Family functioning in adolescents and young adults with differences of sex development

Abstract

Objective

Family functioning influences various psychosocial outcomes for individuals with pediatric chronic health conditions (e.g., Leeman, J., Crandell, J. L., Lee, A., Bai, J., Sandelowski, M., & Knafl, K. (2016). Family Functioning and the Well-Being of Children With Chronic Conditions: a Meta-Analysis. Research in Nursing & Health, 39, 229–243), but this has not been examined among families of youth with differences of sex development (DSD). The objective of this study was to examine (a) differences in family functioning based on DSD-related and demographic characteristics, (b) the associations between family functioning and psychosocial outcomes, and (c) the moderating effects of current age.

Methods

Participants were 57 adolescents and 40 young adults (AYA) ages 12–25 years (M  = 16.97) with DSD resulting in atypical genital appearance and/or discordance between phenotypic and chromosomal sex who participated in a cross-sectional study examining psychosocial outcomes. Participants completed the Family Environment Scale, the McMaster Family Assessment Device, the Perceived Acceptance Scale, and outcome measures including the Youth Self-Report/Adult Self-Report and the Connor-Davidson Resilience Scale. Analyses included t-tests and linear regressions with moderation effects.

Results

AYA reported better family functioning if they were diagnosed at a younger age, learned about their diagnosis at a younger age, had 46, XY karyotype, and identified as male. General family functioning, family conflict, and family acceptance had the strongest associations with psychosocial outcomes. Greater family acceptance was associated with fewer total problems, and this association was stronger for adolescents (ps < .05).

Conclusions

AYA with certain DSD-related characteristics may demonstrate more adaptive family functioning. Family functioning is strongly associated with psychosocial outcomes for AYA with DSD. Results highlight the importance of developing family-focused interventions to promote psychosocial functioning in AYA with DSD.

“Differences of sex development” (DSD) is a term used to describe a heterogeneous group of congenital conditions marked by variations in chromosomal, gonadal, and/or phenotypic sex (Lee et al., 2006). Some DSD, including those that are the focus of this study,¹ involve atypical genital appearance and/or discordance between phenotypic and chromosomal sex. DSD may be diagnosed early through prenatal testing, at birth due to atypical genital appearance, discordance between prenatal testing and phenotypic appearance, or through evaluation of childhood medical concerns. In other cases, a DSD may not be identified until later in life when issues with puberty occur, e.g., lack of pubertal progression, amenorrhea, or clitoral enlargement upon pubertal onset. A range of terms have been proposed as alternatives to DSD; for example, some individuals identify as intersex (Bennecke et al., 2021). Although we acknowledge lack of consensus regarding terminology, we use DSD here as it was the language used to communicate with participants during the course of this research. Variations in definitions and terminology can complicate incidence rate estimates, which vary considerably across conditions and are thought to range from 1 in 4,500-5,500 (Lee et al., 2006).

Despite the growing body of psychosocial research on individuals with DSD, to our knowledge, few studies have examined family functioning among children, adolescents, or young adults with DSD. A recent cross-sectional study found that parents of youth with DSD reported lower levels of family conflict compared to parents of children with cleft lip and palate or parents of unaffected children, greater family cohesion compared to the unaffected group, and no significant differences in family expressiveness between any of the groups (Hansen-Moore et al., 2021). However, the authors noted limitations including limited variability in the data and poor internal consistency on measures of family functioning.

Additional research on family functioning among families of youth with DSD is warranted given the implications that such research can have for clinical care in DSD, a field in need of more robust psychosocial care standards (Dessens et al., 2017; Ernst et al., 2018). While there is a considerable body of literature on family functioning among families of youth with chronic health conditions, extrapolating those findings to the field of DSD is challenging, in part, due to the mixed nature of these studies’ findings (Herzer et al., 2010; Law et al., 2019; Leeman et al., 2016; McClellan & Cohen, 2007; Mendes et al., 2016; Neris et al., 2023; Rolland & Walsh, 2006). Some studies have found less adaptive family functioning among families of youth with chronic health conditions, while other studies have observed families to be resilient, with similar or even better family functioning compared to families of unaffected youth (Herzer et al., 2010; Lennon et al., 2015; McClellan & Cohen, 2007). The Family Systems-Illness (FSI) Model (Rolland & Walsh, 2006) highlights the complexity of how chronic health conditions can impact family functioning and draws attention to the important contributions of condition-specific characteristics (i.e., condition onset, course, outcome, incapacitation, progression uncertainty), family systems dynamics and beliefs (e.g., multigenerational experiences of illness, meaning making of the illness), and the goodness of fit between the psychosocial demands over time and the strengths and vulnerabilities of a family.

