Sibling relationships as a mediator and a moderator between parental attachment and depression among adolescents Authors: Beibei Kuang, Yujing Zhao, Baojing Zhang, Jue Deng, Huahua Hu, Tong Ge Published: 2026 ABSTRACT Depression threatens adolescents’ psychological health and is influenced by negative interpersonal relationships. Previous studies have explored the effect of a negative parent−child relationship on adolescent depression but have ignored the frequent coexistence of these relationships with other family relationships, for example, sibling relationships. This study examined how parent−child and sibling relationships co-impact adolescent depression, by testing the mediating and moderating roles of sibling relationships between parental attachment and adolescent depression. A sample of 5,896 adolescents completed the Inventory of Parent and Peer Attachment, Sibling Relationship Questionnaire, and the Patient Health Questionnaire. 3,054 participants (1,506 girls) were included in the analysis (mean age = 13.70 ± 0.77 years). We found that the level of parental attachment security was negatively associated with adolescent depression, and sibling relationships not only mediated but also moderated this association. Regarding the moderating effect, only one dimension of sibling relationships (warmth) showed a significant moderating role. Specifically, sibling warmth buffered the negative association between maternal (but not paternal) attachment security and adolescent depression. Our results suggest that both negative parent–child and sibling relationships are possible risk factors for adolescent depression, and a positive sibling relationship can be a protective factor against the effect of negative parent–child relationships on adolescent depression. The present study also provides new insights for interventions adolescent depression. Sibling relationships deserve more attention in future research. Background Depression is characterized by periods of low mood or sadness and a loss of pleasure or interest in previously enjoyed activities [1]. As one of the most prevalent mental disorders [2, 3] and the second leading cause of disability worldwide [4], depression has become a major public health concern, especially for adolescents. According to a recent global meta-analysis, 34% of adolescents aged 10–19 years old were at risk of developing clinical depression between 2001 and 2020 [5]. Depression impairs adolescents’ mental health and leads to serious social problems. Although biological factors are important contributors to the disorder, researchers suggest that negative interpersonal relationships also contribute to the prevalence of depression [6]. Attachment theory provides a useful framework for understanding negative interpersonal relationships and their effects on depression. Parental attachment refers to an affectionate bond that develops between individuals and their primary caregivers through interpersonal interactions in early life [7], and influences future interpersonal functions and methods of regulating distress [8]. Internal working models (IWMs) of the self and others are theoretical constructs that capture individual differences in how individuals respond to interpersonal events and distress [9]. If caregivers are unresponsive and insensitive to the needs of individuals, individuals develop a negative IWM of the self and others, reflecting low attachment security. Individuals with negative IWMs consider themselves worthless and unlovable (self-model), and consider others untrustworthy and unavailable (others-model); therefore, they develop negative cognitive, emotional, and behavioral patterns of dealing with interpersonal events and distress [10]. They are more vulnerable to depression. The negative link between the level of parental attachment security and depression has been confirmed by many studies [11, 12]. However, if caregivers are responsive and sensitive to their needs, they form a positive IWM. Previous researchers mainly cared about the effect of the parent−child relationship on adolescent depression. However, the parent−child relationship often coexists with other family relationships, for example, sibling relationships. It is unclear how multiple family relationships jointly influence adolescent depression. The present study explored how parent−child and sibling relationships jointly influence adolescent depression, by testing the mediating role of sibling relationships in the link between parent−child relationships and adolescent depression. Sibling relationships are unique in that they are the longest-lasting relationships in the lives of most individuals. Although sibling relationships change throughout life, siblings are still ranked as significant sources of social support by adolescents [13]. Moreover, siblings are not only companions who share love and warmth, but also competitors in relationships characterized by conflict and rivalry [14]. So the role of sibling relationships in individual development remains an intriguing and vital research question. In Chinese culture, the sibling bond is metaphorically referred to as “hands and feet” (shouzu), signifying its profound and enduring significance. A well-known proverb states that “an elder brother is like a father, and an elder sister is like a mother,” which vividly illustrates the influential role of sibling relationships in Chinese culture. Considering the uniqueness and significance of sibling relationships, researchers began conducting studies on them over 100 years ago [15]. However, sibling relationships have been understudied relative to other family relationships. Furthermore, previous studies have primarily been conducted in Western countries, and hence, further studies in non-Western cultures are required to enrich the current literature. This study explored the role of sibling relationships between the parental attachment and adolescent depression in the Chinese culture. Family systems theory proposes that the family operates as an integrated, interconnected system in which family members dynamically interact and influence each other [16]. In addition to the parent−child relationship, sibling relationships may also influence individual development [17]. In reality, individuals typically form a parent–child bond first, followed by sibling relationships. Thus, parental attachment is considered an antecedent of sibling