Psychological and Physical Health of Nonoffending Parents After Disclosure of Sexual Abuse of Their Child

Disclosure of child sexual abuse can be traumatic for nonoffending parents. Research has shown its impact on mothers' mental health, which includes heightened psychological distress, depression, and post-traumatic stress disorder. Very little is known, however, about its impact on their physical health or on fathers' health. The self-perceived mental and physical health of nonoffending parents after child sexual abuse disclosure was compared to determine gender-related differences in this regard. Interviews were conducted with 109 mothers and 43 fathers of 6- to 13-year-old sexually abused children. Bivariate analyses revealed that a fair proportion of parents reported psychological and physical problems after disclosure. However, proportionally more mothers than fathers reported psychological distress, depression, and use of professional services. Fathers were more likely to resort to health services instead of social services and to use medication for depression. Study findings provide leads for health and social service providers for the development of intervention protocols and referral procedures sensitive to gender issues, and they shed new light on specific needs of nonoffending parents.

Keywords: child sexual abuse; disclosure; health services; mental health; nonoffending parents; physical health

Child sexual abuse (CSA) has a negative impact not only on the child victims but also on their parents, considered secondary victims by some researchers and clinicians (Deblinger, Hathaway, Lippmann, & Steer, [15]; Manion et al., [37]). Indeed, learning that one's child has been sexually victimized is generally an unexpected, confusing, and traumatic event involving deep emotional loss (Elliott & Carnes, [20]). Disclosure is only the first in a series of stressful events that parents are likely to face when a child is sexually abused. It can be followed by loss of parental custody, separation or divorce, change in residence, and losses of or changes in relationships with family and friends (Massat & Lundy, [38]). Other stressors can include police questioning, court proceedings, and media attention (Dyb, Holen, Steinberg, Rodriguez, & Pynoos, [18]). All of these events have the potential to impact the mental and physical health of nonoffending parents after disclosure.

Focus on nonoffending parents after children's disclosure is important for at least two main reasons. First, health and social service providers expect nonoffending parents to help their children resolve the turmoil created by CSA because they are often involved in child treatment (Cohen, Deblinger, Mannarino, & Steer, [7]). In this regard, it is important to gain a better understanding of how they react physically and emotionally to ensure that their capacity to be available and help their children is not hindered. Second, if mothers and fathers report symptoms and problems of their own, they will need help to cope with this dramatic life crisis.

CSA disclosure is a major life event, and whether mothers and fathers are affected similarly by it is an issue worthy of investigation. In a meta-analytic review of 119 studies of gender differences in facing life events, Davis, Matthews, and Twamley ([14]) found that, overall, women subjectively reported being exposed to more stressors, symptoms of depression, anxiety, and psychosomatic problems compared to men. Men are more likely to report exposure to stressors and to be distressed by work-related or financial events, whereas women report more exposure to stressors and greater distress in the family and social spheres (Conger, Lorenz, Elder, Simons, & Ge, [8]). Although the differential exposure hypothesis that suggests that women face objectively more stressors than men could not be discarded, these differences could also be explained by the effect of gender role socialization, whereby women are expected to be more interdependent and attuned to the feelings of others (Denton, Prus, & Walters, [16]). If this differential vulnerability hypothesis holds in the case of CSA disclosure, then mothers could be expected to react more intensely to the event and to suffer greater stress than fathers do. Mothers' psychological reaction to CSA disclosure has been identified in past studies. Fifty percent to 58% of mothers reported clinical symptoms of general distress (Cyr, McDuff, & Wright, [11]; Hébert, Daigneault, Collin-Vezina, & Cyr, [27]; Newberger, Gremy, Waternaux, & Newberger, [39]; Runyan et al., [42]), 32% to 53.8% reported post-traumatic stress disorder (PTSD symptoms (Cyr et al., [11]; Davies, [13]; Kelley, [31]), and 22% to 46% showed clinical symptoms of depression (Kim, Noll, Putnam, & Trickett, [32]; Santa-Sosa, Steer, Deblinger, & Runyon, [43]). These studies included mixed samples of intra- and extrafamilial CSA, with the exception of Kim and colleagues' and Runyan and colleagues' studies based on intrafamilial cases and Kelley and Davies's studies based on extrafamilial. Time elapsed between disclosure and mothers' evaluation of symptoms varied from 9 weeks to 6 months for the half of the studies that reported it.