Family functioning among youth with DSD may be unique compared to other pediatric populations for several reasons, many of which map onto the condition-specific characteristics that are highlighted in the FSI Model (Rolland & Walsh, 2006). First, because DSD can affect genital appearance and sexual and reproductive function, the sharing of information about a DSD condition may be influenced by concerns about privacy, stigma, discomfort, or perceived inability to accurately describe the condition (Chivers et al., 2017; Crissman et al., 2011; Moyer et al., 2023). This may be relevant to information-sharing within the family (between caregivers and the child with a DSD, siblings, extended family, etc.) and outside the family (between caregiver/child and others such as teachers, babysitters, peers). Navigating information-sharing may impact family communication and lead to disagreements, and thus, family conflict (Crissman et al., 2011). Second, research has found that caregivers of children with DSD experience perceived stigma (e.g., Traino et al., 2022), which may influence family functioning through a caregiver’s acceptance of their child or their child’s DSD or caregiver adjustment (Traino et al., 2022). Third, uncertainty related to DSD diagnosis and decision-making (e.g., gender of rearing, surgical decision-making) can predict depression and anxiety in parents of youth with DSD (Roberts et al., 2020). While shared decision-making is recommended in such cases, there are disagreements about the management of DSD and how to balance the preferences of patients with DSD and their caregivers (Sandberg & Vilain, 2022). Additionally, some DSD conditions require daily management (e.g., the use of corticosteroids in congenital adrenal hyperplasia [CAH]), which may impact family functioning in ways comparable to other chronic health conditions (Psihogios et al., 2019). Inversely, many DSD may not require specific medical attention for several years, suggesting, in theory, that there may be less of an impact on day-to-day family functioning. Lastly, some DSD require multiple surgeries, which can be a significant stressor for families. These factors exemplify the heterogeneity that characterizes DSD and highlights the need for investigating family functioning in this population.

Equally important as understanding family functioning is understanding the connection between family functioning and psychosocial outcomes for youth and young adults (AYA) with DSD. Research with other chronic health conditions has demonstrated that certain facets of family functioning may be more influential than others (e.g., family cohesion and conflict compared to family communication; Leeman et al., 2016), and that certain psychosocial outcomes may be more influenced by family functioning than others (e.g., problem behaviors compared to social competence; Leeman et al., 2016). Research on psychosocial functioning among AYA with DSD is limited and mixed. Several studies on pediatric samples have found internalizing, externalizing, and/or social problems to be in the nonclinical range on average (Hansen-Moore et al., 2021; Kleinemeier et al., 2010; Sandberg et al., 2017), while another study found youth with DSD are more likely to have a behavioral health diagnosis compared with matched controls (Sewell et al., 2021). DSD comprise a heterogenous group of conditions, and differences have emerged when examining specific conditions or DSD-related characteristics (Kleinemeier et al., 2010; Sewell et al., 2021). Importantly, studies on adults with DSD have identified high levels of psychological distress compared to adults without DSD (Bennecke et al., 2017), suggesting that increased attention to psychosocial functioning during adolescence and young adulthood may prevent or reduce issues arising later. Thus, understanding how family functioning impacts psychosocial outcomes in AYA with DSD has implications for psychosocial intervention with this population.

The current study aims to address existing gaps in knowledge about family functioning in AYA with DSD. The objectives are to examine (a) differences in family functioning based on DSD-related and demographic characteristics, (b) the associations between family functioning and psychosocial outcomes, and (c) the moderating effects of current age. This study did not make explicit hypotheses given the mixed findings that characterize the literature on family functioning among families of youth with chronic health conditions generally, and the uniqueness of the DSD population.

Methods

Study design and participants

Data were drawn from a larger multisite, multimethod, cross-sectional study assessing factors that affect quality of life, resilience, and psychosocial adjustment in adolescents (ages 12–17 years) and young adults (ages 18–26 years) with DSD. Participants were recruited from four large children’s hospitals with well-established interdisciplinary DSD teams in Ohio, Massachusetts, Colorado, and Illinois. In addition, young adults who had transitioned out of pediatric care were recruited from adult practices in Ohio and Massachusetts. Inclusion criteria included (1) age 12–26 years, (2) DSD diagnosis resulting in atypical genital appearance and/or discordance between phenotypic and chromosomal sex (individuals with DSD who do not typically experience these discordances were not included, i.e., Klinefelter syndrome, Turner syndrome), and (3) the ability to read and understand English. Exclusion criteria included (1) the presence of significant comorbid medical conditions (e.g., cancer), (2) the presence of current severe mental illness (e.g., active psychotic disorder), (3) intellectual or cognitive delay, as this may prevent completion of questionnaires, and (4) being unaware of the DSD condition.

