Differential Child Maltreatment Risk Across Deployment Periods of US Army Soldiers

Abstract

Objectives. We described the risk for maltreatment among toddlers of US Army soldiers over different deployment cycles to develop a systematic response within the US Army to provide families appropriate supports.

Methods. We conducted a person-time analysis of substantiated maltreatment reports and medical diagnoses among children of 112 325 deployed US Army soldiers between 2001 and 2007.

Results. Risk of maltreatment was elevated after deployment for children of soldiers deployed once but not for children of soldiers deployed twice. During the 6 months after deployment, children of soldiers deployed once had 4.43 substantiated maltreatment reports and 4.96 medical diagnoses per 10 000 child-months. The highest maltreatment rate among children of soldiers deployed twice occurred during the second deployment for substantiated maltreatment (4.83 episodes per 10 000 child-months) and before the first deployment for medical diagnoses of maltreatment (3.78 episodes per 10 000 child-months).

Conclusions. We confirmed an elevated risk for child maltreatment during deployment but also found a previously unidentified high-risk period during the 6 months following deployment, indicating elevated stress within families of deployed and returning soldiers. These findings can inform efforts by the military to initiate and standardize support and preparation to families during periods of elevated risk.

Since October 2001, US Army units have experienced frequent and prolonged combat-related deployments because of multiple international conflicts. As of 2010, more than 2.1 million American men and women service members were deployed in support of multiple international conflicts. Nearly half of these deployed soldiers are parents.¹ Considerable stress on families occurs during unit deployment, when caregivers are absent, and on reintegration, when soldiers return to family life. Periods of elevated stress can increase the likelihood of negative outcomes for families, such as child maltreatment.^2–4

An emerging body of literature characterizing the effect of deployment on family well-being identifies several challenges for soldiers, spouses, and children. Increased rates of mental health problems and substance abuse by soldiers following deployment have been well documented.^5–7 Risks for soldiers’ spouses, including marital problems, domestic violence, and mental health challenges, also have been identified.^8–12 For children, a focus has been on behavioral and educational well-being. Stress and anxiety, as well as behavioral, school, and sleeping problems, during deployment stages have been identified, with additional attachment issues and challenges arising when a soldier returns home.^13–21

Intrafamilial violence among military families, including spousal abuse and child maltreatment, is an area of increasing concern. The Family Advocacy Program (FAP) was established in 1976 as a preventive effort for spouse and child maltreatment in all military families and communities, separate from civilian child protective services. The Army Central Registry reported an overall decrease in the rate of substantiated child maltreatment from 1990 to 2004, largely because of a decrease in the rate of child physical abuse. However, the rate of child neglect increased during this time. In 2004, child neglect reached the highest level since 1991, increasing by 40% from 2000 to 2004.⁴ One study reported an increased rate of maltreatment during combat-related deployment compared with nondeployed periods in a small sample of US Army soldiers.² Beyond annual rates, few studies have examined the risk for child maltreatment in relation to deployment stages.

A better understanding of the risk of child maltreatment across a single deployment cycle, as well as the effect of the multiple deployments experienced by many soldiers, is needed to develop a systematic response within the US Army to provide families appropriate supports. The purpose of this study was to characterize rates of child maltreatment episodes among children younger than 2 years in US Army families across stages of soldier deployment cycles during a period of intense deployment tempo, characterized by increased frequency and length of deployment in the last decade.

METHODS

Data were obtained from the Army Central Registry, Patient Administration Systems and Biostatistics Activity, and Defense Manpower Data Center. Deployment dates for soldiers and soldier demographic information were obtained from Defense Manpower Data Center; substantiated maltreatment reports from FAP were obtained from the Army Central Registry; and TRICARE medical diagnoses were obtained from Patient Administration Systems and Biostatistics Activity.

We included children of active-duty US Army soldiers with at least 3 consecutive years of active service between 2001 and 2007. Children younger than 2 years are considered the population at highest risk for severe injury and fatality from child maltreatment. In addition, restricting the participants to this age group reduces heterogeneity in age-related child maltreatment episodes across the deployment cycle.

We selected children of soldiers with 1 or 2 total deployments during the study period because the deployment experience for those with 3 or more deployments differed across key characteristics (such as duration) and also accounted for only a small minority of soldier experiences (12% of all those deployed during the period). Furthermore, we separately considered children of soldiers with 1 and 2 total deployments during the study to accurately define time periods relative to deployment(s). Soldiers may have qualitatively different experiences before, during, and after deployment depending on whether they experienced 1 or 2 deployments. Each month of observation for children was linked to their soldier parents’ or caregivers’ deployment. Months when children had 2 or more soldiers listed (1 375 464 child-months), when 2 soldiers were listed as the TRICARE sponsor for the same child at the same time, were excluded from the analysis because of difficulty in assigning episodes to deployment cycles of a single soldier. Additionally, children with more than 1 soldier parent or caregiver during the first 24 months of life were excluded because of difficulties in assigning maltreatment episodes in relation to deployment period. Children were followed up until they reached age 24 months.