Only three studies (Davies, [13]; Kelley, [31]; Manion et al., [37]) have compared paternal and maternal levels of symptoms following disclosure. In two studies, more mothers displayed psychological distress or depression symptoms than fathers, while Kelley's study observed the opposite two years after the disclosure. For PTSD symptoms, more mothers reported intrusive symptoms than fathers in the three studies. Only Manion and colleagues observed more symptoms of avoidance for women in contrast with men, 13 weeks after disclosure. However, these studies suffer from a number of limitations, including small sample sizes (e.g., n = 17 parents in Davies, and no time frame) and samples limited to CSA involving extrafamilial perpetrators (e.g., day care workers, neighbors, babysitters).

In addition to mental health, traumatic life events such as CSA disclosure can also have a major impact on parents' physical health. Only two studies have addressed this issue. In a phenomenological qualitative study by Lafleur ([34]), all three participating mothers reported physical health problems after disclosure. These were generally short lived and included headaches/migraines, gastrointestinal problems, skin conditions, weakened immune system, and exacerbation of predisclosure health problems. Newberger and colleagues ([39]) found that the somatic symptoms of 42 mothers dropped within the normal range 12 months after disclosure, suggesting that physical symptoms occurred as an acute effect of the initial CSA disclosure trauma. To our knowledge, no study has ever assessed the physical health of nonoffending fathers.

To gain a clearer understanding of parental health following CSA disclosure, it is important to control for the potential effect of other factors. Research on health has identified many factors influencing quality of health in the biological, social, behavioral, and psychosocial spheres (Denton et al., [16]). Four factors were considered crucial in the present study: history of trauma, sociodemographic variables, life event stressors, and characteristics of the abuse.

The variability in parental reactions following disclosure could be explained in part by parents' own trauma history. A number of studies have found that mothers with a history of CSA experienced significantly more general distress (Deblinger et al., [15]; Hébert et al., [27]; Kelley, [31]) and PTSD symptoms (Cyr et al., [11]) than mothers without such a history. However, Kelley did not find this relationship for fathers, nor did Deblinger and colleagues ([15]) for mothers. The Adverse Childhood Experiences (ACE) Study (Edwards, Holen, Felitti, & Anda, [19]) demonstrated that CSA significantly increased risk for depression for both men and women in adulthood. It also found that different forms of child maltreatment were likely to co-occur and that a strong and graded relationship existed between the number of types of child maltreatment and prevalence and risk of mental health and multiple somatic symptoms. Furthermore, many studies have shown social inequality structures, such as income, education, and occupational status, to be related to health status (Denton et al., [16]). Newberger and colleagues ([39]) found that poor mothers were more likely to report higher levels of psychological distress. Crisis theory and research (Roberts, [41]) also showed that the number of stressful life events has an impact on health. Cyr and colleagues ([11]) observed a relationship between mothers' report of psychological distress and the number of stressful events four months after disclosure.

Few studies have examined the association between the characteristics of the abuse and parents' health. For example, Cyr and colleagues ([11]) and Manion and colleagues ([37]) found no significant correlations among the number of abused children in the family, type and frequency of sexual abuse, relationship of the perpetrator to the family, and the child's age and gender to predict maternal emotional functioning. Newberger and colleagues ([39]) found a small relation between the severity of CSA (r = 0.34) and mothers' distress. Hébert and colleagues ([27]) observed that an intrafamilial perpetrator was associated with a higher score for psychological distress (OR = 2.74) while severity and length of the abuse were not. Faced with these inconsistent results, it seemed warranted to verify the potential impact of these variables in the present study.

The aim of the present study was to assess the mental and physical health of nonoffending parents following disclosure of sexual abuse of one of their children and to determine whether gender differences among parents existed. A second objective was to identify whether other variables (related to CSA, sociodemographic characteristics, parents' maltreatment history, and life and disclosure event stressors) could predict parents' mental and physical health.