Procedures

This study was approved by the Institutional Review Board at Nationwide Children’s Hospital (IRB16-00476); the remaining sites ceded review to the single IRB through reliance agreements. Potential participants were identified via chart review and contacted by trained research staff, who completed the eligibility screening with parents of minors or patients themselves if over age 18. Adult participants provided informed consent. Adolescents ages 12–17 years provided informed assent and their parents/legal guardians provided parental permission. All participants completed questionnaires via Research Electronic Data Capture (REDCap; Harris et al., 2009); a subset of participants completed qualitative interviews, but that data are not included in the present study. The current study includes questionnaire data on demographic information, medical health history, family functioning, and psychosocial adjustment. Participants received $50 for completing questionnaires.

Measures

Medical chart review

Diagnosis, medical, and surgical history for each participant were extracted from medical charts by trained research assistants using a standardized form, in order to supplement the medical history information reported directly by participants via questionnaire.

Demographic and Medical History Questionnaire

Participants provided demographic and medical information including current age, diagnosis, age at time of diagnosis, age when first informed about their diagnosis, current gender identity, gender of rearing, whether there was uncertainty about sex designation at birth, race, ethnicity, and insurance type. For adolescent participants, medical history questions were completed by parents.

McMaster Family Assessment Device

The Family Assessment Device (FAD) is a 60-item measure of perceptions of one’s family functioning, based on the McMaster Model of Family Functioning, (Miller et al., 1994). This study included two scales. The General Functioning scale includes 12 items assessing the overall health/pathology of the family (e.g., “Making decisions is a problem for our family”). The Communication scale includes 6 items assessing the clarity of content in verbal messages and directness of communication among family members (e.g., “People come right out and say things instead of hinting at them”). Items are rated on a 4-point scale (1 = Strongly Disagree to 4 = Strongly Agree), with certain items requiring reverse coding and lower scores indicating better family functioning/communication. For this study, internal consistencies were General Functioning, α = .89; Communication, α = .61.

Family Environment Scale

The Family Environment Scale (FES) assesses perceptions of social and environmental characteristics of the family (Moos & Moos, 2009). This study included the three subscales from the Family Relationship Domain, each with 9 items: the Cohesion subscale (e.g., “Family members really back each other up.”); the Expressiveness subscale (e.g., “We say anything we want to around home.”); and the Conflict subscale (e.g., “We fight a lot in our family.”). Items are rated as True or False. Some items require reverse coding. Higher scores indicate higher family cohesiveness/expressiveness/conflict. For this study, internal consistencies were Cohesion, α = .83; Expressiveness, α = .55; Conflict, α = .77.

Perceived Acceptance Scale

The Perceived Acceptance Scale (PAS) assesses perceptions of being accepted by family members (Brock et al., 1998). This study included three subscales. The Mother and Father subscales each include 10 parallel items (e.g., “At times, my [mother/father] has made me feel that [she/he] didn’t approve of me.”). The Family subscale includes 12 items (e.g., “I am a very important part of the lives of my family.”). Items are rated on a 5-point scale (1 = Disagree to 5 = Strongly Agree). Some items require reverse coding. Higher scores indicate greater acceptance. For this study, internal consistencies were Mother, α = .88; Father, α = .91; Family, α = .95.

Youth Self-Report and Adult Self-Report

The Youth Self-Report (YSR) and Adult Self-Report (ASR) are widely used and standardized instruments that measure psychosocial adjustment (Achenbach & Rescorla, 2001, 2003). The current study included items that comprise the Internalizing (e.g., symptoms of depression, anxiety; “I feel lonely”), Externalizing (e.g., rule breaking behaviors; “I argue a lot”), and Total Problems subscales. Items are rated on a 3-point scale (0 = Not True to 2 = Very True or Often True). Higher scores reflect greater psychosocial difficulties. Raw scores are converted to t-scores (mean of 50), with t-scores between 65 and 69 indicating problems in the borderline range, and t-scores ≥70 in the clinically severe range. For this study, internal consistencies were YSR Internalizing, α = .93; YSR Externalizing, α = .85; YSR Total Problems, α = .73; ASR Internalizing, α = .94; ASR Externalizing, α = .87; and ASR Total Problems, α = .85.

Connor-Davidson Resilience Scale

The Connor-Davidson Resilience Scale assesses resilience and ability to cope with stress (Connor & Davidson, 2003). It includes 25 items (e.g., “I am able to adapt when changes occur”) that are rated on a scale from 0 (Not at All True) to 4 (True Nearly All the Time). Items scores are averaged, with higher scores reflecting greater resilience. For this study, internal consistency was α = .94.