Study Measures

The primary outcome of interest was a child maltreatment episode. We considered child maltreatment episodes arising from 2 separate sources: (1) substantiated child maltreatment reports from FAP and (2) medical diagnoses of child maltreatment (based on International Classification of Diseases, Ninth Revision, Clinical Modificatio,²² [ICD-9-CM] codes) from TRICARE medical records. Substantiated FAP reports capture 4 types of child maltreatment: physical, sexual, emotional, and neglect. Because child maltreatment episodes may not be reported to child protective service agencies (or alternatively reported to civilian child welfare agencies without notification to FAP), relying solely on FAP reports will result in under-ascertainment of cases.^23–25 Thus, we also used medical diagnoses of child maltreatment found in TRICARE medical records. Medical diagnoses of child maltreatment included any of the following ICD-9-CM, diagnosis or external-cause-of-injury codes from an outpatient or inpatient encounter: 995.50, 995.51, 995.52, 995.53, 995.54, 995.55, 995.59, and E967.^26–29 We considered these 2 sources of child maltreatment episodes to be 2 distinct outcomes capturing various types of maltreatment, providing different yet valuable information for each episode, and may not overlap in the data.

For each child, all medical encounters occurring within 1 day of each other were collapsed into a single encounter. Then, following previous studies,^30–32 multiple medical encounters with the same diagnosis or related diagnosis occurring within a 180-day interval were grouped into episodes to avoid counting a single abuse episode more than once. For each abuse episode, the date on which the episode began was linked by month to the corresponding soldier’s deployment data.

Exposures of interest were predefined time periods relative to a soldier’s deployment. For soldiers who experienced a total of 1 deployment during the study period (1-deployed), the observation window was divided into the following 5 periods:

1. 7 or more months predeployment (time before a deployment begins),

2. 0 to 6 months predeployment,

3. during deployment,

4. 0 to 6 months postdeployment (time after a deployment ends), and

5. 7 or more months postdeployment.

For soldiers who experienced a total of 2 deployments during the study period (2-deployed), the observation window was divided into the following 7 periods:

1. 7 or more months predeployment,

2. 0 to 6 months pre–first deployment,

3. during first deployment,

4. between deployments (interdeployment),

5. during second deployment,

6. 0 to 6 months post–second deployment, and

7. 7 or more months postdeployments.

These periods are depicted in Figures 1 and 2. It was hypothesized a priori that the 6 months immediately before and after deployment would be higher-risk periods, which is why the predeployment and postdeployment periods were split.

FIGURE 1—

Rate of Child Maltreatment by Deployment Periods Among Children of US Army Soldiers Deployed Once Between 2001 and 2007

Note. FAP = Family Advocacy Program. The vertical axis shows the rate of child maltreatment in episodes per 10 000 child-months of exposure. The horizontal axis shows the deployment periods of the soldier.

FIGURE 2—

Rate of Child Maltreatment by Deployment Periods Among Children of US Army Soldiers Deployed Twice Between 2001 and 2007

Note. FAP = Family Advocacy Program. The vertical axis shows the rate of child maltreatment in episodes per 10 000 child-months of exposure. The horizontal axis shows the deployment periods of the soldier.

Data Analysis

For each deployment period, maltreatment rates were calculated as the observed number of maltreatment episodes divided by the number of child-months. We obtained P values comparing rates in different exposure periods assuming a Poisson distribution. All analyses were performed with Stata version 13.1 (StataCorp LP, College Station, TX).

RESULTS

From 2001 to 2007, 66 724 1-deployed soldiers had 97 013 child dependents 0 to 24 months old, and 44 585 2-deployed soldiers had 66 828 child dependents 0 to 24 months old. Generally, characteristics of 1- and 2-deployed soldiers were similar. However, a smaller percentage of women and officers were deployed twice during our observation period (Table 1). We excluded soldiers with 3 or more deployments during the study period (8.6% of all soldiers in the cohort) because we found that they were different in regard to length and nature of the deployment. Among 1-deployed soldiers, the mean length of deployment was 12 (SD = 4.86) months. Among 2-deployed soldiers, the mean length of deployment was 9 (SD = 4.69) months. The median time between the first and second deployment among 2-deployed soldiers was 12 months.

