Marital Conflict, Sibling Closeness, and Youth Adjustment in Mexican-Origin Families: A Multilevel Moderation Analysis

Abstract

Marital conflict is a well-documented family stressor linked to youth maladjustment. However, limited research has examined protective factors that may buffer these effects in Mexican-origin families, where cultural values emphasizing familial harmony may intensify the negative impact of marital conflict. Guided by family systems theory, this multi-informant, longitudinal study investigated how mother- and father-reported marital conflict related to youth-reported adjustment and whether sibling closeness moderated these associations. Participants included 246 two-parent Mexican-origin families with two adolescent siblings (49.59% girls, Mage = 14.49, SD = 1.85 at Time 1), assessed across three time points. Multilevel moderation models were used to examine between-family, within-person, and cross-level effects. Results revealed that, within families, increases in parent-reported marital conflict over time were not significantly associated with decreases in youth depressive symptoms or risk behaviors. However, compared to other families, higher levels of mother-reported marital conflict were associated with greater youth-reported depressive symptoms and risk behaviors. Further, sibling closeness moderated the association between father-reported marital conflict and youth depressive symptoms at the between-family level. Specifically, in families characterized by high sibling closeness, lower levels of marital conflict were linked to fewer youth depressive symptoms. However, in families with low sibling closeness, youth reported elevated depressive symptoms regardless of marital conflict levels. These results highlight the nuanced ways marital conflict relates to youth outcomes and underscore the importance of examining multiple family subsystems at multiple levels. They also inform culturally relevant frameworks by identifying sibling closeness as a potential protective factor in family-based interventions.

During adolescence, a period marked by expanding social roles and responsibilities (Hartmann & Swartz, 2006), family relationships play a pivotal role in shaping youth emotional and behavioral adjustment (Edler et al., 2025). In this context, marital conflict represents a well-documented source of family stress that has been linked to negative family climates and youth maladjustment (for a review, see van Eldik et al., 2020). However, prior research has focused primarily on White families, limiting our understanding of these dynamics in other ethnoracial contexts. It is therefore important to explore how marital conflict is perceived and experienced within families, and how it relates to youth outcomes in other cultures. In Mexican culture, where family harmony and cohesion are strongly emphasized, disruptions such as marital conflict may be especially harmful to youth adjustment, including depressive symptoms and risk behaviors (Cahill et al., 2021; Martin et al., 2019). Given that Mexican-origin youth comprise 70% of Latino youth, representing the largest minoritized group in the United States (U.S. Census Bureau, 2023), gaining a deeper understanding of how marital conflict affects Mexican-origin youth is critical for informing a broader and more inclusive knowledge base on family processes and youth adjustment.

According to family systems theory, individuals and subsystems within families are interconnected, and their interactions collectively shape youth adjustment (Cox & Paley, 2003). In Mexican-origin families, strong relational ties may intensify the impact of marital conflict on youth adjustment (Cahill et al., 2021; Taylor et al., 2012) while also fostering resilience through supportive relationships, particularly among siblings (Updegraff et al., in press). Siblings hold a unique position within the family system, sharing experiences of family dynamics while offering companionship and emotional support (Jensen et al., 2023). Sibling closeness, defined as the warmth, intimacy, and mutual support shared between siblings, may thus serve as a key protective factor buffering youth from family stress. This is particularly relevant for Latino youth, who tend to have more siblings and spend more time with them compared to their Black and White peers, underscoring the critical role of siblings in these families (Ackert & Wikle, 2022).

Building on prior research, the present study examines how mothers’ and fathers’ reports of marital conflict relate to youth adjustment (i.e., depressive symptoms and risk behaviors) in Mexican-origin families, and whether sibling closeness moderates these associations in adolescence. In addition, to contribute to a more nuanced understanding of these family dynamics, we disentangle both interpersonal and intrapersonal associations, comparing marital conflict across families as well as changes within families over time. Using a multilevel moderation approach, we examine three dimensions as shown in Figure 1: (1) how average levels of marital conflict, sibling closeness, and youth adjustment differ across families (between-family effects, BF), (2) how changes in marital conflict, sibling closeness, and youth adjustment are associated within the same families (within-person effects, WP), and (3) how within-family changes in marital conflict and youth adjustment vary by the average levels of sibling closeness across families (cross-level effects). Particularly, testing cross-level interactions advances our understanding of how enduring family dynamics and time-varying family processes jointly explain youth adjustment. These insights have important implications for the development of culturally sensitive prevention and intervention strategies to support youth well-being in Mexican-origin families.

Figure 1.

Illustrations of Different Moderation Effects.

Marital Conflict and Youth Adjustment

Marital conflict can be defined as the frequency and intensity of negative emotions such as anger and resentment (Braiker & Kelley, 1979). It plays a critical role in youth adjustment, particularly in relation to emotional well-being (e.g., depressive symptoms) and behavioral outcomes (e.g., risk-taking behaviors; van Eldik et al., 2020). The family process model posits that marital conflict may disrupt parent-child interactions, increasing the likelihood of internalizing and externalizing behaviors (Zemp et al., 2018). From emotional security theory, marital conflict may threaten children’s sense of safety, fostering vigilance and distress that contribute to depressive symptoms and risk behaviors (Kouros et al., 2008). Yet, research has disproportionately focused on mothers’ perspectives, with fathers’ experiences less often examined (Perez-Brena et al., 2022) despite evidence that family members hold distinct views of conflict with unique implications for children’s adjustment (Cheung et al., 2016; Ehrlich et al., 2019). Considering gendered roles in the family, mothers typically provide emotional care, while fathers often assume provider or authority roles (Cabrera et al., 2009; Kramer et al., 2021), suggesting that mothers’ reports may be particularly salient for youth outcomes. The father vulnerability hypothesis suggests that fathers may have greater difficulty separating marital and parental domains, making their parenting, and subsequently children’s outcomes, more sensitive to marital conflict (e.g., Taylor et al., 2012). These processes may be especially relevant in Mexican-origin families, where two-parent households are common (Chen & Guzman, 2022), fathers often hold hierarchical authority (Prieto et al., 2022), and mothers are primary caregivers shaping the family’s emotional climate (Kramer et al., 2021).

