Military Service Absences and Family Members’ Mental Health: A Timeline Followback Assessment

Abstract

Though military service, and particularly absence due to deployment, has been linked to risk for depression and anxiety among some spouses and children of active duty service members, there is limited research to explain the heterogeneity in family members’ reactions to military service stressors. The current investigation introduces the Timeline Followback Military Family Interview (TFMFI) as a clinically useful strategy to collect detailed time-linked information about the service member’s absences. Two dimensions of parent absence—the extent to which absences coincide with important family events and cumulative time absent—were tested as potential risks to family members’ mental health. Data from 70 mother-adolescent pairs revealed that the number of important family events missed by the service member was linked to elevated youth symptoms of depression, even when accounting for the number of deployments and cumulative duration of the service member’s absence. However, youth who reported more frequent contact with the service member during absences were buffered from the effects of extensive absence. Mothers’ symptoms were associated with the cumulative duration of the service members’ time away, but not with family events missed by the service member. These results identify circumstances that increase the risk for mental health symptoms associated with military family life. The TFMFI provides an interview-based strategy for clinicians wishing to understand military family members’ lived experience during periods of service member absence.

Introduction

One parent’s absence from family life is known to be potentially stressful for at-home family members and may be related to their symptoms of psychosocial distress. However, the type of absence (e.g., military service, imprisonment, migration, or divorce), can have different meanings for family members, which thus influence the way family members think about and react to the absence (Rodriguez & Margolin, 2015). Several salient features of absence due to military service include the potential for harm to the service member, honorable motivations to protect and serve the family and the nation, and somewhat unpredictable departures and returns. Even among military families, however, there is considerable variability in anticipated dangers, the amount of contact during absence, and the extent to which service members may discuss their activities. In contrast to previous literature linking family members’ psychological problems to objective indices of military service (e.g., number and length of deployments; Lester et al., 2010), this study represents an early attempt to consider family members’ own experiences during the service member’s absence. We introduce a military service timeline interview used to reconstruct, over the prior 5 years, the total months of absence and the co-occurrence of the service member’s absence with important family events. We test hypotheses that (a) service members’ cumulative time away and missed family events are associated with symptoms of anxiety and depression for mothers and adolescent youth and (b) amount of contact between the service member and the family moderates the impact of these absence dimensions.

Service Member Absence and Family Members’ Symptoms

The service member’s absence, as it relates to family members’ functioning, has generally been quantified in two ways—number of deployments and cumulative time that the service member has been deployed. Number of deployments and number of months of combat deployment have been associated with elevated psychosocial symptoms and mental health diagnoses for youth (Barker & Berry, 2009; Lester et al., 2010; Mansfield, Kaufman, Engel, & Gaynes, 2011) and military spouses (Lester et al., 2010; Mansfield, et al., 2010). Several studies, however, fail to find associations between these absence variables and family members’ adjustment (Chandra et al., 2010; Flake et al., 2009; Morris & Age, 2009). The mixed nature of these findings hints at variability in the nature of deployment absences as well as families’ subjective experiences of those absences.

Notably, the mere frequency of departures and the passage of time do not speak to the fabric of families’ lives during absence, nor do these dimensions describe the impact of the service member’s absence from that shared life. For families who experience major life stressors or transitions (e.g., death of a relative, job loss, residential moves) during absences, the service member’s instrumental and emotional support will likely be deeply missed as family members attempt to cope. On the other hand, the service member’s absence from celebratory family events and life transitions (e.g., anniversaries, children graduating high school or college) may evoke feelings of loss. Qualitative data have described youths’ sadness about service member parents missing events or achievements (Mmari, Roche, Sudhinaraset, & Blum, 2009), but the extent to which the service member’s absence from these events is associated with family member symptoms is unknown. More generally, missed events might also evoke complicated emotional reactions, including anger at the service member’s unavailability and perhaps guilt at enjoying happy events during the service member’s absence. Additionally, for military families, as well as other families who endure temporary parent absence, each absence calls for a degree of reorganization of the rhythms and responsibilities of family life (see Rodriguez & Margolin, 2015). To the extent that major stressful life events or rites of passage occur within a family during an absence, these demands for reorganization may be amplified.

