Table 1

A summary of study cohorts and main results from previous studies on mortality in ASD

StudyCountryTotal n
ASD
Diagnosis,
% of total
n
Mean age at
diagnosis
(range),
years
Mean
(years)
follow-up
[time period]
Observed
deaths
n
Expected
deaths in
general
population, n
Reported
risk ratio
(95% CI)
Separate
analyses for
HFASD/
LFASD
Risk ratio separately
for females/males
(CIs if given in the
original article)
Isager et al16Denmark341ASDa low-
functioning:
41.8%
9.5
(2–17)
23.7
[1960–1993]
126.2SMR 1.9
(1.0–3.4)
Not
analysed
F: 3.61 (0.75–10.56)
M: 1.67 (0.76–3.18)
Mouridsen et al17Update of the cohort reported
in Isager et al16
35.5
[1960–2006]
26: low-
functioning
12/143; high-
functioning
14/196
13.5SMR 1.9
(1.3–2.8)
N
significant
differences
F: 4.01 (1.73–7.90)
M: 1.57 (0.93–2.48)
Shavelle et al19USA13 111ASD low-
functioning:
48.1%
8.5
(2–15)
14
[1983–1997]
20284.2SMR 2.4Not
analysed
F: 5.5, M: 1.7
Pickett et al18Update of the cohort reported in
Shavelle et al19
19
[1998–2002]
280114.2SMR 2.5Not
analysed
F: 5.2, M: 2.3
Gillberg et al15Sweden120ASD low-
functioning:
80%
11 (4–18)22.5
[1962–2008]
91.6SMR 5.6
(2.5–10.5)
Not
analysed
F: 20.7, M:2.3
Bilder et al20USA305ASD low-
functioning:
64%
10.8 (1.4–
29.2)
25
[1982–2011]
29Comparison
with
matched
population
controls
HRR 9.9
(5.7–17.2)
Not
analysed
F: 20.7 (6.2–69.2)
M: 7.9 (4.2–15.0)
  • ASD, autism spectrum disorder; SMR, standardised mortality ratio; HRR, hazard rate ratio; CI, confidence interval; F, female; M, male.

  • a. The previous Scandinavian diagnosis Borderline childhood psychosis was included as a proxy for Asperger syndrome.

  • IQ was tested in 60% of cases and in the rest of the cases was based on clinical assessments.