The Impact of Adverse Child and Adult Experiences on Recovery from Serious Mental Illness

Abstract

Objective

To compare effects of adverse childhood experiences and adverse adult experiences on recovery from serious mental illnesses.

Methods

As part of a mixed-methods study of recovery from serious mental illnesses, we interviewed and administered questionnaires to 177 members of a not-for-profit health plan over a two-year period. Participants had a diagnosis of bipolar disorder, affective psychosis, schizophrenia or schizoaffective disorder. Data for analyses came from standardized self-reported measures; outcomes included recovery, functioning, quality of life, and psychiatric symptoms. Adverse events in childhood and adulthood were evaluated as predictors.

Results

Child and adult exposures to adverse experiences were high, at 91% and 82% respectively. Cumulative lifetime exposure to adverse experiences (childhood plus adult experiences) was 94%. In linear regression analyses, adverse adult experiences were more important predictors of outcomes than adverse childhood experiences. Adult experiences were associated with lower recovery scores, quality of life, mental and physical functioning, social functioning, and greater psychiatric symptoms. Emotional neglect in adulthood was associated with lower recovery scores.

Conclusions and Implications for Practice

Early and repeated exposure to adverse events was common in this sample of people with serious mental illnesses. Adverse adult experiences were stronger predictors of worse functioning and lower recovery levels than were childhood experiences. Focusing clinical attention on adult experiences of adverse or traumatic events may result in greater benefit than focusing on childhood experiences alone.

BACKGROUND

Adverse experiences in childhood or adolescence predict a range of deleterious health behaviors and disease outcomes (Dong et al., 2004; Felitti et al., 1998). Half to two-thirds of U.S. adults report experiencing at least one adverse event in childhood (MMWR 2010; Dong et al., 2004; Felitti et al., 1998), and increased exposure to adverse childhood experiences are associated with worse mental health in adulthood (Edwards, Holden, Felitti, & Anda, 2003; Arnow, 2004), including depression (Lanoue, Graeber, de Hernandez, Warner, & Helitzer, 2012), and anxiety disorders (Raposo, Mackenzie, Henriksen, & Afifi, 2014).

People with serious mental illnesses experience even higher exposure rates (Goodman, Rosenberg, Mueser, & Drake, 1997; Mueser et al., 1998; Mauritz, Goossens, Draijer, & van Achterberg T., 2013). These studies, mostly drawn from the public mental health system, report a high prevalence of adverse experiences. Authors of a recent review found that about 90% of adults with serious mental illnesses have experienced one or more adverse events in childhood (Grubaugh, Zinzow, Paul, Egede, & Frueh, 2011), that those experiencing adverse events in childhood have worse physical and emotional functioning, and worse psychiatric symptoms as adults (Lu, Mueser, Rosenberg, & Jankowski, 2008; Rosenberg, Lu, Mueser, Jankowski, & Cournos, 2007). Moreover, adults with serious mental illnesses are more likely to experience adverse event exposures in adulthood (Coverdale & Turbott, 2000; Goodman et al., 2001).

Several studies have also demonstrated relationships between adverse childhood experiences and onset of mental health problems (Kessler et al., 2010; McLaughlin et al., 2010; Green et al., 2010; Varese et al., 2012). Population-based studies suggest relationships between adverse childhood experiences and broad categories of psychopathology, though not necessarily specific psychiatric disorders (Bulik, Prescott, & Kendler, 2001; Keyes et al., 2012). Rather, experiencing adverse or traumatic experiences in childhood may produce psychological vulnerabilities, including substance abuse, dissociation or heightened sensitivities that increase risk for psychoses and other serious mental illnesses (Janssen et al., 2004; Read, van, Morrison, & Ross, 2005; Pietrek, Elbert, Weierstall, Muller, & Rockstroh, 2013; Chen et al., 2010).