The study was designed to improve the state of knowledge in the field by considering both mental and physical health indicators. It was intended to shed new light on a topic that remains understudied, namely the differential reactions of mothers and fathers following CSA disclosure. It also relied on a sample that included cases of both intra- and extrafamilial CSA.

Participants were solicited to take part in our study at four sites offering services to sexually abused children and their parents, namely two child protection agencies, one child advocacy center, and one nonprofit organization. Only parents who were not the perpetrators of the CSA as indicated in the police investigation were included in the study. The sample was composed of 105 mothers, 4 stepmothers, 36 fathers, and 7 stepfathers. Stepmothers and stepfathers were included only in the absence of parental involvement by the biological mother or father and if they had lived with the child for at least 2 years. In 86 cases, only 1 parent participated, while 2 did in 33 other cases. The final sample was made up of 152 participants who were the parental figures to 119 children. The mean age was 37.4 years (SD = 6.3) for mothers and stepmothers and 41.7 years (SD = 7.8) for fathers and stepfathers. Most participants were French Canadian, and other ethnic groups accounted for 17.6% of the sample. More than half of the women (56%) and nearly half of the men (48.8%) had completed at least a high school degree. Among the women, 45% worked part time or full time, compared with 58.1% of the men. For most households (70.3%), annual income was below CAN$39,999, with only 11.9% reporting CAN$70,000 or more. One-fifth of the children (21%) lived in a nuclear family setting, half (50.4%) lived with only one parent, and just under one-third (28.6%) belonged to blended family units (with a stepmother or stepfather and their children, if any).

The CSA victims were 6 to 13 years old (M = 9.4, SD = 2.0); 72.3% were girls. For 22.9% of the victims, the sexual abuse involved a single episode; for 40.7%, it occurred a few times, and for 36.4%, the incidents occurred many times. The children were 6.7 years old on average (SD = 2.3) at the time of the abuse. The offenders ranged from 8 to 70 years old, and nearly half (48.9%) were under 18 years of age. They were a parent or stepparent in 22% of the cases, a sibling or stepsibling in 23%, and a member of the extended family in 21%. Most of the sexual abuse episodes (62.7%) involved physical contact with penetration, attempted penetration, oral sex, anal sex, or atypical sexual activities (e.g., multiple aggressors).

All the respondents gave their full consent before taking part in the study, which was approved by the ethics board of both Université de Montréal and Centre jeunesse–Institut Universitaire de Québec. At each site, clinicians recruited mothers and fathers only once the abuse was confirmed following a police investigation. When both parents participated, they were met with separately by trained interviewers who administered the questionnaires. The interval between the CSA disclosure and parent evaluation ranged from 0 to 47 months and averaged 12.5 months (SD = 9.3), although 6 outliers for whom the time exceeded 4 years were excluded from the analyses.

All the measures were administered in their French version. The first reference concerns the authors of the original measure and the second the authors of the French translation.

Psychological health was assessed by way of three questionnaires.

The PSI (Ilfeld, [29]; Kovess, Murphy, Tousignant, & Fournier, [33]) was used to evaluate the amount of symptom recurrence pertaining to mental health disturbance. Using a 4-point scale (0 = not at all to 3 = very often), anxiety, depression, anger, and cognitive disturbance were evaluated in the week prior to the evaluation. The most frequently recommended clinical cutoff score is 30 and above (Kovess et al., [33]). The French version was validated in a vast Quebec epidemiological health survey in 1987 (Préville, Boyer, Potvin, Perreault, & Légaré, [40]). The internal consistency of the PSI has been rated at 0.92. The total score was used in our analyses.

The MPSS-SR (Falsetti, Resnick, Resick, & Kilpatrick, [21]; Stephenson, Marchand, Marchand, & Di Blasio, [46]) was administered second and used to evaluate two-week severity and frequency of 17 symptoms that fit the 4th edition of the Diagnostic and Statistical Manual for Mental Disorders (DSM-IV) B, C, and D criteria (American Psychiatric Association, [1]). Severity was assessed using a 5-point scale (A = not at all distressing to E = extremely distressing), and frequency was assessed using a 4-point scale (0 = not at all to 3 = 5 times a week, very much, almost always). The MPSS-SR yields a total score from 0 to 119. The internal consistency for the frequency and severity was very high for both the English (0.97 and 0.94, respectively) and French (0.92 and 0.95, respectively) versions (Coffey, Dansky, Falsetti, Saladin, & Brady, [6]; Guay, Marchand, Iucci, & Martin, [25]). The total score was used, and the clinical cutoff score for a community sample was 46.

