Post-adoption Depression: Parental Classes of Depressive Symptoms Across Time

Abstract

Background

Approximately 10%-15% of birth mothers and fathers experience postpartum depression, but reports of depressive symptoms in adoptive parents are more variable. Findings from investigators range from 10%-32%, which may mask the experiences of distinct groups of adoptive parents from pre-to post-placement of a child.

Methods

We performed latent class growth analysis using the Center for Epidemiologic Studies-Depression scores of 129 primarily heterosexual, adoptive parents (50% females) for three time points: 4-6 weeks pre-placement of the child, 4-6 weeks post-placement, and 5-6 months post-placement. Mixed effects models by parent depressive trajectories were also conducted for three types of variables: interpersonal, psychological symptoms, and life orientation.

Results

Five classes of depressive symptom trajectories were found. The majority of parents (71%) belonged to a class with low levels of depressive symptoms across time. However, two classes of parents were above the threshold for depressive symptoms at placement, and three classes of parents were above the threshold at 6 months post-placement. The majority of interpersonal, psychological symptom, and life orientation variables were significant across classes and by time.

Limitations

The homogeneity of the sample calls for replication of study findings.

Conclusions

An explanation for inconsistencies in the range of adoptive parent depressive symptoms may be explained by different subgroups of parents who vary by their trajectory of depressive symptoms before and after placement of the child. Adoption and mental health professionals should be aware that select adoptive parents may struggle pre-and post-placement of a child.

Introduction

Placement of an adopted child in the home has been typically viewed as an endpoint in the process and in research investigations surrounding adoptive parents. Relatively little attention has been given to families' ongoing challenges during the vulnerable post-placement period when almost 20% of adoptive families engage in family counseling (Vandivere, Malm, & Radel, 2009). An assumption is often made that such counseling is based on the child's needs; however, evidence suggests that adoptive parents may struggle with depressive symptoms in the post-placement time period. It is critical to assess depression in parents; adoptive parental major depression has been associated with a significantly greater risk for major depression and disruptive behavior disorders in adopted adolescents (Tully, Iacono, & McGue, 2008), externalizing behaviors in adopted toddlers (Pemberton et al., 2010), and internalizing and externalizing behaviors in children adopted by both same-sex and opposite-sex couples (Goldberg & Smith, 2013). Therefore, addressing post-adoption depression (PAD) in parents is important for both parental and child wellbeing.

Since 1995, a limited number of investigations across disciplines have explored the phenomenon of parental PAD. Studies of PAD have reported a wide range of rates, from 8% to 32% (Dean, Dean, White, & Liu, 1995; Fields, Meuchel, Jaffe, Jha, & Payne, 2010; Foli, South, & Lim, 2012; Gair, 1999; Mott, Schiller, Richards, O'Hara, & Stuart, 2011; Senecky et al., 2009). Many of these studies were limited: using small samples, including only parents of children adopted from abroad, or excluding adoptive fathers (e.g., Fields et al., 2010; Gair, 1999; Senecky et al., 2009; Viana & Welsh, 2010). Despite methodological limitations, these and more recent studies provide preliminary indications of the extent of the problem. In the current study, we report on a longitudinal investigation of adoptive parents in which we attempt to identify parents' trajectories of depressive symptoms following the placement of the child in the home.

Adoptive Parents' Transition and Depressive Symptoms

The transition to adoptive parenthood is individualized and unique, with parents experiencing emotions ranging from joy to fear. The continuum of experiences of new adoptive parents was supported in qualitative findings based on interviews with parents (McKay & Ross, 2010). Investigators identified the meta-themes of challenges and facilitators, individualized factors that interact with the other. For example, family support, or the lack of such support, may be a facilitator for one parent and a barrier for another (McKay & Ross, 2010). In a systematic research synthesis, McKay, Ross, and Goldberg (2010) examined 11 studies that reported findings of adoptive parent mental health, physical health, and intimate partner relationship satisfaction through the immediate post-adoption period (3 years after placement). Conclusions included the scarcity of research that examines adoptive parent transitions in the post-placement time period and the “relatively common” (p. 38) prevalence of PAD; however, studies related to PAD often did not examine both the difficulties of parenting and buffers to those difficulties, such social support and resilience. Conclusions could not be drawn for physical health nor couple satisfaction as only two studies reviewed included these variables (McKay et al., 2010). Since then, two studies have examined relationship functioning of adoptive parents; significant predictors of depressive symptoms included coping and relationship maintenance (Goldberg, Smith, & Kashy, 2010), and socioeconomic status, anxiety, partner support, partner's enthusiasm for being an adoptive parent, feelings of rest, and total number of adopted children (South, Foli, & Lim, 2013).

To date, studies providing rates of PAD have varied in who were sampled (primarily mothers), the depressive screening measures used, and when and how parents were assessed for depression. Senecky et al. (2009) measured depressive symptoms of 39 mothers in Israel who were registered with international adoption agencies using the Edinburg Postnatal Depression Scale (EPDS), the Beck Depression Inventory (BDI), and the Brief Symptom Inventory (BSI). Clinically significant symptoms of depression were found in 25.6% of the mothers prior to adoption and in 15.4% six weeks after adoption. Fields and colleagues (2010) found an even higher rate of depression (27.9% at 0-4 weeks and 25.6% at 5-12 weeks) in 86 adoptive mothers of infants who were assessed retrospectively based on the first year post-placement. Most recently, Nguyen and Gunnar (2014) assessed two groups of adoptive mothers, one of post-institutional care children (high risk group) and one of overseas foster care children, and one group of birth mothers at 1-3 months and 8-9 months post-adoption/post-birth. They found no significant differences in Center for Epidemiologic Studies Depression (CES-D) scores between the three groups of mothers. However, a higher percentage of mothers who had adopted children from post-institutional care declined to participate in the study (45%) compared with the other two groups (overseas foster care [15%] and non-adopted mothers [10%]; Nguyen & Gunner, 2014). The under-representation of this group of mothers is important because of the unique needs and negative outcomes that may arise in children who have been exposed to such institutional settings (e.g., Hawk & McCall, 2010).