A related consideration is whether deployment absence influences family members’ psychosocial symptoms uniquely, relative to general military service absence. Though recent changes in operational tempo (i.e., high rates of re-deployment, short “dwell times” between deployments) understandably motivates assessment of the impact of deployment on spouses and youth, research has overlooked other common causes of military parent absence due, for example, to: trainings and schools for advancement or skill acquisition, pre-deployment workups, and temporary duty assignments (e.g., conducting equipment inspections at another base). These absences may also intersect with important family experiences, leading family members to accumulate feelings of stress or loss. In assessing the impact of dimensions of absence – the number of absences, cumulative duration of absence, and the number of co-occurring significant family events – we also include the many types of absence that characterize military careers by counting months of cumulative absence due to the varied demands of military service.

The Role of Frequency of Contact with Service Member

Modern technologies may facilitate communication between the service member and family members during service-related absences; however, the impacts of this contact (i.e., frequency, medium, quality, content, etc.) are only beginning to be understood. To the extent that the service member is able to provide support for family members or share in their lives from a distance, the literature on temporary parent absence suggests this contact would be associated with positive outcomes for at-home family members (see Rodriguez & Margolin, 2015). Alternatively, periodic contact might remind at-home family members how much they miss the service member and might disrupt new patterns and routines the family has established. Interestingly, preliminary investigations of contact in military families have linked more frequent youth-service member contact with elevated youth psychosocial stress and more frequent spouse-service member email contact with elevated spouse subjective distress (Houston, Pfefferbaum, Sherman, Melson, & Brand, 2013). Perhaps contact induces emotional distress, or more distressed family members seek reassurance by increasing contact. However, studies in this new literature have thus far assessed only main effects of communication on family member functioning whereas the present study investigates the possibility that contact might moderate the effects of specific absence variables on family member functioning, with either a buffering or intensifying effect on symptoms.

The Present Study

The present study introduces the Timeline Followback Military Family Interview (TFMFI) to collect precise information on salient family events over the past 5 years and how those events coincided with the service member’s absences from the family. This procedure, which is adapted from the substance use literature (Fals-Stewart, O'Farrell, Freitas, McFarlin, & Rutigliano, 2000; Sobell & Sobell, 1992), makes use of a calendar and key dates to serve as anchors and memory aids to obtain retrospective estimates of a specific behavior over a specified time period. The goal here is to assess two absence dimensions – total cumulative time of absence related to military service (i.e., deployments, trainings, duty assignments), and number of missed family events during absences. We hypothesize that these more nuanced measures of absence will better capture the implications of the service member’s absence and thus will account for significant additional variance, over number of deployments, in mother and youth symptoms of depression and anxiety (Hypothesis 1). We also anticipate that the frequency of contact with the service member will moderate the impact of service member absence on family members’ symptoms; we are testing this as a non-directional hypothesis as it is an open question whether contact heightens or lessens the psychological symptoms associated with absence (Hypothesis 2).

Method

Participants

The current study assessed 70 mother-adolescent dyads from United States military families living in Southern California. Dyads were eligible if they had: (a) a father/husband serving on active duty, (b) a mother not currently on active duty, (c) a child aged 14–18, and (d) ability to complete study measures in English. Families were recruited through advertisements in military newsletters/social media/websites; announcements via family readiness, school liaison, and military housing personnel; flyers posted – with permission from the Naval Medical Center and local commands – in base hospitals/clinics; and referrals from study participants for a study of “the challenges and benefits of military family life”. Of 123 families who initially contacted us, 70 families (56.9%) completed all study procedures; 20.3% (n = 25) did not meet eligibility requirements, 13.8% (n = 17) did not respond to our contact attempts following their initial inquiry, and 8.9% (n = 11) declined to participate. The most common reasons for declining participation were difficulty scheduling (n = 6) and not wishing to be videotaped (n = 2).

Participating youth (n = 38 female) ranged in age from 14.0 to 18.9 (M = 16.0, SD = 1.2). About a third of youth and mothers self-identified as Hispanic/Latino (31.4%). In terms of race, 64.3% self-identified as White/Caucasian, 4.3% as American Indian or Alaska Native, 2.1% as Native Hawaiian/Pacific Islander, 2.9% as Asian, 5.7% as Black, and 20.7% as multiracial or other. Mothers’ average age was 39.9 (SD = 4.8, range = 31.6–53.0). Ninety percent of mothers had attended some college, and 40% had completed a bachelor’s degree or equivalent; 31.4% worked for pay outside the home.

