Risk, Resilience, and Behavioral Health in the Ohio National Guard, 2008-2012

I. DEMOGRAPHICS: The Who

2

male female white black other 17 - 24 25 - 34 35 - 44 45+0

10

20

30

40

50

60

70

80

90

100

85

15

88

75

3432

24

10

86

14

89

74

26

36

25

13

84.9

15.1

87.3

7.45.3

33.6 34.8

19.6

11.9

85

15

88

75

3134

21

14

W1W2W3W4

ONG sample demographics (1)

race agegender

*Waves 3 and 4 include the dynamic cohorts

% a

mon

g th

ose

in e

ach

wav

e

0

10

20

30

40

50

60

70

80

90

100

4743

10

86

1

12

52

37

11

83

2

15

46 45

9

82

1

16

48

42

10

80

1

18

W1

W2

W3

W4

ONG sample demographics (2)

rankmarital status

*Waves 3 and 4 include the dynamic cohorts

% a

mon

g th

ose

in e

ach

wav

e

male female white black other 17-24 25-34 35-44 45+0

10

20

30

40

50

60

70

80

90

100

88

12

90.1

72.9

30

37.8

21

11.2

88

12

90

73

24

39

23

13

88

12

89

74

29

37

20

14

88

12

88

84

27

37

21

15

W1

W2

W3

W4

Clinical sub-sample demographics (1)

race agegender

% a

mon

g th

ose

in e

ach

wav

e

*Waves 3 and 4 include the dynamic cohorts

0

10

20

30

40

50

60

70

80

90

100

41

48

11

88.6

2

9.4

35.6

53

11.5

88

2.5

9.5

41

48

10

87

3

10

40

50

10

86

2

12

W1

W2

W3

W4

Clinical sub-sample demographics (2)

rankmarital status

*Waves 3 and 4 include the dynamic cohorts

% a

mon

g th

ose

in e

ach

wav

e

II. DEPLOYMENT CYCLES

7

Number of lifetime deployments per person at baseline

0 1 2 3 4 5 ≥ 6 0

5

10

15

20

25

30

35

4036.0

31.3

17.9

8.2

2.7 2.0 2.0

Number of lifetime deployments

% o

f sol

dier

s

%

%

%

%

% % %

Proportion deployed within past year by wave and location

0

5

10

15

20

25

30

26.4

9.8 10.2

25.7

8.6 8.17.1

18.917.7

1.63.1

6.8

Total Deployed

Area of Conflict

Non-conflict Area

2009 2010 2011 20122008

W1 W2 W3 W4

Area of conflict = Iraq or Afghanistan. Non-conflict area = U.S, Saudi Arabia and Kuwait, other

% o

f sol

dier

s de

ploy

ed

Proportion deployed within past year by wave and location

0

5

10

15

20

25

30

26.4

9.8 10.2

25.7

8.6 8.17.1

18.917.7

1.63.1

6.8

Total Deployed

Area of Conflict

Non-conflict Area

2009 2010 2011 20122008

W1 W2 W3 W4

Area of conflict = Iraq or Afghanistan. Non-conflict area = U.S, Saudi Arabia and Kuwait, other

% o

f sol

dier

s de

ploy

ed

Kuwait and domestic hurricane relief

Afghanistan surge

Year 2 Year 3 Year 40.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

9.8 10.2

25.7

9.2 9.2

24.4

0.6 0.7 0.90.1 0.3 0.4

Total Deployed

1 Deployment

2 Deployments

≥ 3 Deployments

Number of deployments per person within past year through the course of the study

III. RISK AND RESILIENCE

12

13

Three elements of risk

Pre-deployment

Peri-deployment

Post-deployment

14

Pre-deployment risk factors

Pre-deployment

Peri-deployment

Post-deployment

Sexual assault: most lifetime sexual traumas reported at baseline occurred outside of a deployment setting

5% 7%

Women Men

Rape

(n=61)

6% 4%

Sexual assault

(n=114) (n=106)

(n=14)

(n=2228)(n=388)n=2616

Not deployment-related

Deployment-related

Pre-deployment preparation: Most soldiers in the sample agreed that they felt well-prepared