In a series of preliminary studies, our research team recruited both mothers and fathers who had adopted a child up to two years before the research assessment. We found PAD rates from 18%-26% in mothers and 11%-24% in fathers (Foli, South, & Lim, 2012; Foli, South, Lim, & Hebdon, 2013). We also examined several variables that might have predicted depressive symptoms in this population. Depression in adoptive mothers, as measured by the CES-D, could be explained by nine variables: feeling of rest, self-esteem, history of depression, perceived friend support, parent-to-child bonding, marital satisfaction, and the items related to expectations (of themselves as parents, the child, and family and friends) (Foli et al., 2012). In adoptive fathers, regression analysis revealed four significant predictors of depressive symptoms as measured by the CES-D: age of the adopted child, partner satisfaction scores, perceived friend support, and scores on unmet expectations of the child (n=31; R²=0.82).

There has been a tendency to compare birth parents' and adoptive parents' rates of depression in PAD studies, with authors generally finding that rates have not significantly differed (e.g., Mott et al., 2011; Nguyen & Gunnar, 2014). However, given the differences in contextual features between adoptive and birth parents, such comparisons of positive screening rates may limit a deeper understanding of depression in adoptive parents. Examples of contextual differences between adoptive and birth parents include the home study process during which the adoptive parent is required to report physical, financial, social, and home details; the variable time frame during which adoptive parents wait for a child; educational and socioeconomic levels of adoptive parents; and issues of infertility. Depressive symptoms have also been examined in same-sex adoptive parents. In gay men and lesbian adoptive parents, variables that were related to lower depressive symptoms included higher support from the workplace and family, and relationship quality; the researchers also found increasing depressive symptoms across time (Goldberg & Smith, 2011). Given these contextual features of adoption, the experience of depression in adoptive parents may be quite different than in birth parents (Fontenot, 2007; McKay & Ross, 2009).

In addition to differences between birth and adoptive parents' experiences of postpartum/post-adoption depression, we further hypothesize that there may be heterogeneity in depressive symptoms among adoptive parents which may be hidden by merely reporting rates across the entire sample at a given point in time. Cross-sectional investigations have been common in the study of PAD (Gair, 1999; Mott et al., 2011; Foli et al., 2012; Foli et al., 2013), but provide limited understanding into how symptoms may vary over time.

Theoretical Framework

A middle range theory of parental post-adoption depression has been forwarded by Foli (2010) and subsequently supported in several studies (Foli et al., 2012; Foli, South, & Lim, 2014; Foli, South, Lim, & Hebdon, 2012; 2013). In summary, the theory describes the influence of the adoption process on prospective parents, including the uncertainties and inherent stressful nature of the process. During this process, adoptive parents create expectations in four dimensions: of self as parent, of the child, of family and friends, and of society and others (Foli, 2010). When a dissonance occurs between parental expectations that are created pre-placement and the reality that is faced post-placement, depressive symptoms may emerge. As a child is integrated into the home, we believe parents' experiences and adaptation also change across time. What is needed moving forward is an understanding of how depressive symptoms in adoptive parents fluctuate across the transition from pre- to post-placement. We have included several variables in the current study that surround parental expectations, including psychological variables, child characteristics, friend and family support, and orientation to life.

The Current Study

The range of rates presented in the literature makes conclusions related to the prevalence of PAD difficult to ascertain and therefore, debate continues regarding how widespread the problem of depression is following placement of the child in the home. Estimating rates across samples may obscure the fact that there are subsets of adoptive parents are at significant risk for clinical levels of depression. The goal of the current study was to determine depressive symptom screening rates and examine how parental symptoms may vary by distinct trajectories of symptoms. In this way, we attempted to provide a deeper understanding of how prevalent the problem of depression is in adoptive parents and also provide a more complete picture of depressive symptoms experienced by these parents.

We attempted to eliminate the methodological weaknesses of past investigations by 1) collecting data across time, at both pre- and post-placement time points; 2) capturing information from parents who choose to adopt through private, public and inter-country paths; and 3) enrolling both mothers and fathers when the adoptive family includes two parents. Identifying at-risk parents requires a comprehensive, longitudinal investigation of both mothers and fathers that can provide evidence as to the course of depressive symptoms in the first six months following placement of the adopted child. We hypothesized that there may be different trajectories (distinct classes) of depressive symptoms within adoptive parents over time; however, given the exploratory nature of this analysis, we did not have specific predications about the number or pattern of different trajectories that might emerge. Further, we expected that key variables will differ between these distinct classes and over time as the parents transition from pre-to post-placement. Based on previous research, including our own, we examined the following variables for differences between trajectories: demographics, psychological symptoms, interpersonal variables, and orientation to life. This study was part of an investigation whose additional findings are reported elsewhere.

Methods

Study Design and Sample

Data were collected via an online survey from adoptive parents, the majority of whom were clients of the largest adoption agency in the country. The study was advertised through electronic and hard copies of flyers distributed by adoption agency staff, advertisements placed in a quarterly magazine sent to clients of the agency, and a webinar presented by one of the investigators on parental emotional health to clients of the agency. Parents contacted members of the research team to enroll in the study either through electronic mail or phone calls. Participants who did not indicate consent during the online survey were unable to proceed.

This study was approved by the institutional review boards at XXX University and University of XXXX. Individuals enrolled in the study met the following inclusion criteria: Internet access; at least 21 years of age; the ability to speak, read, and understand English; and anticipate placement of the child within approximately 4-6 weeks after completion of pre-adoptive questionnaires. Gift cards were given to each participant for completion of the survey at each time point: 4-6 weeks pre-placement (T1), 4-6 weeks post-placement (T2), and 5-6 months post-placement (T3). Data were collected between February 2013 and December 2014, and weekly reports of CES-D scores were reviewed by the research team. Parents exceeding the threshold of symptoms were notified and given information related to mental health resources.

Sample Characteristics

One hundred twenty-nine (129; 2 homosexual, 2 single, 125 heterosexual), adoptive parents enrolled in this study, with 76% of them recruited through the largest adoption agency in the U.S.¹. Table 1 represents subject characteristics of the parent participants and their adopted children. For the child's characteristics, we used mothers' reports at T2 (the first assessment post-placement). If the mother's report was not accessible, the father's report was used. If more than one child was adopted at the same time, the older child's characteristics were used. Five couples adopted two children and one couple adopted three children at the same time.

Table 1. Subject Characteristics.