Despite advances in describing adverse experiences and their associations with psychiatric onset and functioning, less is known about relationships between adverse childhood experiences and recovery from serious mental illnesses. Recovery from mental illness is a complex process that includes adaptation to the diagnosis, developing agency and a sense of self-determination, community integration, as well as optimism about recovery and treatment options (Green, 2004; Liberman & Kopelowicz, 2005; Bromley et al., 2013; Leamy, Bird, Le, Williams, & Slade, 2011). Clinical approaches to trauma treatment or treatment of post-traumatic stress disorder (PTSD) (Frueh et al., 2009; Mueser et al., 2008) have been evaluated among people with serious mental illnesses, but we found no studies of relationships between adverse events and measures of recovery. As mental health researchers increasingly recognize recovery as a crucial patient-centered outcome for people with serious mental illnesses (Green, 2004; Anthony, 1993), an understanding of the relationships between adverse child and adult experiences and recovery is needed to inform approaches to treatment and improve outcomes.

In this paper we: 1) report the prevalence of adverse child experiences in an insured population of individuals with diagnoses of serious mental illnesses; 2) report the prevalence of adverse adult experiences using an adapted version of adverse childhood experience items (Dong et al., 2004); and 3) test the predictive capacity of adverse childhood experiences and adverse adult experiences on psychiatric symptoms, mental health functioning and recovery. We predicted that greater exposure to adverse childhood experiences would have a negative association with recovery, physical health, mental health, functioning, quality of life and psychiatric symptoms. Similarly, we predicted that greater exposure to adverse adult experiences would have negative associations with those outcomes. Finally, we hypothesized that adverse childhood experiences would retain an independent effect on outcomes, distinct from adverse adult experiences.

METHODS

Setting and Study Description

The STARS study recruited participants from Kaiser Permanente Northwest (KPNW), a not-for-profit integrated health plan serving about 480,000 members in the Portland, Oregon metropolitan area. KPNW provides comprehensive medical, mental health, and addiction treatment.

Conducted from 2003–2008, STARS was a mixed-methods longitudinal study examining factors associated with recovery among individuals with serious mental illnesses. At baseline, individuals completed two in-depth interviews and a paper-and-pencil questionnaire. Questionnaires were repeated at 12-months and 24-months post-enrollment and assessed a variety of topics including quality of life, psychiatric symptoms, and recovery. The 12-month questionnaire included questions about adverse childhood experiences. We assessed adverse adult experiences separately at 24-months. Questionnaire data were linked to electronic health plan data on diagnoses, service use, and clinician information. The study was approved and monitored by Kaiser Permanente's Institutional Review Board and Research Subjects Protection Office. Experienced doctoral- and master’s-level interviewers obtained written informed consent prior to participation.

Participant Identification, Inclusion and Exclusion Criteria, and Recruitment

We used diagnostic records and health-plan membership data to identify study participants who: 1) had diagnoses of schizophrenia, schizoaffective disorder, bipolar disorder, or affective psychosis; 2) had at least 12 months of health-plan membership prior to enrollment; and 3) were age 16 years or older. We excluded those with cognitive impairment sufficient to interfere with ability to consent, complete questionnaires, or participate in interviews. Participants were stratified on diagnosis and gender, and selected randomly within these groups for recruitment. We reached recruitment goals after attempting to recruit 418 individuals identified using health-plan records. Letters signed by primary mental health clinicians and the principal investigator were mailed with study information. We enrolled 46% of eligible persons we attempted to contact and stopped enrollment after exceeding goals. Additional recruitment details are described elsewhere (Green et al., 2013). Participants did not differ from the eligible population distributions by age or sex, within diagnostic group (Green et al., 2008; Hendryx, Green, & Perrin, 2008).

Measures

Our primary measure of recovery was the Recovery Assessment Scale (RAS) (Corrigan, Giffort, Rashid, Leary, & Okeke, 1999) and subscales: personal confidence and hope; willingness to ask for help; goal/success orientation; reliance on others; not being dominated by symptoms (Corrigan, Salzer, Ralph, Sangster, & Keck, 2004). To measure anxiety, depression, psychoticism, and dangerousness to self/others, we used the Colorado Symptoms Index (Boothroyd & Chen, 2008). From the Wisconsin Quality of Life Index we used the General Satisfaction Scale, the Social Relations/Support Scale, and one item on ability to work or attend school (Becker, Diamond, & Sainfort, 1993). We also used the SF-12 health inventory, to provide mental (MCS) and physical health (PCS) composite scores, and to provide a measure of social functioning (SF) (Ware, Kosinski, & Keller, 1995; Tunis, Croghan, Heilman, Johnstone, & Obenchain, 1999). Finally, we included interviewer-rated Global Assessment of Functioning (GAF) scores, rated at the 24-month interview (American Psyciatric Association, 1994). The GAF is a common measure of an individual’s overall level of psychological, occupational and social functioning.