16

Strongly disagree

Strongly agree

Scale from the Deployment Risk and Resilience Inventory: A Collection of Measures for Studying Deployment-Related Experiences of Military Personnel and Veterans

At baseline, soldiers had a comparable lifetime prevalence of civilian-related vs. deployment-related traumatic events

Any Assaultive violence

Injury or shocking

Sudden death of a loved one

Other 0

10

20

30

40

50

60

70

80

90

100

Deployment-related events

% A

mon

g D

eplo

yed

Any Assaultive violence

Injury or shocking

Sudden death of a loved one

Other0

10

20

30

40

50

60

70

80

90

100

% A

mon

g To

tal

Civilian-related events

n deployed at baseline = 1668, n total at baseline = 2616

Assaultive events

58%

Non-assaultive events or no

events42%

n = 1839, respondents in baseline who specified their deployment status and did not have events of ambiguous settings or types

Among both settings, assaultive-type events were more common

19

Peri-deployment risk factors

Pre-deployment

Peri-deployment

Post-deployment

On average, soldiers reported high levels of unit support during their most recent deployment

20Scale from the Deployment Risk and Resilience Inventory: A Collection of Measures for Studying Deployment-Related Experiences of Military Personnel and Veterans

Strongly disagree

Strongly agree

Although most sexual trauma occurred in a civilian setting, there was an appreciable exposure of deployment-related

sexual trauma

21

Women (n = 169) Men (n = 1498)

none

54%none

86%

sexual harassment only

27%

sexual harassment only

13%

both

17%

sexual assault only

0.4%

both

1%

sexual assault only

2%

*among those deployed

Many soldiers experienced extremely stressful combat-related events during deployment

22

Saw dead bodies

Saw severely w

ounded people

Took care of in

jured or dying people

0

10

20

30

40

50

60

70

34.8

41.3

24.7

% a

mon

g th

ose

depl

oyed

n = 1668

23

Post-deployment risk factors

Pre-deployment

Peri-deployment

Post-deployment

Soldiers in general reported feeling supported post-deployment, but not necessarily understood

24Scale from the Deployment Risk and Resilience Inventory: A Collection of Measures for Studying Deployment-Related Experiences of Military Personnel and Veterans

Strongly disagree

Strongly agree

25

Post-deployment unemployment: At each wave, about 18% of soldiers deployed in the past year lost a job, were laid off, or

lost a large part of their income in the past year

010203040506070

19.2 17.3 18

% o

ut o

f tho

se d

eplo

yed

in p

ast y

ear

IV. Psychiatric disorders

27

IV. Psychiatric disorders

-An overview

Year 1 Year 2 Year 3 Year 40.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

50.0

26.1

19.421.3

23.0

14.012.3

10.512.012.3

8.3

11.7 13.1

7.6

4.4 4.1 3.8

Any

Depression

AUD

PTSD

Depression and alcohol use disorders were more prevalent than PTSD across all 4 years

% o

ut o

f tho

se w

ho c

ompl

eted

eac

h w

ave

(n = 2616) (n = 1770) (n = 1395) (n = 1172)

Prevalence of all condition co-morbidity at baseline

42%

41%

17%

74%

19%

7%

85%

12%3%

Past 30 days Past year Lifetime

No disorder One disorder At least two disorders

29

Baseline prevalence of other conditions among those with PTSD within the past year (N=188)

30

Alcohol abuse Generalized anxiety disorder

Alcohol dependence Depressive disorder0

10

20

30

40

50

60

7.45

15.96 17.02

48.94

% w

ith c

ondi

tion

amon

g th

ose

with

PTS

D

38% has no other condition, 41% had one other condition, and 20% had 2 or more.

V. Psychopathology by risk and resilience factors:

How do the “pre-, peri-, and post-” matter?