Data analysis

Analyses were completed using SPSS version 29. Listwise deletion was used to address missing data, which was 0.75% across all 22 variables. For Objective 1, t-tests were conducted to examine differences in family functioning (overall family functioning, family communication, family cohesion, family conflict, family expressiveness, family acceptance, maternal acceptance, paternal acceptance) based on the following dichotomized DSD-related clinical and demographic variables: karyotype (46, XX karyotype vs. 46, XY karyotype), age at diagnosis (before age 2 years vs. age 2 years or older), age when first informed about their diagnosis (before age 10 years vs. age 10 years or older), uncertainty about sex designation at birth (yes vs. no), and current gender identity (female vs. male). Variables were dichotomized for analytic purposes due to the limited sample size within certain levels of each variable. Assuming a power of .80, and an α of .05, analyses was powered to detect medium (ƞ² = .25) to large (ƞ² = .40) effect sizes (Cohen, 1992). For Objectives 2 and 3, multiple regression analyses were conducted to determine whether family functioning variables were associated with psychosocial outcomes (internalizing symptoms, externalizing symptoms, total problems, and resilience). Separate regression analyses were conducted for associations of each family functioning measure (FAD, FES, PAS; subscales of each measure were entered simultaneously) with each of the 4 psychosocial outcomes, for a total of 12 regression models. Regressions were run three ways: without covariates, controlling for current age, and controlling for current gender identity. For significant models, current age was examined as a potential moderator using methods outlined by Aiken and West (1991) and Holmbeck (2002). Assuming a power of .80, and an α of .05, Objective 2 and 3 analyses were powered to detect medium (R² = .15) to large (R² = .35) effect sizes (Cohen, 1992).

Results

Participant characteristics

Of 301 AYA screened for eligibility, 247 were found to be eligible. Of those, 82 could not be contacted, 36 actively declined, 10 did not enroll for other reasons, and 3 enrolled but later withdrew. The resulting sample comprised 116 participants, 97 of whom completed quantitative measures: 57 adolescents and 40 young adults ages 12–25 years (M = 16.97). There were no significant differences in race, ethnicity, or insurance type between those who did and did not enroll and between those who did and did not complete quantitative measures. Table 1 shows participant demographic and descriptive information.

Table 1.

Demographic and descriptive information.

	N = 97 M (SD) or n (%)
Age, years	16.97 ± 3.68
Gender of rearing
Female	54 (55.7)
Male	43 (44.3)
Current gender identity
Female	53 (54.6)
Male	42 (43.3)
Non-binary/fluid	2 (2.0)
Self-reported race
American Indian/Alaska Native	2 (2.1)
Asian	8 (8.2)
Black/African American	7 (7.2)
White	67 (69.1)
More than one race	6 (6.2)
Other/unknown	7 (7.3)
Self-reported ethnicity
Not Spanish/Hispanic/Latinx origin	78 (80.4)
Spanish/Hispanic/Latinx origin	13 (13.4)
Unknown	6 (6.2)
Insurance
Private/commercial	66 (68.0)
Medicaid	29 (29.9)
No insurance	2 (2.1)
Karyotype/diagnosis
46, XX	33 (34.0)
CAH, 21-hydroxylase deficiency	16 (16.5)
CAH, 11β-hydroxylase deficiency	1 (1.0)
Mayer–Rokitansky–Küster–Hauser syndrome	14 (14.4)
Ovotesticular DSD	1 (1.0)
SRY translocation	1 (1.0)
46, XY	58 (59.8)
Testicular regression syndrome (vanishing testes)	16 (16.5)
Complete androgen insensitivity syndrome	9 (9.3)
Partial androgen insensitivity syndrome	6 (6.2)
Severe hypospadias with other anomalies	9 (9.3)
Isolated severe hypospadias	5 (5.2)
Gonadal dysgenesis	6 (6.2)
Syndromic gonadal dysgenesis (Frasier syndrome)	3 (3.1)
5α-reductase deficiency	3 (3.1)
17β-hydroxysteroid dehydrogenase deficiency	1 (1.0)
Other karyotype	6 (6.2)
Mixed gonadal dysgenesis	5 (5.2)
Mayer–Rokitansky–Küster–Hauser syndrome	1 (1.0)
Uncertain gender designation at birth
No	60 (61.9)
Yes	37 (38.1)
Age at diagnosis
At birth	38 (39.2)
Within first year of life	19 (19.6)
2–5 years	11 (11.3)
6–11 years	8 (8.2)
12–16 years	10 (10.3)
16 years old or older	11 (11.3)
Age at learning of diagnosis, years
0–5	23 (23.7)
6–9	18 (18.6)
10–14	32 (33.0)
15–18	15 (15.5)
19 years or older	1 (1.0)
Unknown/unclear	8 (8.2)

Note. CAH = congenital adrenal hyperplasia.

Objective 1: differences in family functioning

Table 2 reports M, SD, and correlations among study variables. Though not examined statistically, when comparing the means for the FES subscales in this study to what has been published previously in the literature (i.e., caregiver-reports on the FES as reported in Hansen-Moore et al., 2021), this study found lower cohesion and expressiveness and more conflict. To our knowledge, data using the other family functioning measures used in this study have not been reported in other published studies on DSD. On the internalizing problems subscale of the YSR/ASR, n = 7 (7.2%) reported borderline elevations (t-scores between 65 and 69) and n = 7 (7.2%) reported clinical elevations (t-score ≥ 70). On the externalizing problems subscale, n = 6 (6.2%) reported borderline elevations and n = 2 (2.1%) reported clinical elevations. On the total problems subscale, n = 6 (6.2%) reported borderline elevations and n = 6 (6.2%) reported clinical elevations. With the exception of two associations, all study variables were significantly correlated with one another in the expected direction.