Parent Variablea	n (%)	Child Variableb	n (%)
Gender		Gender
Male	59 (46)	Male	32 (50)
Female	70 (54)	Female	32 (50)
Parent Age (yrs.), mean (SD)	38.1 (5.3)	Age (mos.), mean (SD)	28.1 (38.3)
Race/Ethnicity		Child's Race/Ethnicity
Caucasian or White	120 (93)	White or Caucasian	17 (27)
Other	9 (7)	Black or African	12 (19)
Income		Hispanic	2 (3)
Under $75,000	42 (34)	Asian or Pacific Islander	22 (35)
$75,000 to under $100,000	38 (30)	Multiracial	8 (13)
More than $100,000	45 (36)	Other	1 (2)
Education		Transracial Family
Less than four year college graduate	23 (18)	No (Same as either of parents)	24 (35)
Four-year college graduate	56 (43)	Yes (Different from both parents)	38 (65)
Post-graduate	50 (39)	# of Adopted children
Job Status		1	56 (90)
Full time	85 (66)	2	5 (8)
Part time	14 (11)	3	1 (2)
Other	30 (23)	Special Need
Infertility		Yes	28 (44)
Yes	56 (45)	No	34 (55)
No	68 (55)	Type of Special Need
Bothered by infertility, mean (SD)α	2.5 (1.2)	Physical	13 (21)
Religion		Emotional/Psychological	6 (10)
Protestant	56 (43)	Developmental/Cognitive	5 (8)
Roman Catholic	15 (12)	Other	12 (19)
Non-denominational Christian	45 (35)	Missing	26 (42)
Other	13 (10)	Type of Adoption
Level of Religiosity, mean (SD)β	6.1 (1.2)	Public (domestic)	7 (11)
Motivation of Adoption		Private (domestic)	20 (32)
Build family through adoption	97 (76)	Inter-country	29 (47)
God's calling	90 (71)	Other	6 (10)
Want a child of specific gender	9 (7)	Length of Waiting Time (mos.), mean (SD)	10.7 (11.5)
Other	16 (13)

^an=129 parents.

^bn=64 couples/singles

Child variables were reported based on the older child if more than one child were adopted at the same time.

^α1=strongly disagree-5=strongly agree.

^β1=not at all-7=very strongly religious.

The mean age of parent participants was 38.1 (SD=5.3). Most of them were Caucasian or white (93%), over a third (36%) reported incomes of greater than $100,000 per year and had completed education (39%) past a four-year degree. Fifty percent of the adopted children were females; 44% of the sample was categorized as having special needs by their adoptive parents. The mean age of the adopted (older) children was 28.1 months (SD=38.3) and 47% of the children were adopted from outside of the U.S.

Measures

All of the following measures were administered at all three time points, with the exception of the parent's and child's demographic measures, which were completed at T1² and T2, respectively. In the current study we used several standard instruments as well as individual items. Items used to measure characteristics of the sample included parent's and child's gender, parent's and child's age, parent's and child's race/ethnicity, income, education, job status (full-or part-time), history of infertility, how much the parent was bothered by infertility (rated 1-5; 1=strongly disagree, 5=strongly agree), religion, level of religiosity (rated 1-7; 1=not at all, 7=very strongly religious), number of adopted children, the type of adoption (public, private, inter-country, and other), and whether the parent considered the child to have special needs. Additional single items were posed: feelings of being rested (“I feel rested when I wake up,” 1=strongly disagree, 5=strongly agree); overall life satisfaction (rated 1=worst possible life overall, 10=best possible life overall); sexual relationship (how satisfied with sexual relationship with partner: rated 1=very dissatisfied to 7=very satisfied); partner enthusiasm toward being a parent (rated 1=very strongly unenthusiastic to 7=very strongly enthusiastic); and partner enthusiasm toward being an adoptive parent (rated 1=very strongly unenthusiastic to 7=very strongly enthusiastic)³. All items were recoded as necessary so that higher scores were in the more favorable (e.g., more well-rested) direction.

Center for Epidemiological Studies Depression Scale (CES–D)

The CES-D is a short, self-report scale designed to measure symptoms of depression in the general population (Radloff, 1977). Twenty items are rated on a 4-point scale (0-3), and scores range from 0 to 60. A cut off of ≥16 was used to determine a positive depressive symptom screen. The CES–D has been widely used as a screening instrument with nonclinical populations. Cronbach's alphas for the current study were 0.88 at T1, 0.92 at T2, and 0.90 at T3.

Postpartum Depression Predictors Inventory-Revised (PDPI-R; Modified)

The PDPI-R assesses 10 risk factors prenatally: marital status, socioeconomic status (SES), self-esteem, prenatal depression, prenatal anxiety, pregnancy intendedness, history of depression, social support, marital/partner satisfaction, and life stress; and three risk factors postnatal: child care stress, infant temperament, and maternity blues (Beck, 1998; 2002; Beck, Records, & Rice, 2006). A modified version of the PDPI-R was used by Foli and colleagues (2012) to assess risk for depressive symptoms in 300 adoptive mothers. Adaptations included substituting the word “infant” for “infant/child” and measuring life stress with other items. Modifications used in the current study were limited to substituting “infant/child” in the items. Cronbach's alphas for this study were 0.66 at T1; 0.68 at T2; and 0.69 at T3. However, removing three items that demonstrated little variation in the sample (marital status, SES, and unplanned adoption) improved Cronbach's alphas to 0.70 at T1; 0.73 at T2; and 0.73 at T3.

Perceived Social Support (PSS-Fa & PSS-Fr)

These two parallel scales measure perceived social support from family and friends; each scale consists of 20 items, with responses of “Yes,” “No,” and “Don't Know” (example: “My family/friends give me the moral support I need.”) (Procindano & Heller, 1983). Cronbach alphas of PSS-Fa and PSS-Fr for this study were 0.96 and 0.95 at T1; 0.94 and 0.90 at T2; and 0.94, and 0.92 at T3, respectively.

Intimate Relations Questionnaire (IRQ)

This 25-item instrument has four subscales that measure love (10 items), conflict/negativity (5 items), ambivalence about whether to continue the relationship (5 items), and maintenance in a relationship (5 items; Braiker & Kelly, 1979). These four dimensions have been found to reflect the various areas of a couple's relationship. Cronbach alphas of love, conflict, ambivalence, and maintenance for this study were 0.86, 0.80, 0.83, and 0.75 at T1; 0.88, 0.80, 0.88, and 0.78 at T2; and 0.90, 0.85, 0.87, and 0.81 at T3, respectively.

Inventory of Depression and Anxiety Symptoms (IDAS)

The 64-item IDAS (Watson et al., 2007) contains 10 symptom scales. To reduce participant burden, we included the following subscales, which were selected on dimensions not measured by other tools used in the study: ill temper, social anxiety, panic and traumatic intrusions. Cronbach alphas of ill temper, social anxiety, panic and traumatic intrusions for this study were 0.79, 0.52, 0.60, and 0.75 at T1; 0.73, 0.71, 0.64, and 0.83 at T2; and 0.83, 0.78, 0.54, and 0.79 at T3, respectively.