In over one-third of participant families (34.3%, n = 24), the service member was the youth’s stepfather; among these families, the youth reported an average of 10.9 years of co-residence with the service member (SD = 4.0, range = 4.0–17.5). All but one mother was biologically related to the target youth. All target youth and their service member father or stepfather shared a primary residence. Families averaged 2.8 children in the home (SD = 1.0, range = 1–6), including the target youth, and the majority of target youth was eldest child (85.7%). Fifty-three percent of families were living in military housing (either on a military installation or in an off-base military housing community). Service member parents were serving in the U.S. Marine Corps (52.9%), Navy (34.3%), Air Force (8.6%), Army (2.9%) and Coast Guard (1.6%). In our sample, 25.7% of service members were mid-grade enlisted personnel, 40.0% were senior enlisted,18.6% were junior officers, and 15.7% were senior officers. Service members had deployed an average of 4.0 times during their careers (SD = 3.1, range 0–15). Four service members had never deployed, and three were deployed at the time of the interview. Eleven mothers were veterans/retirees of the U.S. military.

Procedures

Mother-adolescent dyads were assessed in a 2 1/2 hour meeting in their homes without the presence of the service member. Study procedures were approved by the University Institutional Review Board, and participants provided written informed consent/assent. The TFMFI was conducted simultaneously with both the mother and youth, was video-recorded and lasted about 45 minutes. Calendars for each of the past five years were used to prompt the family members’ memories about the service member’s deployments or other military related absences and other important family events (e.g., births, deaths, moves, etc.). The calendars were used flexibly and iteratively in the interview; families who could recall the service member’s absences clearly placed these on the calendar first and then filled in family events, whereas other families identified service member absences by first identifying key family events. Figure 1 shows a one-year sample of a fictitious but representative family time line. When the timeline was nearly complete, the experimenter queried about whether other important family events occurred during identified service member absences. The experimenter made sure that both the mother and youth contributed to the completion of the timeline, and the discussion between family members allowed them to prompt one another’s recall, which resulted in more comprehensive information. Each videotaped interview was reviewed by a trained undergraduate research assistant to double-check the accuracy of the information recorded on the calendars and standardized forms.

Figure 1.

Sample of One Year of a Family Timeline

We additionally assessed the mother’s and adolescent’s private reports of mental health symptoms and general adjustment by having each family member independently complete questionnaires on laptop computers that we provided.

Measures

Number of Recent Deployments

We counted the service member’s deployments during the previous five years, as indicated by the TFMFI procedure.

Service Member’s Cumulative Time Away

Using the calendar portion of TFMFI, we first calculated the total time (in months) the service member had spent away from the family due to deployment, temporary duty assignment, or training during the previous five years.

Important Family Events Missed

Drawing on both objective and subjective responses to TFMFI questions, we used the recorded interviews and the notations on calendar to ‘count’ the number of significant family life events the dyad reported the service member had missed during periods of absence. The following ten types of events, which are potentially disruptive to the family’s established patterns, were included in this measure: pregnancy/birth/adoption (reported by 24.3% in the past 5 years), death (38.6%), wedding (24.3%), graduation (from a school, not completion of a school year; 40.0%), residential move (31.4%), family member moving in (12.9%) or out of the service member’s home (10.0%), divorce/marital separation of family member (11.4%), serious illness/injury/hospitalization (54.3%), child(ren) beginning school year (72.9%), and job loss (10.0%). We only counted residential moves, children beginning school year, and job loss for the nuclear family; other events could involve extended family members.

Contact Frequency

Each participant answered the following question about their frequency of contact with the service member during the most recent deployment: “How often did you communicate with your spouse/parent?” Participants indicated contact frequency on a six-point ordinal scale: (0) Never, (1) Once a month, (2) Once a week, (3) 2–3 times a week, (4) Once a day, or (5) Several times a day. Mothers reported significantly more contact with the service member (M = 3.39) than did youth (MYouth = 2.52), paired t(63) = 5.23, p < .001.