31

32

PRE-Deployment risk factors for PTSD

Pre-deployment

Peri-deployment

Post-deployment

women* men*

Those with prior lifetime sexual trauma exposure had a higher prevalence of past-year PTSD

33

(n=388) (n=2228)

n=2616; *all comparisons significant at p<0.001

% w

ith p

ast-y

ear P

TSD

0

5

10

15

20

25

30

23.6

18.2

4.76.3

Prior sexual traumaNo prior sexual trama

34

PERI-Deployment risk factors for PTSD

Pre-deployment

Peri-deployment

Post-deployment

82%

18%

Soldiers with traumas during deployment were more likely than those with non-deployment traumas to experience flashbacks and

insomnia

35

flashbacks

72%

28%

insomnia

Deployment events Non-deployment events

87%

13%

flashbacks

insomnia

84%

16%

31%

69%59%

41%

However, those with traumas during deployment were less likely than those with non-deployment-related events to report

intrusive memories and fear/helplessness/horror

36

Deployment events Non-deployment events

fear, helplessness, or horror

66%

34%

intrusive memories

fear, helplessness, or horror

61%

39%

intrusive memories

37

POST-Deployment risk factors for PTSD

Pre-deployment

Peri-deployment

Post-deployment

Soldiers with high post-deployment support were significantly less likely to have past-year deployment-related PTSD

38Among those who had at least one potentially traumatic event during deployment, n = 1266. High post-deployment support = a score of 24 or higher on a scale that sums items ranging from 1 (strongly disagree) to 5 (strongly agree) – see slide 25

High post-deployment support

Low post-deployment support

0

5

10

15

20

25

30

35

40

45

50

4.5

26.3

% w

ith p

ast-y

ear P

TSD

A combination:

Pre-, peri-, and post-deployment factors all matter for PTSD

39

Those with low levels of preparedness, unit support and post-deployment support have the highest prevalence of PTSD

e.g. HLH = high preparedness, low unit support, high post-deployment support.Among those who have been deployed and experienced a traumatic event during their most recent deployment (n=1294)

HLH HHH LHH LLH HHL HLL LHL LLL0.0

5.0

10.0

15.0

20.0

25.0

Perc

ent

40

Combination of the three factors

41

PRE-Deployment risk factors for depression

Pre-deployment

Peri-deployment

Post-deployment

women* men*

Those with lifetime sexual trauma exposure had a higher prevalence of past-year depression

42

(n=388) (n=2228)

n=2616; *all comparisons significant at p<0.001

% w

ith p

ast-y

ear d

epre

ssio

n

0

5

10

15

20

25

30

35

4037.1

19.8

10.612.5

Prior sexual traumaNo prior sexual trama

43

PERI-Deployment risk factors for depression

Pre-deployment

Peri-deployment

Post-deployment

Soldiers who experienced potentially traumatic events during their most recent deployment were more likely to have past-year depression than

those who experienced non-deployment-related events

44

% w

ith p

ast-y

ear d

epre

ssio

n

0.0

5.0

10.0

15.0

20.0

25.0

30.0

12.9

5.3

12.4 13.1

17.1

Includes respondents in baseline who specified their deployment status and did not have events of ambiguous settings or types (n = 1839).

45

POST-Deployment risk factors for depression

Pre-deployment

Peri-deployment

Post-deployment

Soldiers with high post-deployment support were significantly less likely to have past-year Depression

46Among those who have been deployed at baseline (N = 1668). High post-deployment support = a score of 24 or higher on a scale that sums items ranging from 1 (strongly disagree) to 5 (strongly agree) – see slide 25

% w

ith p

ast-y

ear D

epre

ssio

n

High post-

deployment s

upport

Low post-

deployment s

upport0

10

20

30

40

50

10.9

24.0

A combination of pre- and peri- deployment risk:

What is it about trauma that matters?

47

0

5

10

15

20

25

30

35

40

19.9

13

17.7

6.1

13.3

7.1

AssaultiveNon-assaultive

Although soldiers with deployment-related events in general were more likely to have depression, the type of the event seems to matter more than the setting

Includes respondents in baseline who specified their deployment status and did not have events of ambiguous settings or types

% w

ith p

ast-y

ear d

epre

ssio

n

V. A look over time:

Knowledge requires a longitudinal approach

Individuals show different trajectories of stress response after traumatic events

Norris FH. Looking for resilience: Understanding the longitudinal trajectories of responses to stress . Social Science & Medicine 2009; 68(12):2190-2198.