Life Orientation Test-Revised (LOT-R)

The LOT-R was developed to assess individual differences in generalized optimism (i.e., generalized favorable expectancies for the future) versus pessimism (Scheier, Carver, & Bridges, 1994). Cronbach's alphas for the current study were 0.84 at T1, 0.86 at T2, and 0.82 at T3.

Statistical Analyses

Descriptive statistics for parent and child characteristics were calculated using frequencies for categorical variables or means and standard deviations for continuous variables. For the outcome variable of CES-D, we replaced each missing item with the item-median value if less than 20% of the items of the CES-D were missing. Latent class growth analysis (LCGA) was used to estimate different trajectory classes of the CES-D scores over the three time points. In the LCGA, we allowed the linear and quadratic terms to differ for each latent class. To estimate the best number of latent classes, LCGAs were conducted successively from one to nine latent classes, using full information maximum likelihood (FIML) approach to handle missing data when CES-D scores were missing (i.e., when more than 20% of items were missing). Bayesian information criterion (BIC) was used to determine the number of trajectory classes, along with the other criteria of entropy, posterior probabilities, and number of parents in each class. A simulation study showed that BIC performed the best among information criteria-based indices (Nylund, Asparouhov, & Muthen, 2007). BIC is based on the likelihood function and introduces a penalty term for the number of parameters. The model with the smallest BIC is preferred. Entropy is defined as a measure of classification uncertainty. The value close to 1 indicates clear delineation of classes. The entropy value for our final model was .915 with average latent class probabilities all above .9. The posterior probability of a latent class is the average probability of the class members being in the class membership and close to 1.0 is preferred. All LCGAs were implemented in Mplus version 7.31 (Muthén & Muthén, 1998-2010), with 500 randomly generated starting values. To account for the non-independence in parents nested within couples, the COMPLEX and CLUSTER options were utilized in Mplus. These options account for non-independence of data by adjusting computation of standard errors and chi-square tests of model fit.

After identifying the latent classes of CES-D trajectories, we then used mixed effects models to evaluate the mean differences of these latent classes on the measures and individual variables. The identified latent classes (between-factor), time (within-factor), and their interaction were fixed effects and random intercepts and slopes by subject were included in the models. Tukey's post-hoc test was conducted to control for multiple comparisons for each fixed effect. In addition, we evaluated the differences between the classes on subject characteristic variables (e.g., education, type of adoption), using Kruskal-Wallis tests for continuous variables or Fisher's exact tests for categorical variables. All of these analyses were implemented in SAS 9.4 (SAS institute, 2013).

Results

Depressive Symptoms

The mean CES-D scores were 6.59 (SD=6.22; Range=0-32) at T1, 6.34 (SD=7.47; Range=0-42) at T2, and 6.05 (SD=6.86; Range=0-30). The percentage of parents who were clinically depressed (i.e., CES-D≥16) was highest immediately after placement of the child: 9.5% (n=116) at T1 (4-6 weeks pre-placement); 11.3% (n=115) at T2 (4-6 weeks post-placement); and 9.6% (n=104) at T3 (5-6 months post-placement).

Parental Classes of Depression

Table 2 shows the fit statistics for the LCGAs. BIC identified the six-class model as the best fitting. However, because the difference in BICs of the five- and six-class models were small (less than 10 units), and the likelihood ratio test between the models did not show a significant detriment in fit, we further explored the distributions and profiles of the latent classes between the five and six class models. The major difference between two models was that the six-class model had a very small group (n=2) comprised of individuals from two groups of the five-class model. In addition, the profiles of the smallest five classes of the six-class model were comparable to those of the five-class model. The entropy and average posterior probabilities of the five-class model were acceptable (entropy=0.915; average posterior probability=0.918-1.000). Hence, we selected the five-class model as our final model. The percentage of participants and the intercept, linear and quadratic term estimates are presented for each class in Table 3⁴. Based on the recommendations outlined by Jung and Wickrama (2008), each of the five classes represented more than 1% of the sample.⁵

Table 2. Model Fit Statistics of Latent Class Growth Analysis with 1-9 Classes.

Number of Classes	LL	BIC	Entropy	Posterior Probabilities
1 class	-1117.58	2264.32	NA	1.000
2 class	-1064.14	2176.89	0.916	0.845-0.994
3 class	-1041.61	2151.25	0.871	0.858-1.000
4 class	-1022.14	2131.75	0.900	0.911-1.000
5 class	-1006.39	2119.70	0.915	0.914-1.000
6 class	-992.97	2112.29	0.926	0.914-1.000
7 class	-984.64	2115.07	0.925	0.925-1.000
8 class	-974.73	2114.69	0.931	0.908-1.000
9 class	-971.09	2126.86	0.917	0.826-1.000

LL=log-likelihood. BIC=Bayesian information criterion (smaller better). Entropy (close to 1 is better). Posterior probabilities (close to 1 is better). VLMR = Vuong-Lo-Mendell-Rubin likelihood ratio test. BLRT=Bootstrap likelihood ratio test (significant p-value but insignificant p-value for (k+1)-class model). The final model selected for interpretation is shown in bold.

Table 3. Trajectory Plots for Five Latent Classes.

Trajectory Class	n (%)	Estimate (Standard Error)
		Intercept	Linear	Quadratic
Class 1: Stable: Lowest Negative Screens	92 (71%)	3.86 (0.38)***	-1.63 (0.78)*	0.65 (0.38)+
Class 2: Stable: Medium/Low Negative Screens	24 (19%)	12.32 (1.24)***	3.65 (3.92)	-3.09 (1.88)
Class 3: Inconstant: Negative/Positive Screens	5 (4%)	7.23 (3.40)*	-9.12 (7.41)	8.62 (3.39)*
Class 4: Inconstant: Negative/High Positive Screens	3 (2%)	9.72 (3.50)***	41.09 (1.44)***	-16.81 (0.99)***
Class 5: Stable: High Positive Screens	5 (4%)	25.37 (3.10)***	-5.20 (8.70)	0.45 (4.07)

⁺p<0.10;

^*p<0.05;

^**p<0.01;

^***p<0.001

LCGA revealed five classes of parents who experienced varying levels of depressive symptoms (see Figure 1). Class 1, the “Stable: Lowest Negative Screens” group, accounted for most of the participants (n=92; 71% of sample). These parents had low levels of depression that remained low over time. Class 2, the “Stable: Medium/Low Negative Screens” group (n=24; 19% of sample), had stably moderate scores over time. Class 3, the “Inconstant: Negative/Positive Screens” group (n=5; 4%), had low levels of depression at T1 and T2 that abruptly increased to clinically significant levels of depressive symptoms at T3. Class 4, the “Inconstant: Negative/High Positive Screens” group (n=3; 2%), started at moderate scores that quickly increases over the clinically significant threshold at T2 but the score slightly decreased at T3. Class 5, the “Stable: High Positive Screens” group (n=5; 4% of sample), showed above threshold levels of depressive symptoms over all the time points. Overall, two classes were assessed as above the threshold for depressive symptoms at T2 (4-6 weeks post-placement) and three classes were above threshold at T3 (5-6 months post-placement). Demographic, parent and child characteristics were also examined by trajectory class (Appendix A, Supplementary Data). Significant differences were noted for gender, job status (full-time/part-time), and number of adopted children (overall and for mothers).