Youth Internalizing Symptoms

Youth completed a 26-item version of the Children’s Depression Inventory (CDI; Kovacs, 1992); due to ethical/reporting concerns, the suicidal ideation item was excluded (Cronbach’s α = .83). Youth also completed the 47-item Revised Children’s Anxiety and Depression Scale (RCADS; Chorpita et al., 2000); the 37 items assessing anxiety symptoms across five DSM-IV-TR anxiety disorders were used here (α = .93). Participants were permitted to skip any items they preferred not to answer; 49 CDI items were skipped (2.7%) across 28 youth and 18 RCADS items (0.7%) were skipped across 11 youth. Mean imputation was done at the subscale level to create participants’ total scores.

Mother Internalizing Symptoms

Mothers reported their anxiety symptoms on the 21-item Beck Anxiety Inventory (BAI; Beck, Epstein, Brown, & Steer, 1988) and their symptoms of depression using a 20-item version of the Beck Depression Inventory-II (BDI-II; Beck, 1996), with the omission of the suicidal ideation item. Internal consistency in this sample for the BAI and BDI-II were α = .92 and .93 respectively. As with youth, we used mean imputation for missing data (0.5% of the BAI and 1.4% of the BDI).

Analytic Plan

We first screened the following potential covariates by assessing their bivariate correlations with mother and youth symptoms: youth gender, youth age, youth birth order, mother age, mother veteran status, service member rank, and stepfamily status. Stepfamily status was significantly associated with youth outcomes, and was retained as a covariate in youth models. Service member rank was significantly associated with mothers’ symptoms, and was retained as a covariate in those models. We ran parallel hierarchical multiple regression models for mothers and youth and separately for anxiety and depression. Covariates were entered in step 1 (number of recent deployments, rank [for mothers], and stepfamily status [for youth]), and absence variables (cumulative time away and events missed) were entered at step 2. In follow-up analyses assessing contact frequency as a moderator of the impacts of absence variables, the contact term and interaction term were added at step 3.

Results

Table 1 presents descriptives and bivariate Pearson correlations for the study variables. As expected, our indices of absence – number of recent deployments, cumulative time away, and number of important family events missed – were all moderately positively associated (rs = .27–.33). Cumulative time away was positively associated with mothers’ depression and anxiety symptoms, but was not associated with youth symptoms; number of important family events missed was positively associated with mothers’ depression symptoms and marginally associated with the youth’s depression symptoms. Number of recent deployments was, unexpectedly, inversely related to youth anxiety symptoms. Service member rank was inversely associated with mothers’ anxiety symptoms. Youth contact frequency was marginally inversely associated with youth symptoms but not significantly associated with any study variables. Symptoms of anxiety and depression were strongly correlated for mothers and for youth.

Table 1.

Descriptive Statistics and Bivariate Correlations Among Study Variables

Variable	M	SD	1	2	3	4	5	6	7	8	9	10
1. Deployments	1.26	0.96
2. Time Away (months)	15.52	9.51	.30*
3. Events Missed	4.84	3.56	.27*	.33**
4. Mother Contact	3.37	1.12	−.01	.10	−.01
5. Youth Contact	2.51	1.05	−.22†	−.13	−.08	.24†
6. Mother ANX	7.10	9.30	−.08	.34**	.20†	−.09	−.08
7. Mother DEP	10.87	9.76	−.04	.26*	.26*	.06	−.17	.77**
8. Youth ANX	24.39	14.45	−.24*	.06	.05	−.14	−.24†	.21†	.16
9. Youth DEP	10.11	6.18	−.04	.03	.21†	−.14	−.23†	.17	.22†	.63**
10. SM Rank	8.86	3.17	−.01	.07	−.17	.08	−.01	−.26*	−.13	−.15	−.14
11. Stepfamily	N/A	N/A	.12	−.21†	.03	−.02	.04	.01	.02	−.22†	−.34**	−.38**

Note.

^**p < .01,

^*p < .05,

^†p < .10.

SM = Service Member. ANX = Anxiety symptoms. DEP = Depression symptoms. N = 70 for all variables but contact; 4 service members had never deployed and 2 youth chose to skip this question (mother contact N = 66, youth contact N = 64). Deployments, time away, and events missed were all assessed for the 5-year period preceding the interview.