Most ONG soldiers show resistance to PTSD symptoms across four years

51

7 20 32 4517

27

37

47

57

67

77

Resistance: 50%Mild, recovery: 36%Mild, stable: 11%Chronic dysfunction: 3%

PTSD

sym

ptom

sca

le

Months after event

*n=469, participants who completed at least two waves of the study, had a traumatic event during a deployment within 2 years of baseline assessment, and were bothered by the same event throughout the follow-up interviews.

Risk factors at baseline affect what group each respondent belongs in

52

7 20 32 4517

27

37

47

57

67

77

Resistance: 50%Mild, recovery: 36%Mild, stable: 11%Chronic dysfunction: 3%

PTSD

sym

ptom

sca

le

*n=469, participants who completed at least two waves of the study, had a traumatic event during a deployment within 2 years of baseline assessment, and were bothered by the same event throughout the follow-up interviews.

Low income Low education High number of lifetime

traumatic events

Months after event

The dotted lines here show how the trajectory shapes would differ if everyone had past-year alcohol misuse

at all waves

53

7 20 32 4517

22

27

32

37

42

47

52

57

62

67

PTSD

sym

ptom

sca

le

Months after event

Alcohol effect

Solid line = no alcohol misuse, depression, traumas, or stressors. Dotted line: alcohol use at every point but no depression, traumas, or stressors.

7 20 32 4510

20

30

40

50

60

70

80

The dotted lines here show how the trajectory shapes would differ if everyone had past-year depression at

all waves

54

PTSD

sym

ptom

sca

le

Months after event

Depression effect

Solid line = no alcohol misuse, depression, traumas, or stressors. Dotted line: depression at every point but no alcohol misuse, traumas, or stressors.

55

7 20 32 450

10

20

30

40

50

60

70

80

The dotted lines here show how the trajectory shapes would differ if everyone had at least one new trauma

since wave one

PTSD

sym

ptom

sca

le

Months after eventSolid line = no alcohol misuse, depression, traumas, or stressors. Dotted line: at least one new trauma at each wave but no alcohol misuse, traumas, or stressors.

Cumulative trauma effect

Most ONG soldiers show resistance to depression symptoms across four years

56

Num

ber o

f dep

ress

ion

sym

ptom

s

Months after event

*n=727, participants who completed at least two waves of the study and were deployed within 2 years of baseline assessment

7 20 32 450

1

2

3

4

5

6

7

8

9

Resistance: 52%

Mild, recovery: 17%

Mild, increasing: 20%

Chronic dysfunction: 11%

Risk factors at baseline affect what group each respondent belongs in

57

Num

ber o

f dep

ress

ion

sym

ptom

s

Months after event

*n=727, participants who completed at least two waves of the study and were deployed within 2 years of baseline assessment

7 20 32 450

1

2

3

4

5

6

7

8

9

Resistance: 52%

Mild, recovery: 17%

Mild, increasing: 20%

Chronic dysfunction: 11%

Childhood adversity High number of lifetime

stressors High number of lifetime

traumatic events

Not currently married

7 20 32 450

1

2

3

4

5

6

7

58

The dotted lines here show how the trajectory shapes would differ if everyone had past-year alcohol misuse

at all waves

Num

ber o

f dep

ress

ion

sym

ptom

s

Months after event

Solid line = no alcohol misuse or PTSD. Dotted line: alcohol misuse at each wave but no PTSD

Alcohol effect

59

The dotted lines here show how the trajectory shapes would differ if everyone had past-year PTSD at all

waves

Num

ber o

f dep

ress

ion

sym

ptom

s

Months after event

Solid line = no alcohol misuse or PTSD. Dotted line: PTSD at each wave but no alcohol misuse

PTSD effect

7 20 32 450

1

2

3

4

5

6

7

8

9

VI. ALCOHOL MISUSE: Some findings

60

What is the relationship between mental health disorders and peri- / post-deployment alcohol

abuse?