Table 2.

Means and correlations for family functioning and psychosocial outcomes.

Variable	M (SD)	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.
Family Functioning
1. General	1.86 (.51)	–	.56***	−.72***	−.48***	.53***	−.82***	−.60***	−.42***	.49***	.29**	.42**	−.53**
2. Communication	2.16 (.44)		–	−.38***	−.49***	.25*	−.45***	−.37***	−.31**	.27**	.25*	.27**	−.37**
3. Cohesion	6.98 (2.33)			–	.51***	−.56***	.67***	.47***	.35***	−.48***	−.32***	−.47***	.32***
4. Expressiveness	5.60 (1.97)				–	−.22*	.37***	.25*	.171	−.27**	−.095	−.25*	.26*
5. Conflict	2.80 (2.31)					–	−.63***	−.35***	−.31**	.54***	.63***	.60***	−.27**
6. Family Acceptance	46.90 (8.78)						–	.72***	.56***	−.60***	−.43***	−.53***	.53***
7. Mother Acceptance	42.34 (6.90)							–	.61***	−.57***	−.37***	−.50***	.59***
8. Father Acceptance	38.61 (8.88)								–	−.51***	−.30**	−.43***	.46***
Psychosocial
9. Internalizing	52.69 (12.45)									–	.64***	.90***	−.59***
10. Externalizing	50.03 (10.13)										–	.85***	−.28***
11. Total Problems	51.46 (11.19)											–	−.50***
12. Resilience	72.29 (16.16)												–

Note. For General and Communication subscales of the McMaster Family Assessment Device (FAD), scores can range from 1 to 4 with lower scores indicating better functioning. For the Cohesion, Expressiveness, and Conflict subscales of the Family Environment Scale (FES), scores can range from 0 to 9 with higher scores indicating more of the construct. For the Family, Mother, and Father Acceptance subscales of the Perceived Acceptance Scale (PAS), scores can range from 18 to 72 with higher scores indicating more acceptance.

^* p < .05,

^** p < .01,

^*** p <.001.

Table 3 presents differences in family functioning based on DSD-related and demographic characteristics. There was a significant difference based on karyotype; AYA with 46, XY karyotype reported higher family acceptance than those with 46, XX karyotype (p = .01). There was also a significant difference based on age at diagnosis; compared with AYA who were diagnosed at 2 years of age or older, AYA who were diagnosed before age 2 reported better general family functioning (p < .001), higher family cohesion (p = .003), lower family conflict (p < .001), and higher family (p < .001), mother (p = .01), and father acceptance (p < .001). There was a significant difference based on age when first informed of their diagnosis; AYA who learned of their diagnosis before age 10 reported better general family functioning (p = .01) and higher family acceptance (p = .02) than those who learned of their diagnosis at age 10 or older. There was a significant difference based on current gender identity; compared with AYA who identify as female, AYA who identify as male reported higher family cohesion (p < .001), lower family conflict (p = .002), and higher family acceptance (p = .01). (Notably, there were n = 2 participants who did not identify as female or male and were not included in that analysis.) There were no significant differences based on whether or not there was uncertainty about sex designation at birth (p’s > .05).

Table 3.

Differences in family functioning based on DSD-related and demographic characteristics.

	Karyotype	Age at diagnosis	Age at diagnosis disclosure	Gender identity	Sex designation uncertainty
General functioning	ns	t(62.07) = 3.88***	t(86) = −2.53*	ns	ns
		<2 years: 1.70 (.39)	≤9 y: 1.73 (.40)
		≥2 years: 2.11 (.57)	≥10 y: 2.00 (.58)
Communication	ns	ns	ns	ns	ns
Cohesion	ns	t(57.43) = −3.05**	ns	t(70.25) = 3.67***	ns
		<2 years: 7.59 (1.77)		Female: 6.29 (2.80)
		≥2 years: 6.05 (2.75)		Male: 7.83 (1.08)
Expressiveness	ns	ns	ns	ns	ns
Conflict	ns	t(93) = 3.73***	ns	t(92.85) = −3.24**	ns
		<2 years: 2.12 (1.98)		Female: 3.43 (2.45)
		≥2 years: 3.82 (2.43)		Male: 2.00 (1.86)
Family acceptance	t(87) = 2.64*	t(51.54) = −4.54***	t(70.50) = 2.35*	t(86.24) = 2.74**	ns
	46, XX: 43.63 (10.35)	<2 years: 50.22 (5.53)	≤9 years: 48.96 (5.67)	Female: 44.88 (10.12)
	46, XY: 48.53 (7.13)	≥2 years: 41.92 (10.36)	≥10 years: 44.76 (10.56)	Male: 49.45 (5.92)
Mother acceptance	ns	t(53.12) = −2.60*	ns	ns	ns
		<2 years: 43.84 (4.92)
		≥2 years: 39.82 (8.63)
Father acceptance	ns	t(92) = −3.42***	ns	ns	ns
		<2 years: 41.05 (7.93)
		≥2 years: 35.01 (9.08)

Note. ns = not significant (p ≥ .05). M (SD) are given below t-test results. For the General Functioning subscale, lower scores indicate better family functioning.