The SCID (First, Spitzer, Gibbon, & Williams, [22]; Lapalme & Hodgins, [35]) was administered third to the participants by licensed clinicians (psychologist, social worker, sexologist) trained by an expert for 30 hours to document past-month episodes of depression and PTSD disorders. Within each diagnosis, a series of closed- and open-ended questions was used to check for the presence of the diagnostic criterion. Test-retest reliability varied from 0.63 to 0.88, and a kappa = 0.80 for depression and kappa = 0.88 for PTSD were noted (Zanarini et al., [49]). Presence of diagnosis was used in the analyses.

The ISQ was used to evaluate physical health. The questions were developed by an expert committee based on questions used by the ISQ in two earlier studies (Daveluy et al., [12]), which helped explore different aspects of health status three months prior to disclosure and postdisclosure: self-perceived general health status rated on a 5-point scale (1 = very good to 5 = very poor); limitations caused by health conditions at home, work, leisure activities, and caring for children (1 = yes, 0 = no); use of health and social services (e.g., general practitioner, psychologist) (1 = yes, 0 = no); and use of medication one week prior to disclosure and in the past week (1 = yes, 2 = no).

The ETI-SR SF (Bremner, Bolus, & Mayer, [3]; Cyr, Hébert, & Zuk, [9]) is a standardized instrument designed to measure the past history of child maltreatment (physical, emotional, and sexual abuse). Three items were added to assess whether the participants had experienced childhood neglect, for a total of 65 yes or no items. Cronbach's alphas varied from 0.70 to 0.87 across the 4 scales. The questionnaire distinguished patients presenting elements known to be associated with trauma from comparison groups. The sum of maltreatment types experienced by the participants was used.

The Life Event Sources of Stress Questionnaire (Hébert & Parent, [28]) is an instrument that comes from the Source of Stress Inventory (Chandler, [5]) and the Life Events Checklist (Johnson & McCutcheon, [30]). Its 15 items (e.g., separation, loss of employment, financial difficulties, serious illness of a family member) are rated by respondents on a 3-point scale (0 to 2; did not occur, mildly stressful, very stressful). Its internal consistency has been calculated at 0.63. Total score was used in the analyses as a measure of stress level. Stress from events following disclosure was assessed through 8 items (e.g., medical examination, testimony in court, placement in foster care) rated on the same 3-point scale. Internal consistency was calculated at 0.70. Total score was used in the analyses.

A sociodemographic questionnaire was used to collect general information on parents and CSA victims, such as birth date, gender, educational level, ethnic background, and family composition. The professional in charge of the family at the site provided information regarding CSA with categorical items. These included severity (from exhibitionist to penetration), length of the abuse and, victim's age at the time of the abuse, the victim's relationship to the perpetrator, whether the perpetrator lived with the victim at the time of the abuse, and the date of disclosure.

For the first objective, which aims to compare the effect of parents' gender and the effect of time (pre- versus postdisclosure), generalized estimating equations (GEE) were used given that both parents participated in some cases. GEE are often used to analyze correlated data (Hanley, Negassa, & Forrester, [26]) due to their flexibility in accommodating both continuous and binary outcomes. They allow the statistical treatment of repeated measurements, such as those used in within-family conditional models. To facilitate the interpretation of effects, continuous independent variables were standardized for the purpose of comparison.

The second objective was to determine whether variables related to the parent, to the sexual abuse and to stressful events were related to parental health. As in studies in the field, it is not possible to measure parents' mental and physical health before CSA disclosure; analyses should identify other predictors of parents' health, if any. As a preliminary step, the correlation between the dependent variables and the following variables was assessed: family income, parental educational level, parental occupational status, number of types of child maltreatment, parent and child gender, intra- versus extrafamilial abuse, severity and length of the abuse, age of the perpetrator, whether the perpetrator lived with the victim at the time of the abuse, interval between disclosure and evaluation, stress from life events in the past year, and stress from events following disclosure. Because no significant correlation emerged regarding the CSA variables, child gender, and interval between disclosure and evaluation, these were not included in the analyses. Because family income, educational level, and occupational status were highly correlated, we chose to include only the first of these three in the analyses. In the end, the GEE included parents' gender, family income, number of types of child maltreatment, stress from life events, stress from events following disclosure, and the pre-score for health dimensions if applicable. Ten GEEs were calculated for psychological distress, one for each self-reported PTSD, diagnosis of PTSD, major depressive disorder, health status, physical limitations and three for the types of professional consultation.