Prior work incorporating both fathers’ and mothers’ perspective often combines parent reports, and few studies have examined how mothers’ and fathers’ perspectives uniquely relate to youth adjustment over time (Martin et al., 2019; Taylor et al., 2012). Among adolescents, Cheung et al. (2016) found that only father-reported constructive marital conflict predicted lower insecurity one year later in predominantly White families. Research with younger White children has produced mixed findings, some showing similar effects for fathers and mothers (e.g., Zhao et al., 2021), and others suggest stronger effects for mothers (e.g., Brock & Kochanska, 2015) or for fathers under certain contextual conditions (e.g., Lange et al., 2022). Given these developmental differences and inconsistent findings, the present study adopts an exploratory approach to examine how Mexican-origin mothers’ and fathers’ reports of marital conflict differentially relate to youth adjustment over time.

Longitudinal designs allow researchers to distinguish between overall differences across families (between-family; BF effects) and changes within individuals over time (within-person; WP effects). However, most prior work has emphasized BF differences, comparing families with varying levels of marital conflict to assess whether higher average marital conflict predicts greater youth adjustment problems (e.g., Cheung et al., 2016). In contrast, WP approaches examine whether periods of higher-than-usual marital conflict correspond to increases in youth problems, controlling for each family’s typical marital conflict level (Curran & Bauer, 2011). Recognizing the distinction between BF and WP effects is essential, as pooling them can mask meaningful patterns at each level (Dietvorst et al., 2018). This approach can help distinguish enduring family characteristics from patterns of change as related to youth adjustment, enhancing theoretical interpretations and informing more precisely targeted interventions. Empirical evidence underscores this distinction. For example, Zemp et al. (2018) found that German families with higher average marital verbal aggression had children with greater internalizing and externalizing problems, whereas within-family changes in aggression were unrelated to children’s adjustment. Similarly, Knopp et al. (2017) examined marital conflict among a sample of predominantly White military families, finding that although families with higher verbal marital conflict exhibited more child internalizing and externalizing problems (BF effects), increases in verbal marital conflict only predicted greater externalizing problems (WP effects). Together, these findings highlight that the strength and nature of the associations between marital dynamics and child adjustment depend on both the analytic level and the outcomes considered.

Extending this work, the present study examines both mother- and father-reported marital conflict in relation to Mexican-origin youth’s depressive symptoms and risk behaviors. Based on theory and prior between-family findings, we hypothesize that higher average marital conflict will predict higher overall youth depressive symptoms and risk behaviors (BF effect). In contrast, evidence for within-family (WP) effects is limited and inconsistent across studies and outcomes, suggesting that within-family changes in marital conflict may not have the same implications for youth adjustment as chronic, high-conflict family patterns (e.g., Zemp et al., 2018). Given this variability and the lack of a consistent empirical pattern, we do not offer a directional hypothesis for WP effects and instead examine them in an exploratory manner. Examining both levels simultaneously clarifies how enduring and dynamic aspects of marital conflict shape youth adjustment in Mexican-origin families.

The Role of Sibling Closeness

Siblings hold a distinct and influential role within the family system, blending familial and peer-like qualities that provide emotional support, guidance, and stability (Buist et al., 2013). During adolescence, they often serve as companions, confidants, and role models, helping one another navigate transitions and stress (Jensen et al., 2023). These dynamics may be particularly significant in Mexican-origin families, where siblings spend substantial time together and play central roles in family life (Ackert & Wikle, 2022). Indeed, sibling closeness in the current sample has been linked to fewer depressive symptoms and lower engagement in risk behaviors (Killoren et al., 2021). Siblings may also buffer the effects of marital hostility on outcomes (e.g., Masarik & Rogers, 2019). From emotional security theory, strong sibling ties can help youth manage the stress of marital conflict by promoting feelings of safety and stability, reducing emotional insecurity, and supporting adaptive coping (Davies et al., 2019).

Despite growing recognition of the important role of siblings, evidence remains limited regarding how sibling closeness moderates the association between marital conflict and youth adjustment (Jensen et al., 2023; Perez-Brena et al., 2022). Two theoretical patterns have been proposed: enhanced (stress-buffering) or diminished (reverse stress-buffering) effects of social support (Rueger et al., 2016). An enhanced effect occurs when sibling closeness mitigates (or buffers) the negative association between marital conflict and youth adjustment. For instance, Davies and colleagues (2019) found that marital conflict predicted increased problem behaviors over time among adolescents with low-quality sibling relationships, whereas this association was not significant among those with high-quality sibling relationships. In contrast, the diminished effect occurs when the negative association between marital conflict and youth adjustment emerges only for youth with high levels of sibling closeness. This was observed in El-Sheikh & Elmore-Staton’s (2004) cross-sectional study of a predominantly White sample. They found that among children with high attachment security to their mothers, greater marital conflict was more strongly associated with externalizing problems. In contrast, for children with low attachment security, externalizing problems remained consistently higher but showed a weaker association with marital conflict.