^* p < .05,

^** p < .01,

^*** p < .001.

Objective 2: associations between family functioning and psychosocial outcomes

Table 4 presents associations between family functioning and psychosocial outcomes. When examining FAD subscales, greater general family functioning was associated with lower scores for internalizing and total problems and with greater resilience (ps < .001). When examining FES subscales, greater family cohesion was associated with less internalizing problems (p = .03), and greater family conflict was associated with greater internalizing, externalizing, and total problems (ps < .001). When examining PAS subscales, greater mother acceptance was associated with greater resilience (p = .005), and greater family acceptance was associated with less internalizing (p = .005), externalizing (p = .03), and total problems (p = .01). Findings remained the same when controlling for current age. When controlling for current gender identity, one additional finding emerged that greater family acceptance was associated with greater resilience (p = .04).

Table 4.

Associations of family functioning variables with psychosocial outcomes.

Family functioning	Psychosocial outcomes
	Internalizing problems			Externalizing problems			Total problems			Resilience
	β	t	ΔR2	β	t	ΔR2	β	t	ΔR2	β	t	ΔR2
FAD			.24			.10			.18			.29
General	.50 ***	4.56		.23 ns	1.90		.40 ***	3.48		−.48***	−4.52
Communication	−.01 ns	−.05		.12 ns	1.00		.05 ns	.40		−.10 ns	−.94
FES			.35			.40			.39			.13
Cohesion	−.25*	−2.19		.02 ns	.18		−.17 ns	−1.50		.17 ns	1.23
Expressiveness	−.06 ns	−.62		.04 ns	.41		−.06 ns	−.63		.15 ns	1.28
Conflict	.38 ***	3.73		.65 ***	6.67		.49 ***	4.97		−.14 ns	−1.21
PAS			.42			.19			.32			.38
Family Accept.	−.34**	−2.85		−.32*	−2.28		−.33*	-2.58		.20 ns	1.66
Mother Accept.	−.21 ns	−1.65		−.11 ns	−.74s		−.17 ns	−1.29		.37 **	2.87
Father Accept.	−.20 ns	−1.91		−.06 ns	−.46		−.14 ns	−1.21		.12 ns	1.16

Note. A simultaneous multiple regression model was conducted for associations of each family functioning measure (i.e., FAD, FES, PAS) with each psychosocial outcome (i.e., Internalizing Problems, Externalizing Problems, Total Problems, Resilience), for a total of 12 regression models. This table presents results from analyses that did not include covariates. For FAD subscales, lower scores indicate better family functioning/communication. ΔR² = R squared change, reported as effect-size estimates for the model; Accept = Acceptance; FAD = McMaster Family Assessment Device; FES = Family Environment Scale; ns = not significant (p ≥ .05); PAS = Perceived Acceptance Scale. Significant results are in bold print.

^* p < .05,

^** p < .01,

^*** p < .001.

Objective 3: age as a moderator

Current age group (adolescents vs. young adults) was examined as a potential moderator for significant models. One significant interaction emerged between family acceptance and current age group for total problems (b = .44, SE = .22, β = .23, t = 2.00, p = .04, ΔR² = .36). Post hoc simple slope regression analyses revealed that greater family acceptance was associated with fewer total problems for adolescents (b = −.87, SE = .14, β = -.70, t = −6.28, p = <.001) and young adults (b = −.34, SE = .17, β = −.28, t = −2.06, p = .04), but the association was stronger for adolescents (see Figure 1).

Figure 1.

Post hoc probe of significant interaction between family acceptance and age group associated with total problems. *p < .05, ***p < .001.

Discussion

The purpose of this study was to expand the limited research on family functioning among AYA with DSD. The present study identifies how family functioning among AYA with DSD varies based on certain DSD-related and demographic characteristics, and the ways in which family functioning is associated with psychosocial outcomes. Importantly, the present study identifies areas of challenges and resilience for families of AYA with DSD.