All the conditions assessed (see Table 1) were postdisclosure, except for self-reported PTSD related to past trauma, which was evaluated with the MPSS-SR. Nearly half of the mothers and one-third of the fathers reported suffering from psychological distress. Regarding PTSD related to CSA disclosure, 32% of the mothers met the criteria when evaluated by trained clinicians with the SCID, compared with only 13.1% based on self-report (MPSS-SR). For the fathers, the rates were comparable based on diagnostic interview and self-report (7.1% and 7.3%, respectively). The prevalence of self-reported PTSD related to past trauma was lower for both genders (5.3% for mothers and 4% for fathers). Regarding major depressive disorder, 40.8% of the mothers and 14% of the fathers exhibited symptoms based on the diagnostic interview.

Table 1. Participants Meeting Diagnostic Criteria for Mental Health Status, by Gender.

3 * p < .05.

Overall, no gender effect emerged in terms of PTSD related to past trauma and self-report. However, other symptoms were significantly more prevalent among the mothers, namely psychological distress, PTSD, and depression (see Table 1).

To identify other variables that might also impact mental health, a series of GEE were carried out. The only result to reach statistical significance was the relationship between clinical scores of psychological distress and higher level of stress from life events (b = 0.55, p = .01). Higher family income (b = 1.09, p = .001) and higher level of postdisclosure stress (b = 0.53, p = .03) were significantly related to self-reported PTSD (MPSS-SR). PTSD, as evaluated with the SCID, was explained in part by respondent gender (b = 2.36, p = .03) and postdisclosure stress (b = 0.74, p = .01), which indicated that the likelihood of PTSD diagnosis was higher for mothers and for those with a higher level of postdisclosure stress. None of the five independent variables were significantly correlated with a major depressive episode.

Although the data for physical symptoms were also collected after CSA disclosure, the parents were asked to report symptoms pre- and postdisclosure. As shown in Table 2, half of the mothers and more than half of the fathers rated their predisclosure health as excellent or very good. A significant time effect was noted, with more than 10% of mothers and fathers seeing their health status decline postdisclosure; no gender effect was observed. The participants were also asked to report whether their activities were limited by their health. In this regard, a significant time effect was noted, with a higher number of both mothers and fathers reporting significant limitations postdisclosure, although no gender effect was noted.

Table 2. Self-Perceived Physical Health Status, Diminished Activity, Consultations and Medication Use of Mothers and Fathers Before and After CSA Disclosure.

4 *** p < .001. ** p < .01. * p < .05.

The GEEs used to verify which variables were related to health status and limitations revealed that only predisclosure health status (b = -0.62, p = .02) was significant for health status. However, presence of predisclosure activity limitations (b = 0.55, p = .001), higher income (b = 0.24, p = .01), and more life event stressors in the past year (b = 0.22, p = .001) were related to postdisclosure activity limitations.

The mothers and fathers' use of professional services and medication predisclosure and postdisclosure is reported in Table 2. A significant gender effect was noted for family doctor, psychologist, social worker, and use of antidepressants, with proportionally more mothers than fathers using these services or medications. However, although fewer fathers used antidepressant medication after disclosure, the fathers showed a greater increase than the mothers did because few of them were using antidepressants before disclosure. The same was true for family doctor consultations. A significant time effect was observed regarding the same professionals, with increased consultation of family doctors, psychologists, and social workers for both gender after disclosure. Due to the small percentage of parents using anxiolytic medication, time comparisons and regression tests could not be calculated.