Importantly, the moderating effect of sibling closeness may operate at different levels (see Figure 1). Although previous studies have highlighted the protective role of sibling closeness, they often combined both BF and WP levels, leaving uncertainty about how these factors interact to predict youth adjustment at different levels (e.g., Davies et al., 2019). Specifically, the WP question asks whether the association between marital conflict and depressive symptoms is stronger when youth experience higher-than-usual sibling closeness. The BF question investigates whether the association of family-level marital conflict and individual-level depressive symptoms is stronger for youth in families with higher (versus lower) average sibling closeness. Finally, the cross-level moderation explores whether the WP relationship between marital conflict and depressive symptoms is stronger for youth from families with higher (versus lower) average sibling closeness. Accordingly, the second goal is to examine whether sibling closeness moderates the association between marital conflict and youth adjustment in Mexican-origin families at BF and WP levels.

Current Study

Contributing to the limited research on the role of marital conflict in youth adjustment within Mexican-origin families, the first goal is to examine the associations between mother- and father-reported marital conflict and youth reports of depressive symptoms and risk behaviors, both across families and within youth over time. The second goal is to test whether sibling closeness moderated the associations between marital conflict and youth adjustment across families and within youth over time. By examining the associations both across families and time, this study offers valuable insights into the complex interplay between familial factors and youth well-being, with implications for developing prevention and intervention strategies. Additionally, considering prior research suggesting variations in Latino youth adjustment by family socioeconomic status (SES) and nativity, these factors are included as covariates (Martin et al., 2019; Park et al., 2018).

Methods

Procedure and Participants

This study was approved by the Institutional Review Board at Arizona State University (IRB protocol: STUDY00015653). The data came from a longitudinal study of 246 two-parent Mexican-origin families (Updegraff et al., 2006) in the greater Phoenix area (Maricopa County, AZ). The original sample size was determined through a priori power analyses aligned with the study’s aims as proposed in the funding application to the National Institute of Child Health and Human Development. At Wave 1, eligibility of families included: (a.) the mother being Mexican-origin; (b.) a younger sibling in the 7^th grade; (c.) at least one older sibling, a biological mother, and a father living in the home together; and (d.) the father works at least 20 hours per week, including both formal and informal (e.g., cash based) work. Most youth were biologically related to their father figure, whereas 5.08% had adoptive or non-adoptive fathers. Families were recruited via letters in Spanish and English sent to 7^th-graders at local participating schools, and bilingual researchers made follow-up phone calls to determine eligibility and willingness to participate.

This dataset includes four waves collected from 2002 to 2010. For the current analysis, data from three of these waves were utilized: Wave 1 (2002–2003), Wave 3 (five years later), and Wave 4 (seven years later), referred to as Time 1, Time 2, and Time 3, respectively. Wave 2 was excluded from the analysis as it only included data on younger siblings. At Time 1, 246 families participated, followed by 185 families (75.20%) at Time 2, and 173 families (70.33%) at Time 3. No participating families were excluded from the analytic sample.

At Time 1, fathers had an average age of 41.49 years (SD = 5.80) ranging from 29.83 to 61.00 years, and mothers averaged 39.28 years (SD = 4.63) ranging from 30.00 to 55.75 years. Older siblings were, on average, 15.48 years old (SD = 1.58), ranging 13.25 to 21.17 years. They were typically in 9th grade through college, with a mean grade level of 9.78 (SD = 1.32), with 50% being female (n = 123). Younger siblings averaged 12.51 years old (SD = 0.58), ranging from 11.92 to 15.33 years, with 49.20% being female (n = 121). There were 68 older-girl-younger-girl sibling pairs, 55 older-girl-younger-boy sibling pairs, 57 older-boy-younger-girl sibling pairs, and 66 older-boy-younger-boy sibling pairs, and the age gap between siblings was 2.96 years (SD = 1.63) ranging from 1 to 9 years. On average, parents had been married for 17.57 years (SD = 5.42), and had 3.79 children (SD = 1.60), with a range of 2 to 15 children. Marriages included those recognized in legal, religious, or other cultural contexts. Regarding birthplace, 69.11% of fathers (n = 170), 70.74% mothers (n = 174), 46.34% of older siblings (n = 114), and 37.80% of younger siblings (n = 93) were born in Mexico. The remaining mothers and older and younger siblings were born in the U.S.; the remaining fathers were born in the United States (28.86%; n = 71), another Latin country (1.22%; n = 3), or elsewhere (0.41%; n = 1).

Participants included families from various education and income levels. At T1, the annual family income was $53,184 (SD = $45,382), ranging $3,000 to over $250,000. Regarding parental education levels, fathers completed an average of 9.88 years of education (SD = 4.37), and mothers averaged 10.34 years (SD = 3.74). The median education level for both parents was 11 years, indicating that over half of the parents did not complete high school. Attrition analyses showed that families with high family SES (including household income and parents’ education) at T1 were less likely to be missing in marital conflict, sibling closeness, depressive symptoms, and risk behaviors at T2 and T3 (all ps < 0.05). Thus, SES was used as a covariate in all analyses.

Measurement

Mothers and fathers reported on their marital conflict with the Relationship Questionnaire (Braiker & Kelley, 1979). The conflict subscale includes 5 items, such as, “How often do you and your spouse/partner argue with one another?” Each item was rated on a 9-point scale (such as 1 = “Not at all” and 9 = “Very much”). Scores were averaged across items, with higher scores indicating higher levels of conflict. Cronbach α ranged from .67 to .81 for fathers and from 0.67 to 0.80 for mothers across time.

Youth reported on their depressive symptoms using the Center for Epidemiological Studies Depression Scale (Radloff, 1977). It included 20 items, such as “During the past month, I thought my life had been a failure.” and they were rated from 1 = “Rarely or none of the time” to 4 = “Most of the time.” Higher summed scores indicated higher level of depressive symptoms. Cronbach α for siblings across time ranged between .85 to .90.