Both civilian and deployment-related factors are associated with incident alcohol abuse

62

Yes No Yes No Never Before During/After

Never Before During/After

0

5

10

15

20

25

30

35

40

% w

ho d

evel

oped

alc

ohol

abu

se s

ympt

oms

duri

ng o

r aft

er

depl

oym

ent

Family history of drug/alcohol

abuse

Experienced combat on most recent conflict

Depressive disorder

Post-traumatic stress disorder

Those with both PTSD and depression have the highest probability of developing alcohol abuse

63

No PTSD or DD PTSD only DD Only Both PTSD and DD0.00

0.10

0.20

0.30

0.40

0.50

7.6%

18.4%

24.3%

46.7%

Cond

ition

al P

roba

bilit

y of

an

Alco

hol A

buse

Dis-

orde

r Firs

t Occ

urin

g D

urin

g/Fo

llow

ing

Dep

loy-

men

t

DD – any depressive disorder.These are the predicted probabilities from multivariable logistic regressions adjusted for other variables (age, gender, race, education, income, deployment location, marital status, family history of substance abuse) entered in the multivariate regression model. Note: depression and PTSD are both defined as first occurring during or following deployment.

VII. SUICIDAL IDEATION: Some findings

64

About 2% of the soldiers reported suicidal ideation in the past month, and 11% in their lifetime

Lifetime suicidal Ideation Suicidal Ideation within past 30 days

0

5

10

15

20

25

30

% fr

om b

asel

ine

sam

ple

Calculated within baseline sample (N≈2616). This is “thought (they) would be better off dead, or had thoughts of hurting (themselves)” from the Patient Health Questionnaire - 9. Kroenke K, Spitzer R. The PHQ-9: A new depression diagnostic and severity measure. Psychiatric Annals 2002;32:1. Compared to general population with a prevalence of lifetime suicidal ideation between 4.8% and 18% (cdc.gov/ncipc/wisqars)

10.7

1.9

65

66

In the past 6 months, 7% of the clinical sub-sample have thought about suicide

-1.0

3.0

7.0

11.0

15.0

6.8

2.8

7.0

1.64.4

% fr

om w

ave

1

Wave 1 in-person sample N=500

Deployment experience at baseline was not associated with current suicidal ideation

Never deployed Have been deployed

Not deployed in past 3 years

Deployed in past 3 years

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

% w

ith su

icid

al id

eatio

n th

e pa

st 3

0 da

ys

1.8 2.1

67

1.9 1.9

Telephone sample

68

However, those with deployment history had fewer suicidal outcomes

0.0

2.0

4.0

6.0

8.0

% fr

om w

ave

1 cl

inic

al s

ampl

e Deployed

Never deployed

Wave 1 in-person sample N=500

69

Women were more likely than men to make a suicide plan in the past 6 months

0.0

4.0

8.0

12.0

16.0

% fr

om w

ave

1 MaleFemale

Wave 1 in-person sample N=500

70

In the past 6 months, those with a history of mood disorder had the highest level of suicide risk

Any Anxiety Disorder Any Mood Disorder Any Substance Use Disorder

Any0.0

5.0

10.0

15.0

20.0

25.0

30.0

11.6

26.2

2.9

11.2

% o

f tho

se w

ith su

icid

e ris

k

Wave 1 in-person sample N=500

71

Soldiers with more co-morbidity throughout their lifetime show a higher level of suicide risk

None One Two Three 0.0

5.0

10.0

15.0

20.0

25.0

30.0

1.3

6.4

17.1

27.3

% o

f tho

se w

ith su

icid

e ris

k

Wave 1 in-person sample N=500

Number of disorders

How do deployment, lifetime stressors and mental health affect suicidal ideation?

Among those deployed, current suicidal ideation are associated pre-, peri-, and post- deployment factors

No Yes No Yes No Yes No Yes No Yes0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

% o

f sui

cida

l ide

ation

with

in w

ave

1 su

bgro

ups

with

in th

e pa

st 3

0 da

ys

High preparedness High unit support High family concerns

High post-deployment

support

High combat

All have p-value <0.05. 73

No Yes No Yes No Yes No Yes No Yes No Yes0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

% w

ithin

pas

t 30

days

Stressors experienced over lifespan are associated with current suicidal ideation

Currentincome < 60,000

All have p-value <0.05. n = 2,508. Calculated with in baseline sample that answered all questions.