To verify which variables other than CSA disclosure potentially related to postdisclosure professional consultation, a series of GEE analyses were carried out. Predisclosure family doctor consultation (b = 0.18, p = .005) and higher level of stress from life event stressors (b = 0.11, p = .001) was related to postdisclosure family doctor consultation. Postdisclosure psychologist consultation was related to predisclosure consultation (b = 0.56, p = .001), higher number of types of child maltreatment (b = 0.24, p = .01), and higher stress postdisclosure (b = 0.21, p = .01). Postdisclosure social worker consultation was correlated to predisclosure consultation (b = 0.70, p = .001) and respondent gender (b = 0.49, p = .05), with a significantly higher proportion of women consulting these professionals.

The aim of the present study was to evaluate self-perceived health of nonoffending mothers and fathers and prevalence of professional consultation and medication use following CSA disclosure. As with other studies in the field, and given the cross-sectional nature of the study, it is not possible to know whether the parents' mental and physical health predated or followed disclosure. Because the parents were solicited at the time they were seeking services for their child, this sample included self-referred participants. Therefore, this convenience sample is not representative of all nonoffending parents of sexually abused children. In addition, the large time span between the disclosure of the abuse and the time of the evaluation could have hindered potential effects in the analyses. However, the second objective of this study was precisely to reflect the extent of some of these limitations by determining whether factors other than the disclosure of sexual abuse could explain the parents' health. In general, the results revealed that a fair proportion of parents faced psychological and physical health problems and that more mothers reported health issues than fathers did when they have faced CSA disclosure.

Of the mental health conditions, current psychological distress, PTSD, and major depressive disorder were diagnosed and reported in our sample at rates that surpassed those observed in standard populations. For example, the rates of psychological distress were approximately twice as high for mothers (49.1% versus 22.8%) and fathers (30.2% versus 17.7%) compared to the general population of Canadians living in Quebec (Légaré et al., [36]). The rates observed in the study for depression were those most markedly different when compared with standard population data (Canadian Community Health Survey; Statistics Canada, [44]): 40.8% versus 6.3% for women and 14.3% versus 3.8% for men. For PTSD, the rates observed for fathers based on self-report measures and diagnostic interviews were in line with the expected general population lifetime rate of approximately 8% (DSM-IV, 2000). However, the rate for mothers was four times as high when based on a diagnostic interview and slightly higher when based on self-report measures.

Our results are consistent with those of Davies ([13]) and Manion and colleagues ([37]), who observed that fathers reported fewer symptoms of depression and psychological distress than mothers did after CSA disclosure. For PTSD symptoms, our results reproduce the inconsistency observed in previous research. The self-reported measures converged with those of Kelley ([31]) who found no gender effect. However, the results with a diagnostic measure are congruent with those reported by Davies, who observed that none of the 14 fathers in his study obtained a clinical score on intrusiveness and only four did on avoidance, whereas more mothers reached clinical levels. Manion and colleagues also reported a gender difference on the intrusiveness scale. The difference observed for women between self-reported PTSD and PTSD diagnosed by a clinical interview is questioning and could suggest a problem with measurement, although the scores obtained by men with the same questionnaires are more consistent. This result might suggest that trained clinicians were probably able to access more symptoms and distress to make their diagnoses, or that these clinicians were biased in their assessment of PTSD symptoms among mothers. However, this gender difference in PTSD symptoms has been observed in other stressful situations (Vishnevsky, Cann, Calhoun, Tedeschi, & Demakis, [48]).

Our study is the first to take a closer look at parents' physical health after CSA disclosure. Both mothers and fathers rated their self-perceived health status prior to disclosure within the same range as normal populations, with slightly more than half considering their health in the past year to be excellent or very good (Bernèche, Camirand, & Dufour, [2]). After disclosure, the women in our study differed from the general population more markedly, whereas the men managed to remain within the normal range. The same pattern was observed regarding activity limitations, with the rates for both genders exceeding those in the general population. In this regard, compared with Canadians surveyed about their actual limitations (Statistics Canada, [45]), the women exceeded the rate by 19.5 percentage points (38.5% versus 19%) and the men by 5.4 (27.9% versus 22.5%). However, the mothers and fathers did not differ in their self-evaluation of health and activity limitations, although we did note a significant decrease in physical health condition and increase in activity limitations after disclosure.