Figure 1. Five Classes of Parental Depressive Symptoms.

This figure shows the five classes of parents with distinct profiles of depressive symptoms (y axis) from 4-6 week pre-placement of the adopted child to immediate post-placement and 5-6 months post-placement on the x axis. The largest class of parents (Class 1, 71%) was stably low across time, and the next largest class (Class 2, 19%) below the clinically significant level (<16). Three classes of parents (Stable: High Positive Screens (4%), Inconstant: Negative/Positive Screens (4%), and Inconstant: Negative/High Positive Screens (2%) lines) were elevated in the clinical ranges on the depressive symptoms scale at the 5-6 months postplacement. The orange dashed line indicates the cut-point for depression (score of 16). CES-D = Center for Epidemiological Studies Depression Scale.

Mixed Effect Modeling of LCGA Classes

Table 4 shows means and standard deviations of all the measures and individual variables by trajectory class and time, and p-values from the mixed effects models. Overall risk for PAD, as measured by the PDPI-R, revealed a strong correlation to CES-D scores (r=0.602; p< 0.001) and significance by class, time, and class×time interactions. For each of these models, Class 1 “Stable: Lowest Negative Screens” differed from the other groups most frequently.

Table 4. Mixed Effects Modeling of Explanatory Variables as a Function of Time and Latent Class Membership.