Adverse Childhood Experiences

To reduce participant burden we modified the original 10-category adverse childhood experiences scale (Felitti et al., 1998) as follows: 1) collapsed four sexual abuse questions into one item (“Before you were age 18, did an adult or person at least 5 years older ever sexually abuse you?”; 2) consolidated four maternal battering items into one item assessing abuse of mother/stepmother; and 3) and dropped one item (“had to wear dirty clothes”) from five measuring physical neglect. Items were dichotomized according to the procedure used in the original study (Dong et al., 2004), and summed (ranging from 0 to 10). We adjusted scoring for scales with altered items.

Adverse Adult Experiences

We modified the childhood items to assess experiences after age 18. Two of the 10 categories (parental divorce, maternal battering) were excluded as their relevance to adults seemed less salient. We modified the remaining 8 categories by changing the reference period or source of exposure (e.g. “Since you reached the age of 18, how often has a spouse or romantic partner hit you so hard that you had marks or were injured?” Additional details provided upon request). These 8 categories were scored identically to the childhood items, accounting for modifications or consolidation of items as outlined above. The final score was summed (ranging from 0 to 8).

Data Analysis

We calculated descriptive and bivariate correlations between adverse child and adult items and outcome variables, then examined effects of childhood and adult adverse experiences on outcomes using hierarchical linear regression. We controlled for diagnosis (dichotomized as mood disorders versus schizophrenia spectrum disorders), age, sex, and educational status, by entering these first. We entered adverse child experiences in the second block, and adverse adult experiences in the third block. This allowed examination of independent effects of adverse childhood and adult experiences. We did not test mediational effects because we were testing the hypothesis that child and adult experiences independently contributed to psychiatric symptoms and recovery, rather than that the effects of earlier childhood experiences are mediated through later adult experiences. In addition, we tested the interaction effects of childhood and adult experiences in separate linear regression models for each outcome (results not reported; interaction effects not significant). In order to more fully understand independent effects of child and adult experiences, we tested the extent to which individual adverse experiences in childhood and/or adulthood contributed to the primary recovery outcome. Race was not included in any models, as the population was predominantly white and bivariate analyses showed no relationships with study outcomes. Statistical analyses were conducted using SPSS version 19 (IBM, 2010).

RESULTS

Participants

Table 1 shows baseline participant characteristics. Study participants were 92 female (52%) and 85 male (48%) members of Kaiser Permanente Northwest with diagnoses of schizophrenia, schizoaffective disorder, bipolar disorder or affective psychoses. Average age at baseline was 48.8 (SD = 14.8) years. The sample was 94% white, reflecting the KPNW's service area. At baseline, 8% had less than a high school education and 34% were college graduates. About 43% were employed, and 25% received disability insurance. More than half experienced “feeling different” from others (proxy for age of onset) prior to adulthood. At baseline, 41% of the sample experienced at least one additional (non-inclusion) mental health diagnosis, and 27% had received an substance use disorder diagnosis in the 12 months prior to enrollment. Additional demographic and descriptive information is available in previously published work (Green et al., 2008; Hendryx et al., 2008). Analyses included in this paper are based on the subset of 167 participants who completed the 24-month questionnaire. Response rate at 24-months was 96% among living participants. Adverse childhood experience items were assessed at 12-months; all other data were derived from 24-month questionnaires.

Table 1.