Effects of Dimensions of Absence on Family Members’ Symptoms

Table 2 presents the regression analyses for mothers’ and youths’ symptoms related to the two absence variables. For youth, as hypothesized, the number of important family events missed was associated with elevated youth symptoms of depression even when accounting for the influence of stepfamily status, and duration and frequency of absence. Youth in stepfamilies also showed significantly lower levels of depressive symptoms. By contrast, only the number of recent deployments was uniquely associated with youth anxiety symptoms, although not in the expected direction; higher number of deployments was linked with lower youth-reported anxiety.

Table 2.

Hierarchical Regression Analyses of Family Member Symptoms on Dimensions of Service Member Absence

	Youths’ Anxiety Symptoms				Youths’ Depression Symptoms
Variable	B	SE B	β	ΔR2	B	SE B	β	ΔR2
Step 1 – Covariates				.10*				.12*
Stepfamily	−5.48	3.69	−.18†		−4.86	1.51	−.38**
Number of Recent Deployments	−3.99	1.91	−.26†		−.18	.78	−.03
Step 2 – Absence Variables				.01				.06†
Cumulative Time Away	.09	.20	.06		−.09	.08	−.13
Important Events Missed	.44	.51	.11		.47	.21	.27*

	Mothers’ Anxiety Symptoms				Mothers’ Depression Symptoms
Variable	B	SE B	β	ΔR2	B	SE B	β	ΔR2
Step 1 – Covariates				.08†				.02
Rank	−.82	.32	−.28		−.35	.36	−.11
Number of Recent Deployments	−2.15	1.12	−.22†		−1.72	1.25	−.17
Step 2 – Absence Variables				.16**				.12*
Cumulative Time Away	.39	.12	.40**		.26	.13	.26*
Important Events Missed	.20	.31	.08		.55	.35	.20

Note.

^**p < .01,

^*p < .05,

^†p < .10.

Full model R² = .11, F(4, 65) = 2.06, p = .097 for youths’ anxiety symptoms and R² = .18, F(4, 65) = 3.64, p = .01 for youths’ depression symptoms. Full model R² = .24, F(4, 65) = 5.19, p = .001 for mothers’ anxiety symptoms and R² = .14, F(4, 65) = 2.64, p = .04 for mothers’ depression symptoms.

Models for mothers’ symptoms showed similar patterns of results across anxiety and depression symptom clusters. Cumulative time away was positively associated with symptoms of both depression and anxiety, whereas missed events were not associated. Service member rank was negatively associated with mothers’ anxiety.

Frequency of Contact Moderates Impacts of Absence

We then tested the hypothesis that contact with the service member would moderate the impact of absence on family members’ symptoms. For youth, frequency of contact moderated the effect of cumulative duration of recent absence on youth anxiety, B = −.42, t = −3.14, p = .003. At low (i.e., M - 1SD) levels of contact, the association between cumulative time away and youth anxiety was positive, B = 2.41, t = 3.20, p = .002; at high levels of contact (i.e., M + 1SD), the association between cumulative time away and anxiety was negative, B = −.55, t = −2.34, p = .023 (see Supplemental Figure 1). Thus, more frequent contact with the service member appears buffer youth against anxiety due to periodic parent absence. No other significant interactions between contact frequency and absence variables emerged for youth or mothers.

Discussion

This study introduces the Timeline Followback Military Family Interview (TFMFI), to operationalize two dimensions of parental absence: important family events that coincided with absence and total cumulative time away. In support of Hypothesis 1, this more nuanced examination of the impact of service members’ absence revealed that the number of important family events missed was associated with elevated youth depression symptoms, even after accounting for the number of recent deployments and cumulative duration of absence; cumulative duration of absence was associated with mothers’ elevated symptoms of depression and anxiety. In partial support of Hypothesis 2, more frequent contact with the service member buffered youth from anxiety symptoms in the face of more extensive durations of absence.

This study provides an important preliminary look at variables that may explain the impact of periodic service member absence on family members’ mental health symptoms. Our findings suggest mothers may be particularly affected by the amount of time they are required to be separated from their spouses and to serve as “temporary single parents” to their children. In contrast, youth were not significantly affected by the overall duration of the parent’s absence, but instead were affected by whether his absence co-occurred with significant family events. Further research on the nature of these influential missed events and the youth’s role in the events may clarify why youth find the parent’s absence difficult. Perhaps these milestone events and celebrations seem incomplete without the father. Alternatively, with 86% of the participating youth being the eldest child, these adolescents may be assuming some of the adult burdens and responsibilities in the family, particularly in dealing with significant stressful family events.