		Trajectory Class, Mean (SD)						Mixed Effects Model, p-value
Variable	Possib le Range	Tim e	Class 1 (n=92): Stable: Lowest Negative Screens	Class 2 (n=24): Stable: Medium/L ow Negative Screens	Class 3 (n=5): Inconstant: Negative/Pos itive Screens	Class 4 (n=3): Inconstant: Negative/ High Positive Screens	Class 5 (n=5): Stable : High Positi ve Scree ns	Class	Time	Class*Ti me
Overall Risk for PAD
PDPI-R total	0-32	1	2.30 (1.97)	5.27 (3.31)	4.00 (1.73)	7.33 (4.04)	7.75 (4.86)	<0.0011vs2; 1vs4; 1vs5; 2vs4; 3vs4	<0.0011vs2; 1vs3	0.003T1: 1vs2; 1vs4; 1vs5 T2: 1vs2; 1vs4; 1vs5; 2vs4; 3vs4; 3vs5 T3: 1vs2; 1vs4; 1vs5; 2vs4; 3vs4
	0-32	2	3.44 (2.90)	7.40 (3.98)	4.75 (2.22)	14.00 (5.66)	11.50 (4.93)
	0-39	3	3.61 (2.97)	7.35 (3.94)	6.50 (2.38)	15.33 (0.58)	11.00 (7.12)
Interpersonal Variables
PSS-Fa: Family Support	0-20	1	15.23 (5.29)	14.00 (6.24)	11.40 (5.03)	7.67 (10.79)	12.25 (7.93)	0.011(1,2)vs(3,4,5)	0.095	0.226
		2	15.68 (5.27)	15.52 (5.55)	9.40 (6.50)	7.00 (9.64)	6.00 (3.56)
		3	15.26 (5.68)	14.67 (6.16)	11.40 (7.30)	8.67 (9.61)	9.25 (6.65)
PSS-Fr: Friend Support	0-20	1	15.52 (4.40)	14.32 (5.01)	14.00 (5.48)	9.00 (1.73)	13.75 (8.26)	0.0021vs4; 2vs4; 3vs4	0.457	0.220
		2	15.80 (3.82)	15.10 (5.19)	15.60 (2.07)	3.33 (1.15)	12.00 (8.83)
		3	14.85 (4.96)	14.33 (5.20)	14.40 (2.70)	5.00 (3.61)	13.75 (9.25)
Partner's Enthusiasm Being Parent	1-7	1	6.56 (0.69) β,γ	6.50 (0.67)	6.60 (0.55)	5.67 (1.53)	6.75 (0.50)	0.0061vs4; 2vs4; 3vs4; 4vs5	0.0011vs3; 2vs3	0.486
		2	6.56 (0.61)	6.38 (0.86)	6.80 (0.45)	5.33 (1.53)	6.75 (0.50)
		3	6.33 (0.77)	6.39 (0.61)	6.40 (0.89)	4.67 (1.53)	6.00 (0.82)
Partner's Enthusiasm Being Adoptive Parent	1-7	1	6.54 (0.71)	6.50 (0.67)	6.20 (0.84)	5.67 (1.53)	6.50 (0.58)	0.0171vs4; 2vs4	<0.0011vs3; 2vs3	0.053
		2	6.49 (0.69)	6.33 (0.97)	6.80 (0.45)	5.33 (1.53)	6.75 (0.50)
		3	6.31 (0.78)	6.33 (0.59)	5.60 (0.55)	4.67 (1.53)	6.00 (0.82)
IRQ: Love	10-90	1	85.04 (4.24)	80.40 (8.36)	78.60 (8.73)	71.33 (8.08)	80.75 (4.79)	<0.0011vs2; 1vs4; 2vs4; 3vs4; 4vs5	0.0051vs2; 1vs3	0.254
		2	84.03 (4.98)	78.86 (8.42)	80.00 (6.40)	64.00 (10.82)	78.00 (12.08)
		3	83.42 (5.22)	78.65 (12.06)	79.00 (9.80)	64.00 (8.89)	78.25 (6.85)
IRQ: Maintenance	5-45	1	36.17 (5.48)	34.91 (6.44)	33.40 (5.37)	27.33 (8.50)	31.50 (8.70)	<0.0011vs4; 2vs4; 3vs4; 4vs5	0.0091vs2; 1vs3	0.337
		2	34.54 (5.40)	32.95 (7.06)	33.20 (6.22)	19.00 (10.58)	33.25 (6.95)
		3	34.43 (6.33)	33.56 (6.33)	30.60 (10.06)	19.00 (3.46)	31.75 (6.18)
IRQ: Ambivalence	5-45	1	6.09 (2.07)	8.77 (6.36) β	10.00 (7.87)	6.00 (1.00)α,β,γ	10.00 (3.16) α,β	<0.0011vs2; 1vs3; 1vs5	<0.0011vs2; 1vs3; 2vs3	<0.001T1: NS T2: 1vs3 T3: 1vs2; 1vs4; 2vs4
		2	6.46 (2.69)	9.05 (5.86)	12.40 (15.45)	11.00 (1.00)	11.75 (10.31)
		3	6.80 (2.66)	10.17 (7.36)	13.40 (15.01)	16.33 (5.51)	17.50 (12.56)
IRQ: Conflict	5-45	1	13.14 (4.56)β	18.59 (6.84)	18.80 (6.34)	16.00 (3.61)α,β	15.75 (3.40)	<0.0011vs2; 1vs3; 1vs3	<0.0011vs2; 1vs3	<0.001T1: 1vs2 T2: 1vs2; 1vs4 T3: 2vs3; 2vs4; 3vs4
		2	14.04 (4.35)	18.90 (5.55)	20.60 (11.55)	28.00 (8.54)	17.50 (9.68)
		3	14.91 (5.41)	18.44 (7.24)	21.60 (11.93)	29.67 (6.43)	20.00 (5.72)
Sexual Relationship	1-7	1	5.53 (1.34) α	4.50 (1.82)	4.00 (1.83)	5.00 (1.00) α,β	4.50 (1.73)	0.0081vs(2,3,4,5)	0.0131vs2; 1vs3	0.052
		2	5.20 (1.41)	4.67 (1.80)	4.00 (1.87)	3.33 (1.15)	3.75 (1.71)
		3	5.35 (1.33)	4.44 (2.04)	4.20 (2.17)	3.00 (1.00)	4.00 (0.82)
Psychological Symptoms
IDAS: Ill Temper	5-25	1	6.38 (1.53)	8.38 (2.62)	7.00 (1.87)	7.67 (2.08)	11.00 (2.31) β	<0.0011vs2; 1vs4; 1vs5; 2vs4; 3vs4	<0.0011vs2; 2vs3	<0.001T1: 1vs2; 1vs5; 3vs5 T2: 1vs2; 1vs4; 2vs4; 3vs4;4vs5 T3: 1vs2; 1vs4; 2vs4; 4vs5
		2	6.49 (1.57)	8.86 (3.02)	8.40 (2.30)	15.33 (0.58)	8.75 (2.63)
		3	6.61 (2.04)	8.67 (2.33)	8.40 (2.51)	12.33 (0.58)	7.25 (1.71)
IDAS: Social Anxiety	5-25	1	6.31 (1.85)	6.76 (1.55)	5.00 (0.00)	7.33 (0.58)	8.25 (1.89)	<0.0011vs2; 1vs5	0.0331vs3	0.004T1: 3vs5 T2: 1vs2; 1vs5 T3: 1vs4; 1vs5; 2vs4
		2	6.00 (1.56)	7.76	6.20 (1.30)	7.33 (1.53)	8.75
		3	6.10 (1.63)	6.89 (1.71)	7.00 (1.83)	10.00 (3.61)	8.75 (3.10)
IDAS: Panic	8-40	1	8.58 (1.10)	9.90 (1.70)	9.20 (1.30)	10.67 (3.06)	10.00 (2.83)	<0.0011vs2; 1vs4	0.172	0.710
		2	8.26 (0.67)	9.52 (1.72)	9.40 (2.19)	10.00 (2.65)	8.50 (0.58)
		3	8.19 (0.60)	9.17 (1.50)	9.25 (1.50)	10.67 (2.89)	9.25 (1.26)
IDAS: Traumatic Intrusion	4-16	1	4.46 (1.12)	5.10 (1.26) γ	4.00 (0.00) β	6.00 (2.83)	7.75 (2.63)	<0.0011vs2; 1vs4; 1vs5; 2vs4; 2vs5; 3vs4; 3vs5	0.331	0.004T1: 1vs4; 1vs5; 2vs5; 3vs4; 3vs5 T2: 1vs2; 1vs4; 1vs5; 2vs4; 2vs5; 3vs4; 3vs5 T3: 1vs4; 1vs5; 2vs4; 2vs5
		2	4.20 (0.70)	5.55 (1.67)	4.20 (0.45)	8.00 (4.00)	8.50 (3.00)
		3	4.24 (0.68)	4.78 (1.17)	5.60 (1.82)	8.00 (3.61)	7.00 (3.46)
Feeling Rested	1-5	1	3.63 (0.88)	2.77 (1.07)	3.00 (1.83)	3.67 (1.53)	2.25 (1.26)	<0.0011vs2; 1vs5	<0.0011vs2; 1vs3	0.065
		2	3.27 (0.93)	2.33 (0.86)	2.60 (1.34)	1.67 (0.58)	1.50 (0.58)
		3	3.34 (0.80)	2.83 (1.20)	2.20 (1.30)	2.33 (1.15)	2.00 (0.82)
Alcohol Drinking	1-4	1	1.62 (0.87) α	1.59	1.50 (0.58)	1.67 (0.58)	2.00	0.937	0.81	0.008T1: NS T2: NS T3: NS
		2	1.54 (0.78)	1.86 (0.96)	1.40 (0.55)	1.67 (0.58)	2.00 (1.15)
		3	1.63 (0.80)	1.71 (0.85)	2.00 (1.15)	1.33 (0.58)	1.67 (1.15)
Life Orientation
LOT-R	0-24	1	17.29 (3.59)	14.71 (3.23)	16.20 (2.39)	11.33 (2.08)	12.00 (3.46)	<0.0011vs2; 1vs4; 1vs5	0.585	0.919
		2	17.41 (3.15)	15.05 (4.09)	14.60 (3.85)	10.67 (5.51)	11.67 (2.52)
		3	17.28 (3.26)	15.11 (3.66)	15.60 (2.61)	10.67 (4.73)	11.75 (2.22)
Life Satisfaction	1-10	1	8.63 (0.75)	7.82 (1.50)	8.80 (0.84)	7.00 (1.73) α,β	7.50 (1.91)	<0.0011vs2; 1vs4; 1vs5; 2vs4; 3vs4; 4vs5;	<0.0011vs2; 1vs3	0.007T1: 1vs2; 1vs4 T2: 1vs2; 1vs4; 2vs4; 3vs4; 4vs5 T3: 1vs4; 1vs5; 2vs4; 3vs4
		2	8.63 (0.76)	8.00 (1.05)	8.40 (0.89)	4.67 (2.89)	7.25 (1.50)
		3	8.53 (0.77)	7.94 (0.80)	7.80 (1.30)	5.33 (2.08)	7.00 (1.41)