The status of some parental health postdisclosure, as evidenced by the levels of distress and functional impairment in the sample, prompted some parents to seek support from health and social services. Compared with a standard population of Canadians living in Quebec (Fournier & Piché, [23]), a higher percentage of mothers and fathers in our sample turned to health care professionals. Postdisclosure, twice as many mothers and fathers in our study consulted medical specialists, and three times as many consulted a family doctor or general practitioner compared with 17.8% and 17%, respectively, for women in the general population, and 12.7% and 12.1%, respectively, for men. In addition, one-third to one-half of the mothers and 16.3% to one-half of the fathers reported consulting social workers and psychologists, compared with less than 1% of women and men in the general population. However, these higher rates for social worker and psychologist consultation postdisclosure must be interpreted with caution because some consultations could have been suggested by child protection services (CPS) and could have included contact for the family with CPS at the time of the police investigation. Gender differences were noted in terms of service use, with mothers being more likely to seek the help of social workers and psychologists and fathers relying more on general practitioners. Although this result may be due to the characteristics of our convenience sample or to gender reporting biases, these results are consistent with findings in the literature that suggest that men tend to resolve personal issues independently from the others and are reluctant to seek help from significant others or professionals to address psychological trauma (Gallo, Marino, Ford, & Anthony, [24]).

Gender differences were observed on some variables, but not all. Some of these gender differences in parents' mental and physical health are consistent with the differential vulnerability hypothesis, which suggests that due to the nature of the stressful event—with a family member involved—mothers experience it as more stressful and react to it with greater intensity compared to fathers. According to Denton and colleagues ([16]), the vulnerability hypothesis proposes that women's health differs from men's not only because they live in a social context (e.g., lower income) that impacts their health but also because women are more concerned with and affected by others' well-being. However, this hypothesis needs more empirical support in the context of CSA disclosure.

To gain a better understanding of the potential impact of CSA disclosure on parental health, we also examined whether other variables might have an impact. Among these, life event stressors experienced in the past year played a significant role in the prevalence of psychological distress, functional limitations, and general practitioner consultation, whereas stress from events specifically related to CSA disclosure was related to PTSD symptoms. These findings converge with those of Dyb and colleagues ([18]), who observed both the presence of numerous stressful events following CSA and persistence of PTSD symptoms many years after disclosure, as well as with those of Cyr and colleagues ([11]) with regard to psychological distress. In addition, there was a higher prevalence of PTSD among higher-income earners and parents who reported diminished activity due to deteriorated health. This last result is counterintuitive given that low socioeconomic status is usually a strong predictor of PTSD (Brewin, Andrews, & Valentine, [4]). A possible explanation for this finding might be that CSA generates more negative social judgments directed at nonoffending parents in higher socioeconomic classes, which leads to more PTSD symptoms. Accordingly, Dunmore, Clark, and Ehlers ([17]) found the development of PTSD in sexual or physical abuse victims to be positively associated with their negative perception of other people's reactions (e.g., victims perceiving that people are ashamed of them). Furthermore, the participants in our study reported very low family income on average, which suggests that many lived below the poverty level. This skewed distribution could have accentuated the difference between the poor participants and those who were slightly more fortunate. This negative effect was not surprising as chronic life stressors have often been related to higher psychological distress and greater functional limitations in life (Denton et al., [16]).

Contrary to expectations, three variables—childhood maltreatment, CSA variables, and interval between disclosure and evaluation—were not related to the parents' postdisclosure adaptation; childhood maltreatment was only associated with psychologist consultation. The presence of multiple forms of childhood maltreatment—a widely recognized risk factor in the development of mental health conditions in adulthood (Edwards et al., [19])—was not related to mental or physical health measures in our study. More proximal variables, such as daily life stressors, had a larger impact in the months following disclosure. Psychologist consultations were associated with a history of adverse childhood experiences; these results might indicate that parents have overcome the potential impact of their past maltreatment. In addition, no difference was observed as a function of many CSA variables. This result is congruent with the observations of other studies (Cyr et al., [11]; Deblinger et al., [15]; Manion et al., [37]) that found no relationship with CSA variables, while only two studies reported some results with one variable each. These findings suggest that although intrafamilial situations could be more difficult to handle and could have more repercussions on different spheres of life, a variety of difficulties could also arise from extrafamilial situations. These results imply that above and beyond sexual abuse characteristics and the interval between disclosure and evaluation, parents have vulnerabilities and resources that influence their reactions. Although Kelley ([31]) has observed a weak but significant relationship with the time elapsed since CSA, she also observed that distress and PTSD were still present for parents 2.2 years after disclosure. More research is needed to gain a fuller understanding of these variables. Cluster analysis could help in this regard because it might reflect parental reactions more comprehensively than the analysis of a series of isolated variables (Cyr, McDuff, & Hébert, [10]).