Baseline participant demographics and 24-month study outcome variables

Baseline measures	Overall N
Age, mean (SD)	177	49.1 (14.7)
Female, n (%)	177	92 (52.0)
Diagnosis group, n (%)	177
--Schizophrenia/schizoaffective disorder		75 (42.4)
--Depression with psychosis		18 (10.2)
--Bipolar disorder (I & II)		84 (47.5)
Race/ethnicity, n (%)	177
--White		169 (95.5)
--Non-white		8 (4.5)
Education, n (%)	173
--Less than high school		13 (7.5)
--High school graduate		33 (19.1)
--Some college/technical		68 (39.3)
--College graduate or higher		59 (34.1)
Age first felt different from others, n (%)	167
--Always felt different		10 (6.0)
--Grade school		42 (25.1)
--High school		34 (20.4)
--Adulthood		81 (48.5)
Prior year household income, n (%)	166
--<$19,999		49 (29.5)
--$20,000–$39,999		50 (30.1)
-->=$40,000		67 (40.4)
Disability income	173	44 (25.4)
At least one additional (non-inclusion) mental health diagnosis in year prior to baseline	173	71 (41.0)
At least one addiction diagnosis in year prior to baseline	173	46 (26.5)
Reported smoking some or every day	173	51 (29.5)
24-month outcome measures, mean (SD)
RASa	167	92.9 (14.0)
CSIb	167	15.8 (10.5)
WQLI-GSc	167	1.35 (1.1)
WQLI-SRc	155	1.56 (1.3)
WQLI-SWc	166	1.73 (1.9)
SF-12 PCSd	165	45.9 (13.0)
SF-12 MCSd	165	43.0 (12.3)
SF-12 SFd	166	43.1 (12.0)
GAFe	166	65.2 (15.4)

^aRecovery Assessment Scale. Possible scores range from 24 to 120; higher scores measure better recovery outcomes.

^bThe Modified Colorado Symptoms Inventory. Possible scores range from 0 to 56; higher scores indicate more psychiatric symptoms.

^cWisconsin Quality of Life Index; GS=General Satisfaction, SR=Social Relations/Support, SW=School/Work. Possible subscale scores range from −3.0 to 3.0; higher scores measure higher quality of life.

^dThe SF-12 physical component summary (PCS), mental component summary (MCS) and social functioning (SF) can have possible score ranging from 0 to 100 and are normed to achieve means of 50 and standard deviations of 10 in the general U.S. population.

^eGlobal Assessment of Functioning scores can range from 1 to 100.

Prevalence of Adverse Experiences

Table 2 compares prevalence of adverse childhood experiences and matched adverse adult experiences in the original Dong et al. (2004) study and our study. In addition to calculating separate exposure rates for childhood and adulthood, we calculated prevalence estimates for lifetime exposures and for repeated experiences in both childhood and adulthood for each event category.

Table 2.

Prevalence of adverse child and adverse adult experiences in STARS compared with general population estimates from the original adverse childhood experiences study

	(Dong et al., 2004) Adverse Childhood Experiences	STARS <18 y.o. Child (N=170)	STARS >18 y.o. Adult (N=167)	STARS Ever (N=171)	STARS Both (N=167)
Household mental illness	20.3%	52.9%	50.9%	69.6%	33.5%
Physical Abuse	26.4	46.5	27.5	56.7	16.8
Household substance abuse	28.2	45.9	58.7	73.1	30.5
Physical neglect	26.4	35.9	29.9	49.7	15.6
Emotional abuse	10.2	35.3	29.9	50.9	13.8
Parental separation/divorce	13.0	34.1	--	--	--
Emotional Neglect	14.8	32.4	16.2	36.3	12.0
Sexual abuse	21.0	24.7	31.1	45.0	10.2
Maternal domestic violence in house	24.1	22.9	--	--	--
Criminal household member	6.0	12.9	25.7	32.7	5.4
Percent exposed to at least one adverse event	67.3%	91.2%	82.0%	93.6%	63.5%

Child= Adverse Childhood Experience; Adult=Adverse Adult Experience; Ever=Experience of each adverse event category in either childhood or adulthood; Both=experience of each adverse event category in both childhood and adulthood.

As expected for individuals with serious mental illnesses, study participants reported higher exposure rates in almost all categories than found in general populations (Dong et al., 2004; Felitti et al., 1998). Only in maternal domestic violence and sexual abuse categories were percentages of childhood experiences similar to results in the general population. More than 91% of participants reported exposure to at least one adverse childhood experience, while 82% reported adverse adult experiences. Nearly 94% reported either an adverse experience in childhood or adulthood. More than six in 10 reported experiencing repeated adverse experience in adulthood also experienced in childhood. One-third of participants reported experiencing household mental illness and household substance abuse during childhood and adulthood.

