Results for the main effect (full cohort) and effect modification (interaction) from the zero-inflated Poisson model for any morbidity (yes/no, zero-inflation model) and number of morbidities (count model) for adults with spinal muscular atrophy (SMA) compared to matched adults without SMA (w/o SMA).

Table 4: Results for the main effect (full cohort) and effect modification (interaction) from the zero-inflated Poisson model for any morbidity (yes/no, zero-inflation model) and number of morbidities (count model) for adults with spinal muscular atrophy (SMA) compared to matched adults without SMA (w/o SMA).

	Zero-inflated model	Count model
	OR (95% CI)	IRR (95% CI)
Full cohort^*
Females
SMA vs. w/o SMA	0.26 (0.20, 0.33)	2.11 (2.06, 2.16)
Males
SMA vs. w/o SMA	0.16 (0.12, 0.22)	2.15 (2.10, 2.21)
Age interaction
Females
SMA vs. w/o SMA
18-39 years	0.23 (0.16, 0.33)	3.33 (3.07, 3.60)
40-54 years	0.34 (0.23, 0.50)	3.00 (2.82, 3.20)
55-64 years	0.14 (0.06, 0.34)	2.24 (2.10, 2.38)
65-74 years	0.71 (0.46, 1.09)	2.00 (1.91, 2.09)
≥ 75 years	^**	1.75 (1.67, 1.83)
Males
SMA vs. w/o SMA
18-39 years	0.14 (0.08, 0.22)	2.91 (2.61, 3.24)
40-54 years	0.20 (0.13, 0.30)	2.83 (2.66, 3.02)
55-64 years	0.23 (0.13, 0.41)	2.35 (2.23, 2.49)
65-74 years	0.37 (0.22, 0.61)	2.13 (2.04, 2.22)
≥ 75 years	^**	1.72 (1.63, 1.81)

OR: Odds ratio; IRR: Incidence rate ratio; CI: Confidence interval. Age (continuous) was a covariate in the full cohort model; *Estimates are adjusted by age (continuous); **Too few with 0 morbidities to provide reasonable estimates. The zero-inflation model is a logistic regression model, predicting whether the individual had zero vs. non-zero (i.e.¸ ≥ 1) morbidities. An OR < 1.00 indicates a lower likelihood of having zero morbidities. The count model is a Poisson regression model, predicting the number of morbidities among those with non-zero values.