Some of the limitations of our study deserve mention. Except for PTSD evaluated by clinicians, our study was based on self-report measures and relied on memory recollection. Consequently, some parents could have forgotten, minimized, or exaggerated important aspects of their health. Some dimensions of physical health were based only on one item, thus limiting the robustness of this measure. These measures were chosen because they provided comparative data; however, their validity or reliability was not reported. The interval between disclosure and the evaluation was relatively long and could not be compared with other studies because many did not report it. Although we checked for the potential impact of this variable and found no relation, our results reflect the parents' situation one year after disclosure. The cross-sectional and retrospective nature of the design does not allow us to infer that CSA disclosure was responsible for the parents' mental and physical state, because many parents could have suffered from difficulties before disclosure. In addition, these participants were volunteers and recruited in a clinic offering services for their sexually abused children. The participants included both intra- and extrafamilial CSA; however, more intra- than extrafamilial were included in our study. This finding seems to reflect the naturalistic situation where more families who have experienced intrafamilial CSA seek help for their children. Because of this convenience sample, the present sample is not representative of all CSA, with many of the instances of abuse neither disclosed nor reported to the authorities.

The results reveal considerable levels of stress, psychological distress, and physical problems for mothers and fathers following disclosure of CSA. However, not all parents experience psychological distress or physical problems. Further research is needed to determine whether our results can be generalized, given that the parents were recruited as a result of having sought help for their children. In addition, it would be necessary to conduct prospective studies given that the health status of caregivers for victims of CSA over time is insufficiently documented and that CSA disclosure is only the first in a series of events that could keep parents under stress for months and years.

Some of the results of the study suggest that the physical and psychological effects of CSA disclosure could be intertwined. If these results can be confirmed by other studies, then the development of global intervention plans involving partnerships and collaboration between health care workers and social service providers should be developed. Given that some help-seeking strategies could differ by gender, as evidenced by the fact that a larger number of the fathers in our study turned to their family doctors/general practitioners, it may be helpful to develop and implement intervention protocols and referral mechanisms sensitive to these gender differences within the health and social services systems. These parents could also benefit from group interventions, because these groups provide a supportive environment in which mothers and fathers can discuss mental and physical health issues, learn new coping strategies for supporting their children, and address the difficulties and symptoms that they face (Tavkar & Hansen, [47]). Treatment-Focused Cognitive Behavior Therapy (Cohen et al., [7]), a best practice recommended for sexually abused children, has also proved to be efficient in reducing parents' depressive symptoms and increasing parental skills, particularly supportive and protection interventions. For mothers and fathers presenting depressive and psychological distress symptoms, a number of well-validated therapies could be applied. For those with PTSD symptoms, a more in-depth exploration of the presence of disclosure-related stressors and their subjective and objective significance is needed because our results suggest a relationship between these variables. It is clearly essential to address the specific needs of nonoffending parents through tailored interventions, given that parents remain the most significant source of support for their children following CSA disclosure.

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000. Respondents gave consent before taking part in the study, which was approved by the ethics board of both Université de Montréal and Centre jeunesse–Institut Universitaire de Québec.

This research was supported by a grant from the Canadian Institutes of Health Research (IRSC 172315) to the first four authors, the Chaire de recherche interuniversitaire Marie-Vincent sur les agressions sexuelles envers les enfants (Marie-Vincent Inter-University Research Chair on Sexual Abuse Against Children) and the Centre de recherche sur les problèmes conjugaux et les agressions sexuelles (CRIPCAS: Research Centre on Intimate Relationship Problems and Sexual Abuse).