Many participants reported two or more adverse experiences in childhood and adulthood (Table 3); 76.5% and 62.8% respectively. About one-third (34.8%) experienced multiple forms of adverse events that began in childhood and were repeated in adulthood. In addition, 11% reported four or more categories of adverse events occurring in childhood and again in adulthood.

Table 3.

Total number of participant-reported adverse experiences in childhood, adulthood and combined

Number of Categories	Adverse childhood experiences (N=170)	Adverse adult experiences (N=167)	Combined childhood and adult experiences (N=167)
0	8.8%	18.0%	36.5%
1	14.7	19.2	28.7
2	20.6	11.4	13.8
3	17.1	16.2	9.6
4	7.1	13.8	6.6
5	11.2	9.6	3.0
6	5.9	7.8	1.2
7	5.9	3.6	0.6
8	4.7	0.6	0.0
9	2.9	--	--
10	1.2	--	--
Mean(SD)	3.44 (2.5)	2.70 (2.1)	1.38 (1.5)

Notes: Only 8 adverse categories were measured in adulthood. Combined child and adult experiences required valid responses at both time points and measures the experience of adversity in one or more domains (e.g. household mental illness) in both childhood and adulthood.

Relationships between Adverse Events, Recovery, Psychiatric Symptoms and Functioning

Table 4 shows hierarchical linear regression results examining relationships between adverse childhood and adult experiences and study outcomes, adjusting for diagnostic group, age, sex, and education. Recovery was negatively associated with greater adverse adult experiences (β =−1.1; p=0.05); while childhood events did not predict recovery. Mental health symptoms (Colorado Symptoms Index) were greater among people who had experienced adverse childhood events (β =0.61; p=.05). When adult adverse experiences were entered into the model, however, childhood events no longer predicted symptoms, but adult experiences did (β =1.6; p<0.001). Childhood experiences did not predict general satisfaction with quality of life, social relationships/support, global assessment of functioning, or physical health, though adverse adult events did (WQLI-General Life Satisfaction, β =−0.15; p<0.001; WQLI-Social Relations/Support, β =−0.2; p<0.001; GAF β =−1.9; p<0.001;SF-12 Physical Component Summary β =−138; p=0.01). Child adverse events did not predict ability to work or attend school, but adult experiences did (β =−0.16; p<0.05). Childhood events, negatively associated with mental health MCS scores (β =−0.83; p=0.05) and SF-12 social functioning (β =−0.80; p=0.05) in the first model, were no longer significant after inclusion of adult adverse experiences. Adult experiences, however, predicted both lower SF-12 MCS scores (β =−1.32; p=0.01) and lower SF-12 social functioning scores (β =−1.67; p<0.001), including when childhood experiences were included in the model. We assessed the difference between diagnostic groups for each study outcome but they were not predictive with the exception of the GAF (β =−8.44; p<0.001), as expected.

Table 4.

Hierarchical linear regression results for study outcome measures¹

	RAS				CSI				SF-12 SF				SF-12 PCS				SF-12 MCS
	Step 2		Step 3		Step 2		Step 3		Step 2		Step 3		Step 2		Step 3		Step 2		Step 3
	β	SE	β	SE	β	SE	β	SE	β	SE	β	SE	β	SE	β	SE	β	SE	β	SE
Child	−0.81	0.45	−0.56	0.46	0.61*	0.31	0.24	0.31	−0.8*	0.38	−0.42	0.38	−0.30	0.40	0.02	0.40	−0.83*	0.37	−0.53	0.37
Adult	--	--	−1.1*	0.55	--	--	1.6***	0.37	--	--	−1.67***	0.45	--	--	−1.38**	0.48	--	--	−1.32**	0.44
Model R2	0.03		0.05*		0.15*		0.24***		0.08*		0.15***		0.13		0.18**		0.08*		0.15***
F(df)	3.3 (1,159)		3.8 (1,158)		3.8 (1,159)		19.1 (1,158)		4.5 (1,159)		13.9 (1,158)		0.6 (1,158)		8.3 (1,157)		4.5 (1,159)		13.9 (1,158)