LOT-R = Life Orientation Test-Revised. PSS-Fa = Perceived Social Support from Family. PSS-Fr = Perceived Social Support from Friends. IRQ = Intimate Relations Questionnaire. IDAS = Inventory of Depression and Anxiety Symptoms. PDPI-R = Postpartum Depression Predictors Inventory-Revised. Alcohol drinking was scored as: 1=no drink, 2=1-2 beverages, 3=3-4 beverages, and 4=more than 4 beverages per week. Higher score indicates a higher score in the direction of the way the variable is described in the table (e.g., higher scores on Life Satisfaction= more satisfaction). Mean and SD are unadjusted mean and SD. P-value was obtained from a mixed effects model. Tukey's post-hoc test was performed to adjust for multiple comparisons. Significant (p<0.05) class and time differences are noted below each p-value. Due to space, for significant interaction, the significant difference in time by each group is specified on the first cell of each group using this notation:

^α= time 1 vs. 2;

^β= time 1 vs. 3; and

^γ= time 2 vs. 3.

NS = no significant differences in any class (at the given time for interaction).

A significant main effect of class was found for all of the interpersonal variables (family and friend support, partner's enthusiasm of being a parent and being an adoptive parent, love, maintenance, ambivalence, conflict, sexual relationship⁶), psychological symptoms (ill temper, social anxiety, panic, traumatic intrusion, and feeling rested), and life orientation (optimism and life satisfaction). Alcohol intake was not significant for class, time, or the class×time interaction. Overall, the “Stable: High Positive Screens” and “Inconstant: Negative/High Positive Screens” groups reported scores that indicate interpersonal difficulties and psychological symptoms.

The parents in the “Stable: Lowest Negative Screens” group were at the most desirable levels for each explanatory variable, reporting low levels of psychological and interpersonal difficulties and high levels in positive domains (e.g., life satisfaction, optimism, support from family and friends). Compared to the “Stable: Lowest Negative Screens” group, the “Stable: High Positive Screens” group showed low levels of optimism and feeling of rest, and high levels of traumatic intrusion. Compared to the “Stable: Lowest Negative Screens” group, the “Inconstant: Negative/High Positive Screens” group showed low optimism, friend support, love, maintenance, and partner's enthusiasm of being parent and adoptive parent; however, high scores in conflict and panic. In other words, these parents appear to struggle with relationships (family, friend, intimate relationships) during the transition of a child into their home.

The means of the measures and individual variables of the “Inconstant: Negative/Positive Screens” group were not significantly different from “Stable: Lowest Negative Screens” group with the exception of ambivalence and conflict at T3. When compared to the “Inconstant: Negative/High Positive Screens” group, this group had significantly higher friend support, love, maintenance, and partner's enthusiasm of being a parent and an adoptive parent.

Significant time effects were found for the following variables: (a) compared to T1, love, maintenance, feeling of rest, and sexual relationship decreased at T2; ambivalence, conflict, ill temper, and life satisfaction increased from T1 to T2, (b) compared to T1, love, maintenance, feeling of rest, sexual relationship, and partner's enthusiasm for being a parent and an adoptive parent decreased at T3; ambivalence, conflict, social anxiety, and life satisfaction increased from T1 to T3; and (c) compared to T2, partner's enthusiasm for being a parent and an adoptive parent decreased at T3; ambivalence and ill temper increased from T2 to T3.

Significant interactions between class and time were found for two interpersonal variables (ambivalence and conflict), psychological symptoms (ill temper, social anxiety, and traumatic intrusion), and life satisfaction. The “Stable: Lowest Negative Screens” group reported lower mean scores on ambivalence, conflict, and ill temper than the other groups at T2 and T3. For social anxiety, the “Stable: High Positive Screens” group and “Inconstant: Negative/Positive Screens” groups were significantly different only at T1, but this difference was no longer significant at T2 and T3. The “Stable: Lowest Negative Screens” group had significantly lower social anxiety than the “Stable: High Positive Screens” and “Stable: Medium/Low Negative Screens” groups at T2, but significantly lower scores than the “Stable: High Positive Screens,” “Inconstant: Negative/High Positive Screens,” and “Inconstant: Negative/Positive Screens” groups at T3. At each time point, for traumatic intrusion, the “Stable: Lowest Positive Screens” and the “Stable: Medium/Low Negative Screens” groups (both under the threshold of symptoms) differed significantly from several other classes, including “Stable: High Positive Screens.” For life satisfaction, at T1 the “Stable: Lowest Negative Screens” group and the “Stable: Medium/Low Negative Screens” group were significantly different, but the means of the “Inconstant: Negative/High Positive Screens” group dropped down at T2 and T3, showing a significant difference from the other groups.

Discussion

In this study, we measured depressive symptoms using the CES-D in a sample of primarily heterosexual, adoptive parents' across three time points, from pre- to post-placement of a child. We hypothesized that the range in post-adoption depression (PAD) rates found in the literature may contain sub-groups of parents who are characterized by different trajectories of depressive symptoms. This hypothesis was supported when five distinct trajectories of depressive symptoms across time were revealed. Two classes were above the threshold of depressive symptoms at placement and three classes were above the threshold at post-placement. The majority of the parents, however, belonged to classes that were below the screening risk threshold for depression at all time points. We further examined the trajectory classes through a mixed effects model and found several variables that were significantly different across the depression groups, thus furthering our understanding of possible mechanisms that may explain the different course that depressive symptoms take in adoptive parents.

