3.0 (1.1)^b
Men	−2795.9 (230.0)^c		−14 287.7 (4130.0)^b		350.3 (125.1)^b		−376.2 (261.6)	−51.6 (24.2)^a
African American			19 115.4 (5822.6)^b		1007.3 (186.0)^c		3037.0 (378.9)^c
BMI	48.8 (21.0)^b		668.9 (454.1)		−13.4 (14.7)	2.9 (1.5)	−42.3 (30.3)
WHR	2444.9 (1135.5)^a
Prediabetes	−232.4 (169.3)	−42.4 (16.6)^a	−10 907.2 (3611.8)^b	−725.1 (353.4)^a			−492.9 (255.3)
Diabetes	−52.5 (258.9)	−47.3 (24.3)	−8605.3 (5551.7)	−1136.3 (522.0)^a			−273.1 (399.6)
Statin use			8238.3 (4166.0)^a
Cancer	533.7 (237.8)^c		24 814.7 (5535.4)^c	−1517.3 (544.6)^b			213.1 (399.6)	−79.4 (38.3)^a
Smoker	205.1 (148.3)	−28.1 (14.2)^a
Cholesterol	17.6 (2.0)^c		282.0 (42.3)^c		29.8 (1.5)^c		62.4 (3.0)^c
Triglycerides	4.5 (1.1)^c		130.2 (22.5)^c		−2.8 (0.8)^b		−6.3 (1.7)^c	0.3 (0.2)

BMI, body mass index; WHR, waist–hip ratio.

We forced age, age 2, sex, and BMI into all models then used backward selection with P < 0.10 to determine which variables to include in the final model of each dihydrosphingomyelin carbon-chain length. ^aP < 0.05; ^bP < 0.01; ^cP < 0.001.