	WQLI-SR				WQLI-SW				WQLI-GS				GAF
	Step 2		Step 3		Step 2		Step 3		Step 2		Step 3		Step 2		Step 3
	β	SE	β	SE	β	SE	β	SE	β	SE	β	SE	β	SE	β	SE
Child	0.02	0.04	0.07	0.04	−0.03	0.06	0.01	0.06	−0.02	0.03	0.01	0.03	−0.24	0.46	0.19	0.46
Adult	--	--	−0.2***	0.05	__	__	−0.16*	0.07	__	__	−0.15***	0.04	--	--	−1.9***	0.54
Model R2	0.06		0.17***		0.04		0.07*		0.09		0.17***		0.13		0.18***
F(df)	0.27 (1,149)		18.95 (1,148)		0.27 (1,159)		5.1 (1,158)		0.4 (1,159)		14.9 (1,158)		0.28 (1,159)		11.76 (1,158)

^*p=0.05;

^**p=0.01;

^***p<0.001

¹Diagnosis, age, sex and education level were entered in step 1 as controls. Results are not reported here but are available from the authors upon request.

Child=adverse childhood experiences. Adult=adverse adult experiences. RAS=Recovery Assessment Scale. CSI= Colorado Symptoms Inventory. SF-12-SF= SF-12 social functioning summary SF-12 PCS= SF-12 physical component summary. SF-12 MCS= SF-12 mental component summary.

Child=adverse childhood experiences. Adult=adverse adult experiences. WQLI-SR=Wisconsin Quality of Life Index Social Relations/Support Scale. WQLI-SW=Wisconsin Quality of Life Index School/Work. WQLI-GS=Wisconsin Quality of Life Index General Satisfaction. GAF= Global Assessment of Functioning.

Finally, regression analyses for individual adverse experiences showed that five of the eight experiences measured did not contribute to the findings on recovery. We therefore tested a final model to see which of three adverse experiences (emotional neglect, physical neglect, and household substance use) were the best predictors of recovery. Physical neglect was not significant in either childhood or adulthood. Childhood exposure to substance use was significant before adult exposure was added, but not after. When entered together, emotional neglect in adulthood was the only adverse experience to be significant in the final model (β =−10.1; p=0.002).

DISCUSSION

The high frequency of adverse childhood experiences we found is consistent with results from other studies of individuals with serious mental illnesses (Grubaugh et al., 2011). Moreover, adverse adult experiences were similarly high, with lifetime exposure reaching almost 94% of the sample. Our finding of high cumulative exposure is also consistent with other research in this population (Goodman et al., 1997; Goodman et al., 2001; Neria, Bromet, Sievers, Lavelle, & Fochtmann, 2002), suggesting significant adversity, high levels of potential trauma exposure, and accumulated stress (Grubaugh et al., 2011; Lu et al., 2013). Experiencing serious mental illness or substance use in the household—in childhood, adulthood, or both—was twice as common as the prevalence of other adverse experiences. This was higher than reported in previous studies (Rosenberg et al., 2007) and suggests that study participants experienced significant adversity in childhood which continued into adulthood. This finding, in an insured population with generally better functioning and higher incomes, is striking given differences between this population and those reported elsewhere (Cusack, Grubaugh, Knapp, & Frueh, 2006; Padgett, Henwood, Abrams, & Drake, 2008). The most notable exception is the experience of childhood sexual abuse: our study population reported lower rates than previously reported in uninsured populations with mental illnesses. Thus our sample more closely resembles the general population (Coverdale & Turbott, 2000; Rosenberg et al., 2007).

Existing literature on the relationships between adverse childhood experiences and adult health led us to hypothesize that childhood experiences would be at least as salient as adult experiences in predicting psychiatric symptoms, mental health functioning and recovery. Childhood experiences did predict worse mental health and social functioning, yet effects did not hold when adult experiences were entered into models. Adverse adult experiences predicted all study outcomes. These results differ from previous studies in which childhood events were associated with lower mental health functioning (Rosenberg et al., 2007; Wu, Schairer, Dellor, & Grella, 2010). Previous studies, however, did not assess independent effects of childhood and adult experiences using the same repeated measure to assess both experiences.