Youth also reported on their engagement in risk behaviors with a 23-item measure from the Michigan Study of Youth Transitions (Eccles & Barber, 1990). Sample items include, “Ruined or damaged other people’s things on purpose including tagging” and “Tried drugs other than those for which you had a prescription.” Responses were rated from 1 (“Never”) to 4 (“More than 10 times”), with higher average scores representing more risk behaviors in the past year. Cronbach α for siblings across time ranged between .86 to .91.

Youth reported on their sibling closeness using an 8-item scale developed by Blyth and colleagues (Blyth et al., 1982; Updegraff et al., 2002). Items were rated on a 5-point scale (1 = “Not at all” and 5 = “Very much”), with examples like, “How much do you share your feelings or secrets with your sibling?” Higher average scores indicated greater closeness. Cronbach α ranged between .81 and .86 for siblings across time.

Moderators included youth age, sex (0 = female, 1 = male), birth order (0 = older sibling, 1 = younger sibling), dyad sex constellation (0 = mixed-sex, 1 = same-sex), and sibling age gap. Covariates included nativity (0 = Mexico-born, 1 = U.S.-born) and family socioeconomic status (SES). Family SES is composited from mothers’ and fathers’ education levels and the log of household income, standardized to correct for skewness.

Analysis plan

Using a three-level nested dataset (time points within siblings within families), this study used SAS 9.4 to assess the association between father- and mother-reported marital conflict and youth adjustment (depressive symptoms and risk behaviors), and tested whether sibling closeness moderated these links at BF and WP levels. Marital conflict was modeled at within-person (WP) and between-person (BP) levels, and sibling closeness at within-person (WP), between-sibling (BS), and between-family (BF) levels. WP variables were the deviation of scores at each time point from person mean across three time points, BS sibling closeness reflected each sibling’s deviation from the family mean, and BF sibling closeness and BP marital conflict were family-mean centered at the sample level.

Separate multilevel models were conducted for depressive symptoms and risk behaviors, Model 1 included WP and BP marital conflict reported by mothers and fathers, WP, BS, and BF sibling closeness, youth and family covariates (age, sex, birth order, nativity, dyad sex constellation, sibling age gap, SES). Model 2 included interaction terms between marital conflict and sibling closeness following Park et al.’s (2018) multilevel moderation framework. Interaction terms were tested separately for mothers’ and fathers’ reports of marital conflict for each youth outcome, resulting in four moderation models. In each model, three moderation effects were tested: (a) within-person(π_ij4), assessing whether within-person associations between marital conflict and youth adjustment depended on time-varying sibling closeness; (b) between-family (γ₇₀₀), assessing whether family-mean conflict–adjustment links varied by person-mean sibling closeness; and (c) cross-level(γ₀₁₁), assessing whether within-person relations varied by person-mean sibling closeness. Significant interactions were probed at ±1 SD of sibling closeness. Moderation effects were examined separately for father- and mother-reported marital conflict in relation to both youth outcomes. In total, six models were tested: two main-effects models (Model 1) and four moderation models (Model 2). Equations for the father-reported interaction model predicting depressive symptoms are provided below.

Level 1:

$$\begin{gathered} {\mathit{Depression}}_{ijk}={\pi }_{ij0}+{\pi }_{ij1}{\mathit{WP.F.MarConflict}}_{ik}+{\pi }_{ij2}{\mathit{WP.M.MarConflict}}_{ik}+{\pi }_{ij3}{\mathit{WP.SibCloseness}}_{ijk} \\ +{\pi }_{ij4}{\mathit{WP.F.MarConflict}}_{ik}\times {\mathit{WP.SibCloseness}}_{ijk}+{\epsilon }_{ijk} \end{gathered}$$ (4)

Level 2:

$$\begin{gathered} {\pi }_{ij0}={\beta }_{i00}+{\beta }_{i10}{\mathit{BS.Closeness}}_{ij}+{\beta }_{i20}{\mathit{Age}}_{ij}+{\beta }_{i30}{\mathit{Sex}}_{ij} \\ +{\beta }_{i40}{\mathit{BirthOrder}}_{ij}+{\pi }_{i50}{\mathit{Nativity}}_{ijk}+{\mu }_{ijk} \end{gathered}$$ (5)

$${\pi }_{ij1}={\beta }_{i01}$$ (6)

$${\pi }_{ij2}={\beta }_{i02}$$ (7)

$${\pi }_{ij3}={\beta }_{i03}$$ (8)

$${\pi }_{ij4}={\beta }_{i04}$$ (9)

Level 3:

$$\begin{gathered} {\beta }_{i00}={\gamma }_{100}{\mathit{BP.F.MarConflict}}_i+{\gamma }_{200}{\mathit{BP.M.MarConflict}}_i+{\gamma }_{300}{\mathit{BF.Closeness}}_i+{\gamma }_{400}{\mathit{SexCon}}_i+{\gamma }_{500}{\mathit{AgeGap}}_i \\ +{\gamma }_{600}{\mathit{SES}}_i+{\gamma }_{700}{\mathit{BP.F.MarConflict}}_i\times {\mathit{BF.Closeness}}_i+v_{i00} \end{gathered}$$ (10)

$${\beta }_{i01}={\gamma }_{001}+{\gamma }_{011}{\mathit{BS.SibCloseness}}_i$$ (11)

$${\beta }_{i02}={\gamma }_{002}$$ (12)

$${\beta }_{i03}={\gamma }_{003}+{\gamma }_{013}{\mathit{BP.F.MarConflict}}_i$$ (13)

$${\beta }_{i04}={\gamma }_{004}$$ (14)

$${\beta }_{i10}={\gamma }_{010}$$ (15)

$${\beta }_{i20}={\gamma }_{020}$$ (16)

$${\beta }_{i30}={\gamma }_{030}$$ (17)

$${\beta }_{i40}={\gamma }_{040}$$ (18)

$${\beta }_{i50}={\gamma }_{050}$$ (19)

Sibling interdependence was modeled using a multilevel dyadic approach (Planalp et al., 2016), retaining a compound symmetry covariance structure due to non-significant sibling covariances. Additionally, time effects were excluded to preserve meaningful covariation between changes in marital conflict and youth adjustment (Wang & Maxwell, 2015).