Modest PAD rates in this sample ranged between 9.5% pre-placement and peaked at 11.3% when placement occurred, with most parents remaining below the threshold. Some parents, however, struggled with depressive symptoms across all three time points, indicating a need for individualized assessments of depressive symptoms. The majority of parents belonged to Class 1: “Stable: Lowest Negative Screens,” with levels well below the threshold for depressive symptoms at all time points. Their risk for PAD as assessed by the PDPI-R was also the lowest of any group of parents, with indicators of strong interpersonal variables, stable psychological variables, and high levels of optimism and overall life satisfaction. It would appear that most adoptive parents weather the transition from pre-to post-placement without experiencing depressive symptoms. These findings are consistent with Mott et al. (2011) who found that adoptive mothers reported greater well-being and less psychopathology than postpartum mothers. In a previous study, we found that adoptive mothers reported significantly higher emotional stability than normative values of females (Foli et al., 2012). The parents in our study were also representative of adoptive parents in general (e.g., age, marital status, and educational and socioeconomic levels) and thus, possess potential buffers to depressive symptoms. However, it should be noted that these variables did not significantly differ between the depressive trajectory groups, indicating that these factors do not preclude the presence of depressive symptoms. An additional caveat is that disclosure of depressive symptoms prior to and after placement may be challenging as parent admission to depressive symptoms is often related to feelings of guilt and shame (Foli, 2010).

In the second class, “Stable: Medium/Low Negative Screens,” depressive symptoms were also below the threshold across all three time points, but not as low as Class 1. Placement of the child appears to decrease depressive symptoms, increasing the possibility that pre-adoption struggles are alleviated with the presence of the child arriving home. Although both Class 1 and 2 were below the screening threshold, they were significantly different on several variables, including overall PDPI-R scores.

In contrast to Class 1 and Class 2, parents in Class 3 “Inconstant: Negative/Positive Screens,” reported a sharp increase in symptoms by 6 months post-placement. For Class 3 parents, there was decreasing enthusiasm for parenting and being an adoptive parent; they became less satisfied with intimate relationship functioning, and they experienced more ill temper, social anxiety, and less feelings of rest. Their overall life satisfaction differed significantly by time compared to those parents who were below CES-D cut-off scores (Class 1). Fields et al. (2010) reported that while personal and family history of psychiatric difficulties were not correlated with depression, higher levels of stress than was expected and adjustment difficulties with the child were related to depressive symptoms for those who screened positive at any time point post-adoption. These findings emphasize the need for post-adoption support services that are based on a comprehensive assessment of the parents' and children's needs.

Time following adoption placement is also important to consider for those parents in Class 4 (Inconstant: Negative/High Positive Screens). Depressive symptoms increased from pre-placement to cross well above the screening threshold when the child was placed in the home. Class 4 is distinct in that these parents reported the highest level of depression immediately post-placement of any group. Anecdotal accounts of adoptive parents being startled when attachment and bonding (both parent to child and child to parent) are not immediate are common in accounts of PAD (Foli & Thompson, 2004) and should be further explored with other factors that occur around a child's placement in the home. Similar to Class 2, Class 4 parents' also reported fewer depressive symptoms by 6 months post-placement, but unlike Class 2 and similar to Class 5, they remained above the threshold of 16, indicating a need for a clinical evaluation. A decreased rate of depressive symptoms across time was noted by Senecky and colleagues (2008), who reported that rates fell after placement of a child. Fields et al. (2010) also found a similar pattern in a retrospective study in which depressive symptoms decreased post-placement from 27.9% at 0-4 weeks, 25.6% at 5-12 weeks, and 12.8% of mothers at 13-52 weeks.

The final group of parents, Class 5: Stable: High Positive Screens, belonged to a class that screened positive for depression prior to the child's arrival – the only class to do so – and continued to report symptoms above the threshold across time. Consistent with findings from Viana and Welsh (2010), pre-adoption variables of higher levels of depression, higher expectations of child development and behavioral/emotional problems (expecting difficulties), and a greater number of children in the family were significantly related to higher parenting stress at six months post-placement. Further research should focus on factors related to the adoption process and pre-placement variables to further elucidate the continuation of depressive symptoms across time.

Overall, our findings support Foli's (2010) theory of parental post-adoption depression as the trajectories of depressive symptoms changed across time and in response to placement of a child. Our findings also suggest that the dissonance between pre-adoption expectations and post-adoption reality may be modified or mitigated through buffers such as family and friend support and traits such as optimism. Findings also provide opportunities for continued study and theory refinement. For example, one of our parent classes (Class 5: Stable High Positive Screens) experienced depressive symptoms prior to placement, indicating that either expectations were violated prior to placement, such as happens in a failed adoption, or the parent's symptoms were unrelated to the adoption process (e.g., parents may have been struggling with depression related to infertility).

The limitations of the current study include the homogeneous sample as well as the modest number of parents who make up certain classes. A simulation study by Twisk and Hoekstra (2012) suggested caution in use of clustering statistical methods because they failed to accurately identify developmental patterns that had been artificially imposed on a real data set although latent class analysis and LCGA seem to be preferable among the clustering methods. Our LCGA results are preliminary and will need to be replicated in larger samples of adoptive parents in order to be supported. We did not use separate analyses to examine parents who had adopted multiple children (n=6 couples); future investigators may choose to address hypotheses related to multiple child adoptions as depression and stress may differ with these parents. While self-report data of depressive symptoms provide the foundation of our analysis, we believe that the online survey provided parents with anonymity and privacy, thereby affording them an opportunity to be candid in their reports. Further, parents responded to advertisements to participate in the study, and thus, may be non-representative of adoptive families. The CES-D is a screening tool, and the depressive symptoms reported were not confirmed through a clinical interview. Despite this, self-report data have been found to predict health outcomes (Idler, Leventhal, McLaughlin, & Leventhal, 2004).

Despite these limitations, the current study expands our understanding of parental post-adoption depression beyond that of rates at single points in time, as well as longitudinal studies that also examine parents as a single group (i.e., mothers only). Previous investigators describe a range of significant variables that are related to post-adoptive depressive symptoms; indeed, this study supports many of these findings and moves the scientific conversations forward. Our findings also call into question assumptions commonly associated with risk to parents, such as the age of the child being adopted, the presence of special needs, and the type of adoption (domestic private, domestic public, and inter-country), which did not differ between parental trajectories of depressive symptoms. These assumptions have not been able to consider the heterogeneity among adoptive parents as they transition from pre- to post-placement. The call is now to conduct research that will enable us to continue to understand the risks and buffers of unique classes of adoptive parents who may experience PAD, and to validate best practices to support these parents and their children.

Highlights.

• Analysis addressed post-adoption depressive symptoms from pre- to post-placement.

• The majority of adoptive parents were below the threshold for depressive symptoms.

• Latent class growth analysis demonstrated five distinct trajectory patterns.

• Additional variables varied by depressive symptom classes and time.