Skip to main content
. 2017 Jan 25;12(1):e0170569. doi: 10.1371/journal.pone.0170569

Table 4. Summary Results for Overall (4 Year) Regression Models for Psychological Disorders.

Independent Variables Postinjury Outcomes Variables OR (95% CI)
PTSD Mood Adjustment Anxiety* Substance Abuse Other
Injury Group
    Limb salvage (ref)
    Early amputation NS NS NS NS NS NS
    Late amputation NS 2.83 2.36 NS NS 2.66
(1.31–6.09) (1.20–4.62) (1.49–4.77)
Blast (vs. no blast, ref) NS NS NS 0.39 NS 2.29
(0.20–0.76) (1.37–3.86)
ISS (log) NS NS NS NS NS 1.98
(1.46–2.96)
Age, yr 0.93 NS NS 0.94 NS
(0.89–0.98) (0.89–0.99)
Injury year (2001/5, ref vs. 2006/8) NS NS 0.38 0.53 0.47 0.42
(0.26–0.55) (0.36–0.79) (0.27–0.83) (0.31–0.58)
Preinjury psychological diagnosis NS NS NS NS NS NS

Traumatic brain injury and injury location (above vs. below knee) were not significant in the models.

*Variables with significant injury group × time interaction (see results section for description of the interactions).

†Other psychological disorders included postconcussion syndrome, and cognitive and sleep disorders.

Substance abuse: late amputation had significantly increased OR relative to early amputation (OR = 2.55 [1.17–5.54]).

Anxiety: late amputation had significantly increased OR relative to early amputation (OR = 2.39 [1.34–4.25]).

PTSD: early amputation had significantly reduced OR relative to limb salvage (OR = 0.39 [0.15–0.97]) only during year 1, although the injury group × time interaction was not significant.

CI = confidence interval, ISS = Injury Severity Score, NS = not significant, OR = odds ratio, PTSD = posttraumatic stress disorder.