Given the modest sample size and the complexity of multilevel moderation models, power to detect small interaction effects was limited. Accordingly, results, particularly moderation findings, should be interpreted with caution. The study was not preregistered, and data and analytic code are available from the corresponding author upon request.

Results

Descriptives

The means, standard deviations, and bivariate correlations across time points are shown in Table 1. Father- and mother-reported marital conflict are positively related to youth depressive symptoms at Time 1 and Time 2. However, some negative associations were observed between father-reported marital conflict and youth risk behaviors. In addition, fathers’ and mothers’ reports were positively correlated at all three time points (ps < .05), however, paired-samples t-tests showed fathers reported lower levels of marital conflict compared to mothers at Time 1 (t = 3.74, p < .001), Time 2 (t = 4.11, p < .001), and Time 3 (t = 4.57, p < .001).

Table 1.

Correlations, Means, and Standard Deviations for Study Variables

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1. FR marital conflict T1	-
2. FR marital conflict T2	.51**	-
3. FR marital conflict T3	.61**	.63**	-
4. MR marital conflict T1	.32**	.25**	.27**	-
5. MR marital conflict T2	.23**	.29**	.28**	.56**	-
6. MR marital conflict T3	.21**	.32**	.38**	.49**	.63**	-
7. Sibling closeness T1	−.02	−.01	.00	−.02	−.11**	−.05	-
8. Sibling closeness T2	−.01	.01	.00	.03	−.08*	−.04	.42**	-
9. Sibling closeness T3	.03	.08*	.10**	−.01	−.07	−.01	.35**	.58**	-
10. Depression T1	.06*	.12**	.01	.12**	.12**	.10**	−.04	−.10**	−.01	-
11. Depression T2	.07*	−.01	−.02	.16**	.14**	.06	−.09**	−.16**	−.03	.33**	-
12. Depression T3	.07*	.11**	.07	.08*	.00	.08*	−.09**	−.10*	−.06	.26**	.50**	-
13. Risk Behaviors T1	.05	−.02	−.03	.14**	.14**	.09*	−.11**	−.10**	−.10**	.41**	.14**	.11**	-
14. Risk Behaviors T2	−.01	−.06	−.11**	.05	.02	−.08	−.20**	−.06	−.06	.03	.16**	.16**	.29**	-
15. Risk Behaviors T3	−.07*	.01	−.12**	.05	.01	−.06	−.22**	−.10*	−.10**	.04	.12**	.18**	.34**	.72**	-
M	3.99	3.45	3.47	4.44	4.17	4.30	3.33	3.56	3.62	16.81	13.14	12.97	1.42	1.50	1.45
SD	1.62	1.58	1.76	1.68	1.76	1.82	0.75	0.75	0.68	9.86	9.45	9.18	0.42	0.38	0.39

Note.

^*p < .05,

^**p < .01,

^***p < .001.

MR = mother-reported, FR = father-reported.

Multilevel Models

The results of the models on youth depressive symptoms are presented in Table 2. Positive associations were observed at the BP level for mother-reported marital conflict (B = 0.78, SE = 0.27, p < .01), but not father-reported marital conflict (B = 0.36, SE = 0.29, p = .11). No significant associations emerged at the WP level for either father report (B = −0.10, SE = 0.33, p = .76) or mother report (B = 0.01, SE = 0.32, p = .97). This BP effect indicates that, on average, youth reported higher levels of depressive symptoms when mothers reported higher levels of marital conflict, compared to other families. However, the association was not significant when parents reported higher-than-usual conflict at any given time.

Table 2.

Multilevel Model on Youth Adjustment.

	Depressive symptoms		Risk Behaviors
	Model 1 B (SE)	Model 2 B (SE)	B (SE)
Fixed effect
Intercept	16.63(1.12) ***	16.70(1.11) ***	1.35(0.05) ***
Covariates
Age	0.49(0.38)	0.48(0.38)	0.02(0.02)
Sex	−3.60(0.74) ***	−3.51(0.74) ***	0.14(0.03) ***
Birth order	1.23(1.30)	1.23(1.30)	0.01(0.05)
Nativity	0.29(0.84)	0.21(0.84)	0.05(0.04)
Dyad sex constellation	0.17(0.77)	0.08(0.77)	−0.02(0.04)
Sibling age gap	−0.13(0.28)	−0.15(0.28)	0.00(0.01)
Family socioeconomic status	−1.44(0.52) **	−1.43(0.52) **	0.01(0.02)
Predictors
WP FR marital conflict	−0.15(0.33)	−0.10(0.33)	−0.01(0.01)
WP MR marital conflict	0.03(0.32)	0.01(0.32)	0.00(0.01)
BP FR marital conflict	0.47(0.29)	0.36(0.29)	−0.01(0.01)
BP MR marital conflict	0.69(0.26) **	0.78(0.27) **	0.02(0.01) *
WP sibling closeness	−1.56(0.63) **	−1.65(0.63) **	0.03(0.02)
BS sibling closeness	0.45(1.08)	0.44(1.08)	0.08(0.04) †
BF sibling closeness	−2.58(0.72) ***	−2.59(0.72) ***	−0.07(0.03) **
Interactions
WP FR marital conflict * WP sibling closeness		1.03(0.84)
WP FR marital conflict * BF sibling closeness		−1.00(0.58)
BP FR marital conflict * WP sibling closeness		−0.39(0.43)
BP FR marital conflict * BF sibling closeness		1.10(0.49) *
Random effects
L2 intercept	19.54(5.55) ***	20.29(5.62) ***	0.04(0.01) ***
L3 intercept	7.41(4.63)	6.89(4.64)	0.02(0.01) **
Sibling Covariance (CS)	0.89(8.74)	1.13(8.71)	0.02(0.01)
Residual	60.08(9.37) ***	59.16(9.35) ***	0.07(0.01) ***

Note.