In examining differences in family functioning based on DSD-related and demographic characteristics, significant differences were found based on age at diagnosis, age when first informed of their diagnosis, karyotype, and current gender identity; no differences were found based on whether there was uncertainty about sex designation at birth. Age at diagnosis was a particularly salient characteristic, in that AYA who were diagnosed before 2 years old reported better family functioning compared with those who were diagnosed at 2 years old or later. It may be that when caregivers are aware of a diagnosis early on in a child’s life, the diagnosis is less disruptive to the family system because it allows for greater caregiver adjustment or acceptance. In addition, this study found that AYA who learned about their diagnosis prior to age 10 reported better family functioning compared to those who learned about their diagnosis after age 10. Previously reported analysis of qualitative data from the current study revealed that AYA expressed preference for learning about condition-related information (e.g., their fertility status) as young as possible (Papadakis et al., 2021), expressing the belief that coping with potentially emotionally upsetting information about oneself would be less difficult at a younger age. AYA also talked about being impacted by their caregivers’ emotional reactions when first learning condition-related information. Together, these results suggest that families for whom a child’s DSD diagnosis occurs later and/or whose child learns about their diagnosis later would benefit from increased psychosocial support. It also suggests that parents be counseled about the potential benefit of informing youth about their diagnosis in a developmentally informed manner as early as possible. Because this study measured age at diagnosis and age at first learning about their diagnosis in ranges and not continuously, it is not possible to determine the exact time between actual diagnosis and learning of the diagnosis. Collecting these data in future studies may allow for further examination of these relationships.

When examining differences in family functioning based on karyotype, there was one significant finding in that AYA with a 46, XY karyotype reported greater family acceptance compared with AYA with a 46, XX karyotype. When examining differences based on current gender identity, additional findings emerged, in that compared with females, males reported greater family acceptance and family cohesion, and less family conflict. In the current study, there was incongruency between karyotype and current gender identity in 27 AYA (27.8% of the sample; 22 with current female gender identity and 5 with current male gender identity). More specifically, of males, 85.7% had a 46, XY karyotype and 2.4% had a 46, XX karyotype, and of females, 38.9% had a 46, XY karyotype and 59.3% had a 46, XX karyotype. While some past research suggested that family dynamics and parenting are influenced by child gender (Fossum et al., 2007), more recent findings suggest that child gender does not influence parenting styles, likely due to generational shifts in parenting approaches and views on gender roles (e.g., Endendijk et al., 2016). Alternatively, the current study’s findings may be explained by gender differences in the perceptions of family dynamics (given that, e.g., females are better at identifying nonverbal emotional cues; Wingenbach et al., 2018). The gender differences in the current study may also reflect the influence of specific DSD conditions represented among males and females in this sample. The majority (close to 53%) of AYA with female gender identity in this sample had either CAH or MRKH. It may be that CAH and/or MRKH present unique challenges to families compared with other DSD conditions. Interestingly, past studies have found that parents of boys with DSD report higher levels of depression and illness uncertainty compared with parents of girls (Suorsa et al., 2015; Wolfe-Christensen et al., 2012).

This study did not find significant differences in family functioning based on whether there was uncertainty at birth about the sex of rearing. Such a difference may have been expected given research that has found differences in parenting behavior (e.g., overprotectiveness) due to children’s genital ambiguity (Sanders et al., 2012). Given the heterogeneity that characterizes the DSD population, future research will benefit from continued examination of family functioning within specific DSD conditions.

Broadly, this study confirms that family functioning is associated with psychosocial outcomes in AYA with DSD, and is concordant with findings found for other pediatric chronic health conditions. General family functioning, family conflict, and family acceptance were more closely associated with child psychosocial outcomes than family cohesion and mother or father acceptance. Notably, across all objectives, there were no significant findings for family communication and family expressiveness. The lack of findings may be due to the relatively poor internal consistency found for these two subscales. It may also be that these two subscales are assessing nuanced aspects of family communication.

Family and parental acceptance has not been examined as often as other family functioning constructs in the literature. In the current study, family acceptance was associated with internalizing, externalizing, and total problems, but the mother and father acceptance subscales were not. Because the PAS does not specify who to define as part of “family,” some respondents may think of their nuclear family (e.g., parents, siblings), while others may think of the larger family (e.g., grandparents, cousins). For a clinician assessing psychosocial adjustment in AYA with DSD, this finding highlights the potential importance of understanding who a young person considers to be in their family and the degree of support they receive from them. Interestingly, mother acceptance was associated with greater resilience while father acceptance was not. Child development research has found differential impacts of mother and father acceptance on child outcomes. For example, a meta-analysis found maternal acceptance was more likely to contribute to socioemotional development, whereas paternal acceptance tended to be related to children’s problem behaviors and psychopathology (Li & Meier, 2017). These findings have implications for clinical intervention with families of AYA with DSD. As described previously, parents of youth with DSD and youth themselves may be at risk for perceiving stigma due to their DSD condition (Traino et al., 2022). Thoughtful assessment and intervention of stigma-related feelings and facilitation of acceptance of a youth’s DSD condition would likely benefit the youth in their ability to cope with ongoing challenges.