TABLE 1—

Characteristics of US Army Soldiers Deployed Between 2001 and 2007

Characteristic	1 Total Deployment (n = 66 724), No. (%)	2 Total Deployment(s) (n = 44 585), No. (%)
Gender
Female	6 102 (9.1)	2 218 (5.0)
Male	60 622 (90.9)	42 367 (95.0)
Race/ethnicity
White	38 558 (57.8)	26 066 (58.5)
Black	15 157 (22.7)	9 854 (22.1)
Hispanic	8 006 (12.0)	5 528 (12.4)
American Indian/Alaska Native	1 168 (1.8)	864 (1.9)
Asian/Pacific Islander	1 691 (2.5)	1 025 (2.3)
Other	1 682 (2.5)	968 (2.2)
Unknown/missing	373 (0.6)	270 (0.6)
Education
< high school	4 036 (6.0)	3 157 (7.1)
High school/GED/diploma	44 315 (66.4)	31 534 (70.7)
Some college/associate’s degree	4 287 (6.4)	2 194 (4.9)
College	7 818 (11.7)	4 593 (10.3)
> college	3 268 (4.9)	1 724 (3.9)
Unknown	3 000 (4.5)	1 383 (3.1)
Rank
Enlisted	57 446 (86.1)	39 351 (88.3)
Officer	8 387 (12.6)	4 495 (10.1)
Warrant officer	891 (1.3)	739 (1.7)

Note. GED = general equivalency diploma.

Among 1-deployed soldiers, 444 children (0.46%) had a substantiated maltreatment episode, and 461 children (0.48%) had a medical diagnosis of maltreatment, resulting in rates of 4.58 substantiated maltreatment episodes per 1000 children and 4.75 medical diagnoses of maltreatment per 1000 children across the study period. For 2-deployed soldiers, 334 children (0.50%) had a substantiated maltreatment episode, and 270 children (0.40%) had a medical diagnosis of maltreatment, corresponding to 5.00 substantiated maltreatment episodes per 1000 children and 4.04 medical diagnoses of maltreatment per 1000 children.

Substantiated Child Maltreatment Reports

Child maltreatment rates among dependent children of 1-deployed soldiers.

For children of 1-deployed soldiers, substantiated FAP reports occurred at a rate of 2.90 episodes (95% confidence interval [CI] = 2.67, 3.17) per 10 000 child-months. Substantiated FAP reports were higher for the 6-month period after deployment compared with the 6-month period before deployment (4.43 episodes vs 2.69 episodes per 10 000 child-months; P = .007). The rate during deployment was also higher compared with the 6-month period before deployment (3.72 episodes vs 2.69 episodes per 10 000 child-months; P = .05; Figure 1).

Among substantiated FAP reports filed during soldier deployment, the perpetrator was listed as the nonsoldier caregiver in 88% of cases. In all other periods of nondeployment, the perpetrator was listed as the soldier in, on average, 55% of reports. (Table describing perpetrator type by deployment period available as a supplement to the online version of this article at http://www.ajph.org.)

Child maltreatment rates among dependent children of 2-deployed soldiers.

In aggregate, the rate of child maltreatment for dependents of 2-deployed soldiers was 3.01 episodes (95% CI = 2.71, 3.33) per 10 000 child-months. The rate of maltreatment during a soldier’s first deployment was 2.78 episodes (95% CI = 2.02, 3.75) per 10 000 child-months. During a soldier’s second deployment, the rate of child maltreatment increased significantly to 4.83 episodes per 10 000 child-months (95% CI = 3.89, 5.94; P = .003; Figure 2).

Medical Diagnoses of Child Maltreatment From TRICARE

Child maltreatment rates among dependent children of 1-deployed soldiers.

In aggregate, medical diagnoses of child maltreatment occurred at a rate of 3.22 episodes (95% CI = 2.96, 3.49) per 10 000 child-months. Medical diagnoses of child maltreatment occurred at a rate of 3.34 episodes (95% CI = 2.74, 4.04) per 10 000 child-months during deployment, which was lower than the 6-month postdeployment period rate of 4.96 events (95% CI = 3.90, 6.22; P = .009) per 10 000 child-months (Figure 1). (Table describing maltreatment type by deployment period available as a supplement to the online version of this article at http://www.ajph.org.).

Child maltreatment rates among dependent children of 2-deployed soldiers.