Our results are consistent with a large household survey that found a relatively weak association between adverse childhood experiences and persistence of psychiatric disorders (McLaughlin et al., 2010). Similarly, adverse childhood experiences do not correlate with specific diagnoses, but rather with broad predisposing symptoms (Janssen et al., 2004; Keyes et al., 2012), which suggests that early experiences contribute less to adult symptomology and, therefore, figure as less important in adult recovery than adverse experiences in adulthood. Our results complement these findings by suggesting that childhood experiences may be only weakly associated with recovery and functioning, particularly when studied alongside adverse adult experiences. In contrast, our finding that 17–24% of the variance in functioning, psychiatric symptoms and recovery is a function of exposure to adult adverse experiences suggests that recovery-supportive treatment should focus on addressing trauma-related symptoms not just for the sake of symptom reduction but with the goal of minimizing risk of future exposures. Interestingly, we found that emotional neglect in adulthood (not feeling loved and supported in one’s family) was negatively associated with recovery, thus highlighting the underlying importance of current relationships in the process of recovery, above and beyond other adverse experiences. Finally, we assessed the difference between diagnostic groups for each study outcome but they were not predictive with the exception of the GAF. On this measure, as expected, people with schizophrenia spectrum disorders had worse recovery outcomes than those with mood disorders. Thus, our results suggest that diagnosis itself does not play a critical role in how childhood or adult experiences of adverse events relate to recovery.

Limitations

Our findings should be interpreted with some caution. First, there are differences between the current study population and those included in prior studies. Our sample was more likely to be employed, had a higher educational level, and reported higher incomes than other studies, thus reflecting a slightly more “advantaged” group than typical populations with serious mental illnesses (Cusack et al., 2006; Padgett et al., 2008). However, study participants clearly experienced similarly high rates of life course adversity as those described elsewhere (Padgett, Smith, Henwood, & Tiderington, 2012) and reported experiencing both childhood and adult adverse experiences at the same rates as those in public mental health settings. It should be noted that participants reported that they had significant experience with community mental health settings during periods of being uninsured, so it is possible that the populations may not be as different as they appear though, as noted above, the experience of childhood sexual abuse was lower than typically reported in uninsured individuals receiving care in community clinics, possibly influencing results. Second, the experience of adversity, substance use and mental illness are intertwined, making the magnitude and direction of childhood and adulthood effects difficult to assess. Unmeasured factors and timing/dose of adverse exposures are important to understanding the sequence and severity of experiences, and their effects on recovery—something we were unable to examine in this study. Third, our observational data do not allow us to say more about the longitudinal interplay between current substance use, early trauma, adult experiences and the severity of illness or recovery, or how adult onset of poverty and other difficulties affect recovery. Fourth, recall bias of more recent events compared with childhood recall cannot be ruled out as affecting findings. This study also is limited by an ethnically homogeneous sample.

Additionally, in condensing sexual abuse and maternal battering measures down to one item each, these results may not be directly comparable to rates reported in previous studies, particularly for childhood sexual abuse, and may help explain their lack of predictive power in our findings on recovery. Future studies might pay particular attention to the role of sexual abuse, and all independent contributors, to recovery. Other important measures of trauma and/or recovery, such as post-traumatic stress disorder symptoms and mental health service use, were not included in our analyses.

CONCLUSIONS AND IMPLICATIONS FOR PRACTICE

Exposure to adverse childhood and adult experiences in persons with serious mental illnesses is high. Our findings suggest, however that adult experiences may be more important predictors of mental health, physical health, quality of life, social functioning and recovery than childhood experiences. Our finding that a large proportion of the variance in mental health symptoms and functioning, regardless of psychiatric diagnosis, can be explained by adult exposure to adverse events suggests that attention to such events in clinical settings is important. Particularly important is the role that emotional neglect—not feeling special, loved, close to family, or that family is a source of support and strength—plays in recovery. This is a modifiable intervention target; individuals without family support can be assisted in creating opportunities for connection and support through non-familial relationships. Community-based efforts, such as peer-to-peer social support programs (Davidson, Bellamy, Guy, & Miller, 2012; Daniels, Grant, Filson, Powell, & Fricks, 2010), can help to ameliorate neglect and isolation, and enhance recovery. In addition, trauma-informed care practices (Substance Abuse and Mental Health Services Agency, 2012) may be critical adjuncts to recovery-oriented services (Bloom, 1997), particularly because they help individuals resist re-traumatization. The pathways of how these adverse experiences affect individuals, biologically and psychologically, and therefore how they affect recovery processes, will be important directions for future research (Anda et al., 2006; Nemeroff & Binder, 2014).