^†p < .06,

^*p < .05,

^**p < .01,

^***p < .001.

MR = mother-reported, FR = father-reported. WP = within-person effect, BS = between-sibling effect, BP = between-person effect, BF = between-family. For sex, 0 = female, 1 = male; for birth order, 0 = older sibling, 1 = younger sibling; for dyad sex constellation, 0 = mixed-sex, 1 = same-sex; for nativity, 0 = Mexico-born, 1 = U.S.-born.

Regarding sibling closeness, the analysis revealed two key main effects. First, sibling closeness was negatively associated with youth depressive symptoms at the BF level (B = −2.59, SE = 0.72, p < .001). This suggests that, on average, youth in families with higher sibling closeness reported fewer depressive symptoms compared to those in other families. Second, sibling closeness was negatively associated with depressive symptoms at the WP level (B = −1.65, SE = 0.63, p < .01). This indicates that, controlling for cross-time averages, on occasions when youth reported higher-than-usual sibling closeness than usual, they also reported fewer depressive symptoms than usual. Regarding the moderating role of sibling closeness, the interaction between BP father-reported marital conflict and BF sibling closeness was significantly associated with youth depressive symptoms (B = 1.10, SE = 0.49, p < .05). As shown in Figure 2, for families with lower sibling closeness, father-reported marital conflict was not linked to youth depressive symptoms, which consistently stayed high (B = 0.03, SE = 0.35, p = .94). In contrast, for families with higher sibling closeness, father-reported marital conflict was positively associated with youth depressive symptoms, on average (B = 0.72, SE = 0.31, p < .05); such that compared to other families, when fathers reported lower levels of marital conflict, on average, youth reported fewer depressive symptoms, on average. Additionally, sex and family socioeconomic status were each negatively correlated with youth depressive symptoms (B = −3.51, SE = 0.74, p < .001; B = −1.43, SE = 0.52, p < .01), with males and youth with higher SES more likely to report lower depressive symptoms.

Figure 2. The Moderating Role of Sibling Closeness between Father-reported Marital Conflict and Youth Depressive Symptoms.

Note. * p< 0.05.

The models examining youth risk behaviors are presented in Table 2. At the BP level, mother-reported marital conflict was positively associated with youth risk behaviors (B = 0.02, SE = 0.01, p < .05), indicating that, on average, higher levels of maternal conflict were linked to greater engagement in risk behaviors. In contrast, father-reported conflict was not significantly associated with youth risk behaviors at the BP level (B = −0.01, SE = 0.01, p = .49). At the WP level, neither father- nor mother-reported marital conflict was significantly associated with youth risk behaviors (father: B = −0.01, SE = .01, p = .60; mother: B = −0.002, SE = 0.01, p = .86).

Regarding the main effects of sibling closeness, it was negatively associated with youth risk behaviors at the BF level (B = −0.07, SE = 0.03, p < .05), indicating that, on average, youth from families with higher sibling closeness reported fewer risk behaviors compared to youth from other families with lower sibling closeness. No significant effects emerged at the WP level (B = 0.03, SE = 0.02, p = .20). At the BS level, sibling closeness was marginally associated with youth risk behaviors (B = 0.08, SE = 0.04, p < .06), suggesting a trend where, controlling for family average, siblings who reported higher sibling closeness also tended to report more risk behaviors than their siblings who reported lower closeness. However, sibling closeness did not moderate the association between marital conflict and youth risk behavior at any level (ps > .10). Additionally, youth sex was significantly associated with youth risk behaviors (B = 0.14, SE = 0.03, p < .001), with males reporting higher risk behaviors.

Sensitivity Analysis

Building on the combined model that simultaneously analyzed father- and mother-reported marital conflict as planned, separate models were conducted to examine the independent effects of father- and mother-reported marital conflict on youth development (see supplemental materials). For depressive symptoms, in the father-reported model, marital conflict was significantly associated with youth depressive symptoms at the BP level (B = 0.62, SE = 0.27, p < .05), differing from the combined model, where this association was not significant. At the WP level, the association remained nonsignificant (B = −0.05, SE = 0.32, p = .88), aligning with the combined model. The remaining results from the father- and mother-reported model, as well as sibling closeness findings, were consistent with those from the combined model. For risk behaviors, in the mother-reported model, marital conflict showed a trend-level association with youth risk behaviors at the BP level (B = 0.02, SE = 0.01, p = .06), slightly weaker than the significant association observed in the combined model. In the father-reported model, sibling closeness was associated with youth risk behaviors at the between-sibling (BS) level at trend-level (B = 0.08, SE = 0.04, p < .06), slightly weaker than the significant association in the combined model. Overall, the remaining pattern of findings mirrored those observed in the combined model.

Given evidence that sibling dynamics can differ by characteristics such as sibling sex (Zemp et al., 2021), dyad sex constellation (Updegraff et al., 2005), birth order (Whiteman et al., 2009), and age gap (Buist et al., 2013), we conducted follow-up analyses to examine whether these sibling and dyad factors further moderated the effect of BF sibling closeness on the BP father-reported marital conflict-adjustment associations. To do so, we added the relevant two- and three-way interaction terms. For example, in models assessing sibling sex, we added a three-way interaction between sibling sex and the marital conflict × sibling closeness term, as well as the corresponding two-way interactions between sibling sex and BP marital conflict and between sibling sex and BF sibling closeness. Across all models, none of the exploratory interactions involving sibling or dyad characteristics were significant for depressive symptoms or risk behaviors (ps > .10).