In aggregate, the rate of child maltreatment for dependents of soldiers who were deployed twice was 2.71 episodes (95% CI = 2.44, 3.02) per 10 000 child-months. The rate of maltreatment during a soldier’s first deployment was 2.98 episodes (95% CI = 2.18, 3.97) per 10 000 child-months. During a soldier’s second deployment, the rate of child maltreatment was slightly lower with 2.69 episodes (95% CI = 1.99, 3.54) per 10 000 child-months (Figure 2).

DISCUSSION

We have characterized the risk for substantiated child maltreatment reports and medical diagnoses of maltreatment among the young children of active-duty US Army soldiers who were deployed between 2001 and 2007. We found an elevated risk of maltreatment directly after deployment among children of 1-deployed soldiers. In contrast, among children of 2-deployed soldiers, we found an elevated risk for substantiated maltreatment reports (although not medical diagnoses of maltreatment) during the second deployment. The rate of substantiated maltreatment reports during the second deployment was nearly double that observed during the first deployment.

The finding of increased postdeployment risk for child maltreatment among children of a 1-deployed soldier suggests elevated stress within families when a soldier reintegrates after the first deployment. Many factors within families likely contribute to this increased risk. For example, young children may have behavior and mood changes, including aggression, depression, and hyperactivity, during parent deployment, related to the trauma of separation.^20,33,34 Behavior changes also may increase in frequency relative to the length of deployment.^14,18 At the same time, soldiers who are readjusting following deployment may have particular difficulty dealing with behavior changes in their children.

Currently, the US Army offers programs to families of deployed soldiers designed to facilitate adjustment to all phases of the deployment cycle, including parenting classes, child care services, and free classes specific to soldier reintegration into home life. However, such programs may not be offered at a scale to support need during the immediate postdeployment period. Data such as these might suggest a need for a more intentional strategy to prepare families, particularly those with young infants, for both a soldier departure and a soldier returning home.

Although postdeployment elevation of risk was identified for children of 1-deployed soldiers during our observation period, we did not detect such an elevated risk among soldiers who were deployed twice. Several explanations are possible for this difference. First, our rates in each period were sensitive to sample size, with fewer child maltreatment episodes observed in each period among this smaller subgroup of 2-deployed soldiers. Second, and more important, there may be selection differences among soldiers who deploy twice that are associated with risk of future child maltreatment episodes. For example, a child maltreatment episode or significant soldier mental health event following the first deployment may affect eligibility for a second deployment.

Our study was unique in that it focused on children younger than 2 years, a period of high stress for families, especially those who experience deployment and reintegration. Prior studies focused on the risk related to the period of deployment, a time during which children appeared to be at higher risk for maltreatment and, in particular, neglect.² Our study confirmed an elevated risk for child maltreatment during deployment, particularly among later deployments in 2-deployed families. In addition, our findings showed a previously unidentified high-risk period for children during the 6 months following a soldier’s return home after a single deployment.

Our study had limitations. First, we did not adjust for covariates. For example, we were unable to identify direct combat exposure in the soldiers and therefore could not examine the relation of combat exposure to risk of child maltreatment. In addition, we were unable explore family-specific risk factors because of limitations of the available data, such as income, that may have influenced child maltreatment rates. However, we did not expect a lot of variation among those factors within this relatively homogeneous population during the 24-month follow-up period. Second, although we could identify parent–child relationships, we could not determine the geographic residence of the child, who may have been cared for by a noncustodial, nonmilitary parent or extended family member at various points in the deployment cycle. Care by nonparent caregivers may heighten risk for neglect during deployment periods or may act as a buffer toward direct physical abuse by a soldier returning from conflict. Third, ascertaining child maltreatment episodes is fraught with difficulty, given both underreporting and the heterogeneity of types of maltreatment across the deployment cycle. There is no single reliable measure of child abuse. For that reason, we studied 2 different measures (substantiated maltreatment reports and medical maltreatment diagnosis) to triangulate common trends across different measures. Finally, because reporting practices and trends may vary across posts, our results could have been confounded by base, particularly if bases with higher deployment tempo were not reporting child maltreatment episodes as rigorously as those with lower deployment tempos.

This study identified an elevated risk for child maltreatment during the 6 months following a soldier’s return home from deployment among soldiers who were deployed only once. A somewhat different pattern was observed among the children of soldiers deployed twice, in whom the risk was most elevated during the second deployment. These findings, which illustrate the experiences of deployed US Army families with infants and toddlers, add to the body of literature characterizing elevated stress within families of returning war veterans. Many of the perpetrators identified in our study cohort were not soldiers themselves, suggesting that a soldier-only response is not enough to reduce risk for families. These findings can inform efforts by the military to initiate and standardize support and preparation to families during periods of elevated risk.