Low social support

Parents abused each

other

Caregivers fought

Ever unemployed > 3 months

Serious financial problems

74

No Yes No Yes No Yes No Yes0

1

2

3

4

5

6

7

8

9

% w

ithin

the

past

30

days

Among those never deployed, current suicidal ideation was associated with a lifetime occurrence of mental health conditions

Posttraumatic stress disorder Depression Anxiety disorder Alcohol

dependence

All p-value <0.05. Diagnoses by DSM-IV using: PTSD using the PTSD Checklist, Depression using the Patient Health Questionnaire, Anxiety Disorder using the Generalized Anxiety Disorder questionnaire and alcohol dependence using the MINI. For alcohol dependence the N=831 compared to N=907 for the other analyses.

75

How does alcohol dependence relate to suicidal ideation?

Baseline alcohol dependence is significantly associated with incident suicidal ideation at follow-up

6%9%

Baseline Alcohol dependence Follow-up incident suicidal ideation

2%

p=0.0002

n=1587; percent alcohol dependence and suicidal ideation reported for participants meeting criteria

How does co-morbidity relate to suicidal ideation?

Soldiers with past-year PTSD are 5 times more likely than those without to have a history of suicidal ideation

79

33%

67%

8%

92%

No PTSD Current PTSDHistory of

suicidal ideation

Calabrese JR et al. PTSD Comorbidity and Suicidal Ideation Associated with PTSD Within the Ohio Army National Guard. J Clin Psychiatry 2011; 72: 8.

Compared to PTSD alone, those with 2 or more conditions within the past year are 7 times more likely to have a history of suicidal

ideation

80Calabrese JR et al. PTSD Comorbidity and Suicidal Ideation Associated with PTSD Within the Ohio Army National Guard. J Clin Psychiatry 2011; 72: 8.

18%

82%

62%38%

History of suicidal ideation

Current PTSD only Current PTSD and 2+ conditions

VIII. Mental health service utilization

Among soldiers with mental health care need, less than half reported service use

83%

17%

37%

63%

(N=218)(N=1319)

yes

no

yes

no

*mental health need was defined as ≥ 1 mental health disorder, including PTSD, depression, GAD, or an alcohol use disorder (AUD; including abuse or dependence), or past month suicidal ideation

mental health service need at wave 2

mental health service use among those with need

between waves 2 & 3

37%

63%

yes

no

Among those who reported service use, approximately 60% used VA or military care

58%42%

any use of military mental health services between waves 2 & 3

(N=81)(N=218)

TRICARE / VAOther

*mental health need was defined as ≥ 1 mental health disorder, including PTSD, depression, GAD, or an alcohol use disorder (AUD; including abuse or dependence), or past month suicidal ideation

Utilization of mental health differs by psychiatric disorder

0

10

20

30

40

50

60

70

80

90

100

61

43

57

23

47

mental health condition*

*n=218

GAD

(n=68)

PTSD

(n=41)

past month suicidal ideation

(n=19)

depression

(n=128)

AUD

(n=86)

% w

ith a

ny m

enta

l hea

lth c

ondi

tion

repo

rting

any

men

tal

heal

th s

ervi

ce u

se in

pas

t yea

r

Greater psychiatric burden was associated with higher utilization of mental health services

0

10

20

30

40

50

60

70

80

90

100

30

5055

two (n=42)

one(n=145)

three or more(n=29)

Number of mental health diagnoses*

*n=218; p<0.01; mental health diagnoses included PTSD, depression, GAD, or an alcohol use disorder (AUD; including abuse or dependence)

% w

ith a

ny m

enta

l hea

lth c

ondi

tion

repo

rting

any

men

tal

heal

th s

ervi

ce u

se in

pas

t yea

r

Service use conclusions

the majority of ONG soldiers who have mental health care need do not report accessing services

greater psychiatric burden was associated with higher utilization of mental health services

alcohol use disorder by far most common reported psychopathology

those with AUD also least likely to use mental health services a little over half of those who access care report utilizing VA or

DoD services the youngest group of veterans were least likely to use VA/DoD

services