Discussion

Understanding how changes within families relate to changes in youth adjustment over time is central to family science, and essential for advancing knowledge on youth adjustment, with important implications for identifying when and for whom to intervene. By pinpointing both enduring family dynamics and patterns of change, research can better inform culturally sensitive prevention and intervention efforts that are responsive to families’ lived experiences. To our knowledge, the current study is among the first to explore both BF and WP associations between marital conflict, sibling closeness, and youth adjustment in Mexican-origin families, extending prior work using cross-sectional designs (e.g., Pereyra & Bean, 2017) or BF longitudinal designs (Martin et al., 2019). Findings revealed that the marital conflict-youth adjustment link (a) varied across BF and WP levels, (b) differed by parent reporter (mother vs. father), (c) was related to sibling closeness, and (d) depended on the specific youth outcome assessed. These results underscore the importance of considering developmental processes within their cultural context and highlight how culturally rooted family dynamics shape pathways of youth adjustment in Mexican-origin families.

The Association between Marital Conflict and Youth Adjustment

This study examined both fathers’ and mothers’ perspectives on marital conflict within a cultural context where two-parent families are predominant and familism values are strong (Chen & Guzman, 2022; Cahill et al., 2021). Regarding main effects, first, mother-reported marital conflict was associated with youth adjustment at the BP level, whereas father-reported conflict was not. Specifically, in families where mothers reported higher levels of marital conflict, on average, youth tended to report more depressive symptoms and engage in more risk behaviors compared to other families. This pattern aligns with gender role theory (Bartley et al., 2005) and emotional security theory (Davies & Cummings, 1994), and may also reflect culturally informed family dynamics in which mothers typically serve as primary caregivers and emotional anchors within Latino families (Kramer et al., 2021). Marital conflict experienced by mothers may more directly disrupt family interactions and children’s sense of emotional security. Mexican-origin fathers, who often occupy provider or authority roles, may express conflict-related distress less overtly, making it less salient to youth. Thus, mothers’ reports of marital conflict may better capture dynamics relevant to youth adjustment, consistent with evidence that Latino mothers spend more time in direct parenting than fathers.

Additionally, controlling for cross-time average scores, neither father- nor mother-reported marital conflict was associated with youth adjustment at the WP level. That is, on occasions when parents report more marital conflict than usual, youth did not report more adjustment problems than usual. This finding does not align with Zemp et al. (2018), but is more consistent with findings from Knopp et al.’s (2017) study on German families and Fosco et al. (2023), showing no direct WP association between marital conflict and depressive symptoms). This inconsistency may be related to several factors. First, differences in informants may contribute to the divergent findings. Knopp et al. (2017), Fosco et al. (2023), and the current study relied on children’s reports of adjustment outcomes and parents’ reports of marital conflict, whereas Zemp et al. (2018) relied exclusively on parent reports. This pattern is consistent with van Eldik et al.’s (2020) meta-analysis, showing that the association between marital conflict frequency and youth internalizing problems tends to be stronger when parents report both predictors and outcomes. Additionally, despite the absence of direct within-person effects, marital conflict may influence youth adjustment indirectly, such as through coparenting processes (Knopp et al., 2017) or children’s perceived threat (Fosco et al., 2023).

Furthermore, sampling difference and analytic strategy may also play a role. The current study is among the first to incorporate multiple children within the same family using a three-level modeling framework, accounting for observations nested within children and children nested within families. Including multiple siblings may provide a more comprehensive representation of family processes while also partitioning variance at the child and family levels, which could attenuate direct within-person associations if siblings differ in their perceptions or responses to marital conflict. Finally, cultural differences may also contribute to the discrepancy. The previous studies were conducted primarily with families from Caucasian cultural backgrounds, whereas the current study focuses on Mexican-origin families with an average marital duration of 17 years. Cultural emphases on strong family bonds and cohesion may reflect different family dynamics and pathways through which marital conflict influences children’s adjustment.

Together, the significant between-person association and the absence of within-person effects suggest that youth depressive symptoms are more closely linked to stable, between-family differences in marital conflict than to changes in conflict over the eight-year period in this sample. In other words, the overall family climate of conflict may have a stronger association with youth depressive symptoms than longer-term changes examined here. Considering both the BP association and the relatively high variability of marital conflict across families than within families, there might exist more stable patterns of marital conflict over time, which consistently impacts youth depressive symptoms. Given the cultural emphasis on cohesion and interdependence in Mexican-origin families, chronic marital conflict may be particularly harmful, undermining both emotional security and overall family functioning. Notably, this pattern did not emerge for youth risk behaviors, which may reflect protective cultural values such as familism and respeto that discourage engagement in such behaviors despite family conflict (Cahill et al., 2021). These findings underscore the importance of interventions that address persistent patterns of marital discord, enhance marital quality, and engage both parents in strengthening relationships within the broader cultural framework of family unity and mutual support.

At the same time, the present study’s broad measures of marital conflict may not fully capture the qualitative aspects, such as the topics, tone, or resolution strategies, through which conflict exerts its influence (van Eldik et al., 2020). Distinguishing between constructive versus destructive conflict, or between conflicts centered on parenting versus other domains, may reveal more nuanced pathways linking marital dynamics to youth adjustment (van Eldik et al., 2020). It should also be noted that some internal consistency indices of the marital conflict measure are modest (α = .67–.80), which may limit the precision of current findings. Future research should move beyond global indices of conflict frequency to examine these contextual and content-specific features, which likely shape how Mexican-origin youth experience and respond to family discord.