Current age is expected to be a relevant variable when examining family functioning among AYA, given that, as adolescents transition to young adulthood, they are likely to spend less time with their parents, be less dependent on their parents, and often no longer live with their parents (Arnett, 2007). In the current study, when examining associations between family functioning and psychosocial outcomes and controlling for current age, findings did not change. When examining current age as moderator, family acceptance was associated with fewer total problems for adolescents and young adults, though the association was stronger for adolescents. Given that our data is cross-sectional, it cannot be determined whether adolescents who experience less problems are more likely to feel accepted by their families, or if feeling accepted by one’s family leads to less problems. Future longitudinal research can aim to unpack the direction of effects.

Importantly, this study found rates of internalizing, externalizing, and total problems to fall predominantly in the nonclinical range, consistent with past studies that have examined samples of youth with a range of DSD diagnoses (Hansen-Moore et al., 2021; Kleinemeier et al., 2010; Sandberg et al., 2017). These findings are promising, if they indeed reflect that AYA with DSD experience are not at risk for mental health concerns. However, given that research on adults with DSD has more consistently found high rates of psychological distress, there is still reason to attend to how psychosocial functioning unfolds over the transition to adulthood. Given this study’s findings that family functioning is related to these outcomes, attention to the family unit as potential intervention target is warranted.

Strengths of this study include a relatively large sample of AYA with DSD representing a range of ages and DSD-related characteristics, use of multiple measures of family functioning and psychosocial outcomes, and use of AYA self-report (rather than caregiver report). Limitations of this study include the cross-sectional design, which limits the ability to determine causation; lack of DSD-specific measures of family functioning; dichotomization of DSD-related characteristics for analyses; and inclusion of a broad range of DSD conditions, which limits the ability to identify associations unique to a specific condition. While the racial and ethnic composition of this study reflected the hospitals at which recruitment occurred, and there were no significant differences in race, ethnicity, or insurance type between those who did and did not enroll, the majority of participants were White. Future studies are needed to address these limitations and expand examination of family factors (e.g., family socioeconomic status or family culture/values) as they relate to psychosocial and medically related outcomes among AYA with DSD.

The current study has important implications, some already mentioned, for behavioral health care provided to families of AYA with DSD. It would be beneficial for behavioral health providers working with AYA with DSD to give attention to the family system components within a biopsychosocial framework. Using a clinical framework like that offered by the FSI Model (Rolland & Walsh, 2006) may be especially useful given the heterogeneity that characterizes DSD and the need to appreciate how condition-specific factors may interact with family dynamics. In addition, behavioral health providers working in DSD should feel encouraged to use sound self-report measures or tools for assessing family functioning (Alderfer et al., 2008). Lastly, given that this study found factors such as the age youth are first informed of their diagnosis to be relevant to family functioning, this highlights the importance of behavioral health providers establishing relationships with families of youth with DSD early on in their care, preferably as part of an interdisciplinary team (Lee et al., 2006). While it has been strongly recommended that youth with DSD receive medical care through an interdisciplinary team that involves psychosocial services (Hughes et al., 2006; Sandberg et al., 2017), many youth have inadequate access to psychosocial care (Ernst et al., 2018). The current study is intended to motivate efforts and inform approaches for increasing access of psychosocial care to youth with DSD.

Data availability

Data available on request.

Author contributions

Jaclyn L. Papadakis (Conceptualization [lead], Formal analysis [lead], Writing—original draft [lead], Writing—review & editing [lead]), Cindy L. Buchanan (Investigation [equal], Supervision [equal], Writing—review & editing [equal]), Yee-Ming Chan (Investigation [equal], Supervision [equal], Writing—review & editing [equal]), Canice E. Crerand (Conceptualization [equal], Data curation [equal], Funding acquisition [equal], Investigation [equal], Methodology [equal], Project administration [equal], Resources [equal], Supervision [equal], Writing—review & editing [equal]), Jennifer Hansen-Moore (Investigation [equal], Supervision [equal], Writing—review & editing [equal]), Leena Nahata (Investigation [equal], Supervision [equal], Writing—review & editing [equal]), Joseph Rausch (Formal analysis-supporting, Software [equal], Supervision [equal], Validation [equal], Writing—review & editing [equal]), Amy Claire Tishelman (Conceptualization [equal], Data curation [equal], Funding acquisition [equal], Investigation [equal], Methodology [equal], Project administration [equal], Resources [equal], Supervision [equal], Writing—review & editing [equal]), and Diane Chen (Conceptualization [equal], Investigation [equal], Supervision [equal], Writing—review & editing [equal])

Funding

This study was supported by R21HD089526 from the Eunice Kennedy Shiver National Institute of Child Health and Human Development (MPIs: Crerand & Tishelman). Study sponsors had no role in (1) study design; (2) collection, analysis, and interpretation of data; (3) writing of the article; or (4) the decision to submit the article for publication.

Conflicts of interest: none declared.