Contrary to previous findings that frequency of contact with the service member is associated with youth psychosocial stress (Houston et al., 2013), our results show contact with the service member to be protective against youth anxiety symptoms. Understanding what makes contact with the service member risky versus protective is an important topic for future research. Interpretation of this significant finding as well as our non-significant findings about contact is limited by our use of a one-item, retrospective global question. It would be helpful to know more about the nature of the contact and to assess more immediate reactions to contact. It also might be useful to know about the preexisting nature of the relationships among family members as that may influence whether contact during absences is fraught with irritations and frustrations or is an opportunity for tenderness and comfort. Contact between families and service members clearly can be a complicated matter, sometimes reassuring but also sometimes unsatisfying, e.g., family members have reported that the service member may be protecting them from disturbing information and also that they, too, protect the service member from stresses at home (e.g., Joseph & Afifi, 2010). For families who anticipate future absences, it might be useful to plan for contact so they have ways of connecting emotionally even if there are barriers to sharing certain information.

Our study describes a timeline follow-back procedure, which may be useful in clinical research and in applied settings to assess military families’ experiences of the overlapping challenges of military service and family life. Though commonly used in other clinical contexts (e.g., Fals-Stewart, O'Farrell, Freitas, McFarlin, & Rutigliano, 2000), these procedures are underutilized in family assessments despite the following unique advantages. First, these assessments remind the interviewer that deployments vary in their character due to both military- and family-specific features. Second, the clinician obtains information about the objective features of the service member’s absences as well as family members’ subjective perceptions of absences. Third, inquiring about families’ deployment experiences provides the clinician a way to recognize and validate the challenges some families face during absence, without overemphasizing the impact of service-related absence for families who have not faced as many concurrent challenges. Fourth, the TFMFI provides a way to generate meaningful discussion between family members and perhaps can be used to generate a shared platform to address current or upcoming challenges. Though used here without the service member, including the service member would likely provide an additional layer of clinically rich information. Thus, an instrument like the TFMFI is a format worthy of consideration for clinical assessment with families who face periodic temporary absences and need not be limited to military absences.

There are several limitations of our study that are worth noting. Our selection of a 5-year timeframe to assess absence was rather arbitrary, and was dictated more by time constraints for data collection than by empirical or theoretical considerations, as absences occurring earlier may also certainly exert enduring influences on mother and youth. As noted, the construct of contact could be improved with more detailed information, including the impact of contact. Similarly, our preliminary count-based assessment of missed events could be enhanced by allowing participants to nominate events, rate event impact/salience, or differentiate nuclear family versus extended family events. Our sample was somewhat skewed to higher military ranks and supervisory positions because service members with adolescent children have largely committed to career-length service, and have advanced to these positions. Relatedly, since we assessed mid- to late-adolescent youth, our findings may not generalize to younger children. Our sample differs from the samples common in the literature on military families, in that few of our families were experiencing a current deployment; our results may thus best be indicative of potential risk factors during the post-deployment/reintegration phase rather than risks of deployment per se. Finally, our use of standard mental health symptom inventories focuses on more clinically-significant manifestations of distress reactions although we suspect family members’ responses to the challenges addressed here may be evidenced more in daily individual and relationship indicators, such as feeling worried (particularly about future absences), overwhelmed, or frustrated by the lack of predictability about the service member parent’s presence in the family.

Despite these limitations, our study highlights the importance of more nuanced indices of parental absence and offers an assessment strategy for collecting information about the family’s unique history of service member absence in tandem with the civilian family members’ lives. With a parental absence and even serial absences a common phenomenon for a number of families, family clinicians and researchers need to attend to multidimensional impacts of the absences, even after the family has reunited. Moreover, the discrepancy between our results on contact and previous findings underscore how little is known about the impact of contact between absent service members and family members at home. Pre-deployment briefing opportunities that often are available to military members and spouses could address expectations as well as possible benefits and challenges of contact. The data here—though preliminary—point to the duration of absence, missed events, and contact as relevant dimensions in understanding how families can be affected the absence of a military member. Though service members’ absences were connected to family members’ mental health symptoms, these results do not preclude the possibility that families also might be strengthened by such challenges, which is an important direction for future research.