The Role of Sibling Closeness

Building on contemporary marital conflict research that highlights the need to identify mediating and moderating mechanisms (Davies & Cummings, 2006), this study extends previous cross-sectional and longitudinal BF studies (e.g., Davies et al., 2019; Gass et al., 2007) by examining the role of sibling closeness in youth adjustment at both the BF and WP level. Understanding these dynamics is essential for advancing family systems theory and designing culturally responsive interventions that consider the role of siblings. The results revealed that (a) sibling closeness was significantly linked to youth depressive symptoms at both the WP and BF levels, and (b) sibling closeness moderated the association between father-reported marital conflict and youth depressive symptoms at the BF level. Specifically, in families with high sibling closeness, lower marital conflict was associated with fewer depressive symptoms in youth; while in families with low sibling closeness, youth depressive symptoms remained high regardless of marital conflict levels. One possible explanation is that youth with low sibling closeness may not be emotionally attuned to family dynamics, limiting their ability to benefit from low-conflict home environments. Conversely, youth reporting high sibling closeness may experience stronger emotional connections within the family, and may be better able to take advantage of low family discord to protect against the development of depressive symptoms.

Put differently, the lowest levels of depressive symptoms were observed when marital conflict was low and sibling closeness was high. Conversely, high marital conflict, regardless of sibling closeness, was linked to elevated depressive symptoms. This aligns with the diminished stress-buffering effect proposed by Rueger et al. (2016), indicating that sibling closeness is less effective under high-conflict conditions. In such environments, parental conflict may overwhelm the emotional resources siblings can offer. On the other hand, in low-conflict families, sibling closeness appears to play a more active protective role. Supporting this idea, Keeton et al. (2015) found that high sibling companionship buffered the negative effects of parental stress on youth mental health. This relationship fosters a sense of security in the face of family stress, helping to protect youth from developing mental health issues like depression (Davies et al., 2019).

Only father-reported marital conflict was moderated by sibling closeness, a pattern that may reflect gendered family dynamics common in Mexican-origin families and highlights the often underexamined role of fathers (Perez-Brena et al., 2022). In these families, mothers typically assume primary caregiving responsibilities (Kramer et al., 2021), and their experiences of marital conflict may be more emotionally salient and closely tied to youth’ sense of security, leaving limited opportunity for sibling relationships to buffer these effects. In contrast, marital conflict reported by fathers may represent stress less embedded in daily emotional exchanges, allowing sibling closeness to serve as a stronger protective factor.

Overall, sibling closeness emerged as a culturally meaningful protective factor in reducing youth depressive symptoms (Killoren et al., 2021), but its effectiveness may depend on a low-conflict family environment where emotional support is more accessible. These findings echo the family systems perspective, highlighting the interdependence among marital, parental, and sibling subsystems in youth adjustment. In this cultural context, interventions should therefore consider the broader family context and multiple family subsystems, targeting not only individual outcomes but also elements such as sibling relationships. By efficiently integrating resources aligned with cultural values, programs can better support youth in navigating developmental challenges and promote their adjustment.

Limitations and Future Directions

This study makes important contributions to the literature on family dynamics and enhances our understanding of the critical role of sibling relationships in youth adjustment within Mexican-origin families. However, these findings should be interpreted in light of several limitations. First, although the analytic approach allowed for a nuanced examination of multiple levels of moderation, the study’s power to detect small interaction effects was limited. The original project’s power analysis targeted small to medium main effects and may not have provided sufficient power for detecting multilevel interactions. Consequently, the moderation findings should be considered preliminary, and replication with larger samples is needed to confirm these effects. Second, the family system consists of various interrelated individuals and subsystems, and perceptions of marital conflict may differ between youth and parents (Ehrlich et al., 2019). The effects of marital conflict may also differ based on youth reports, making it important to include multiple reporters to further explore the interplay of their experiences. Third, while this study assessed the frequency and severity of marital conflict, its content and resolution may also impact youth outcomes (van Eldik et al., 2020). For example, a meta-analysis showed that child-related conflict has been shown to predict maladjustment more strongly than general conflict, suggesting a need for more nuanced measures (van Eldik et al., 2020).

Additionally, our study, utilizing three time points across eight years, may not capture the full dynamics of family processes that can vary across timescales. For example, Gao and Cummings (2019) analyzed daily reports over 15 days each year for 3 years and found that greater marital distress was associated with lower parent-child relationship quality on the same day but higher quality the following day. Such daily effects diminished as children entered adolescence. Thus, future research should examine marital conflict and youth adjustment across varying temporal intervals, such as daily, weekly, monthly, and annually. Finally, as our data were collected two decades ago, the processes observed here may not reflect current sociohistorical contexts. Even within the original data collection period (2002–2010), broader social and economic changes may have influenced family dynamics. For example, greater access to digital communication and social media may have begun to reshape how adolescents connect with peers and family, potentially influencing both sibling closeness and parent-adolescent communication patterns. More recently, during and after the COVID-19 pandemic, marked by increased stress and isolation, there can be different patterns of sibling dynamics (Zhou et al., 2024).

Conclusion

Overall, findings from this multilevel, longitudinal, and multi-informant study provide valuable insights into family processes among Mexican-origin families by integrating both mothers’ and fathers’ perspectives on marital conflict, highlighting the often-overlooked role of sibling closeness, and examining multiple dimensions of youth adjustment. By analyzing both BF and WP effects, our findings provide a nuanced understanding of the intricate dynamics within and between families, particularly in Latino families, a population often underrepresented in family research (Perez-Brena et al., 2022). This study emphasizes the importance of sibling relationships in promoting youth well-being within a context of strong family values and positions siblings as a potential resource for resilience. Strengthening these bonds could be an effective focus for family-based interventions. More broadly, this research illustrates the value of examining interconnected family subsystems, marital, parental, and sibling, within Mexican-origin families to inform culturally grounded approaches that support youth adjustment and mental health.