Skip to main content

Table 3 Model parameters

From: Natural history of lung function in spinal muscular atrophy

  

Fixed Effects

Random effects

 

n

Intercept (SE)

95% CI Intercept

Slope

95% CI

Slope

Std. dev. Intercept

Std. dev. Slope

FEV1

 SMA type 1c

6

42.13 (4.58)

33.89; 50.59

−0.40

−1.42; 0.60 (n.s.)

6.11

1.12

 SMA type 2a

46

61.71 (4.51)

52.44; 70.60

−1.29

− 1.78; − 0.81

24.35

1.21

 SMA type 2b

34

81.37 (6.15)

68.36; 93.53

−1.37

−2.04; − 0.73

25.06

1.18

 SMA type 3a

41

97.61 (6.30)

84.80; 110.02

−0.73

−1.11; − 0.35

23.20

0.64

 SMA type 3b

32

100.35 (9.17)

81.91; 118.90

−0.11*

− 0.55; 0.32 *(n.s.)

16.62

n/a*

FVC

 SMA type 1c

5

49.71 (7.34)

34.65; 68.07

−1.15

−3.29; 0.70 (n.s.)

12.60

1.70

 SMA type 2a

47

64.20 (5.29)

53.65; 74.64

−1.32

−1.90; −0.76

28.46

1.39

 SMA type 2b

34

84.53 (6.07)

71.85; 96.50

−1.40

−2.10; −0.71

23.68

1.28

 SMA type 3a

43

96.65 (6.17)

84.08; 108.73

−0.67

−1.06; − 0.31

23.14

0.63

 SMA type 3b

34

109.00 (7.42)

94.46; 123.50

−0.23*

−0.58; 0.11* (n.s.)

15.35

n/a*

VC

 SMA type 1c

6

44.09 (7.06)

28.41; 60.01

−0.78

−2.35; 0.63 (n.s.)

12.84

1.21

 SMA type 2a

32

61.01 (4.62)

51.78; 70.16

−1.57

−2.23; −0.94

23.17

1.40

 SMA type 2b

22

85.54 (6.98)

69.33; 98.18

−1.65

−2.59; −0.60

25.04

1.46

 SMA type 3a

16

96.34 (9.05)

78.17; 114.60

−1.06

−1.71; −0.45

30.07

0.98

 SMA type 3b

4

80.99 (19.67)

35.90; 124.82

0.21

−0.77; 1.23 (n.s.)

35.12

0.75

  1. Legend: Model parameter estimates, standard errors, and confidence intervals for the linear mixed-effects models are shown. n: number of patients in each group
  2. SE standard error, CI confidence interval, Std. dev standard deviation, n.s. slope parameter is not significant; n/a not available
  3. * due to a too limited number of repeated-measurements ‘age at measurement’ was omitted as a random factor from the mixed-effects model. The slope parameter (i.e. the annual rate of decline in % of predicted) will therefore likely be an overestimation of the true value