relationships, which, in turn, are jointly linked to adolescent emotional adjustment. In fact, evidence from longitudinal studies has shown that negative sibling relationships are associated with depressive symptoms from adolescence to emerging adulthood, and negative sibling relationships in childhood predict major depression in adulthood [18, 19]. In addition, from the attachment perspective, the others-model of IWM always acts as a blueprint for individuals to cope with other interpersonal relationships; the emotional tone and experiences of parent–child interactions can be transmitted into the sibling system [20]. So that the parent–child relationship may affect sibling relationships. Previous studies did find this affection [21, 22]. Based on the above theory and evidence, we can reason that sibling relationships may play a mediating role between parental attachment and depression. Unfortunately, no studies have explored this issue. Therefore, this study tested the mediating role of sibling relationships. The following hypothesis is proposed:Hypothesis 1: Parental attachment security was negatively associated with adolescent depression via sibling relationships. Hypothesis 1: Parental attachment security was negatively associated with adolescent depression via sibling relationships. Hypothesis 1: Parental attachment security was negatively associated with adolescent depression via sibling relationships. According to attachment theory, in addition to their primary caregivers, children can form emotional ties to familiar others in their contexts, including their siblings [23]. Older siblings may provide support for younger siblings when parents are absent or the parent−child relationship is negative. For example, older siblings may perform increased caregiving behaviors towards their younger siblings when their parents work overtime. Furthermore, social support theory has proposed that social resources play a critical role in buffering stress and protecting individuals from psychological distress [24]. Although parental attachment is the primary source of emotional support for adolescents, sibling relationships constitute another important support system. When parental support is insufficient or insecure, positive sibling relationships can serve as an alternative protective resource, compensating for deficits in parent−child attachment and reducing the risk of depression. Previous studies have suggested that sibling relationships play a protective role in the link between parental psychopathology and children’s psychological health [25, 26]. However, no research has explored the role of sibling relationships in the association between parental attachment and depression. The second aim of this study was to investigate this issue. The following hypothesis was proposed:Hypothesis 2. Sibling relationships moderate the relationship between parental attachment security and adolescent depression. Hypothesis 2a. Positive sibling relationships buffer the link between parental attachment security and adolescent depression. Hypothesis 2b. Negative sibling relationships exacerbate this association. Hypothesis 2. Sibling relationships moderate the relationship between parental attachment security and adolescent depression. Hypothesis 2. Sibling relationships moderate the relationship between parental attachment security and adolescent depression. Hypothesis 2a. Positive sibling relationships buffer the link between parental attachment security and adolescent depression. Hypothesis 2a. Positive sibling relationships buffer the link between parental attachment security and adolescent depression. Hypothesis 2b. Negative sibling relationships exacerbate this association. Hypothesis 2b. Negative sibling relationships exacerbate this association. Depression is common, frequently recurrent, and associated with a high risk of suicide. Therefore, it is essential to prioritize this issue. Negative interpersonal relationships heighten the risk of subsequent depression, according to interpersonal theories of depression [27, 28]. However, previous studies have mainly focused on the effect of parent–child relationship on depression, ignoring the important role of sibling relationships. As the number of multi-child families increases due to changes in China’s birth policy, sibling relationships have become an important factor influencing individual development, family harmony, and social stability. Furthermore, Confucian cultural values, including familial obligation, filial piety, and the expectation that siblings (especially sons) mutually support each other and share in caring for aging parents, highlight the unique and influential role of sibling relationships in Chinese families. The present study focused on how parent–child and sibling relationships jointly affect adolescent depression. Particular attention was paid to how sibling relationships play a protective role in the association of parent–child relationships and adolescent depression. The findings could help adolescents with depression find ways to overcome it. Methods Participants Participants were 5,896 adolescents from a middle school in Shandong Province, China. After completing the participants exclusion procedure, 3,054 (1,506 girls) participants were included in the analysis. The average age of the participants was 13.70 ± 0.77 years with a range of 12–16. A total of 265 participants did not provide their age, and their missing ages were imputed using mean substitution. Participants were excluded because: They did not finish all scales (116 participants finished only the demographic variables items; 950 participants completed the demographic variables and the first scale, 103 participants completed the demographic variables and the first two scales, 58 participants completed the demographic variables and the second scale) (n = 1,227).They did not answer the attention check items correctly. These items consisted of three statements with obvious answers: “For this item, please select the option ‘Somewhat’ ”; “For this item, please directly select ‘Completely true’ ”; “What is the capital of China?” (n = 640; n = 360; n = 47).Participants chose C, D or E in the self-reported attentiveness check item: “What is your attitude toward completing this questionnaire? A = Very conscientious, B = Relatively conscientious, C = Neutral, D = Relatively unconscientious, E = Very unconscientious.” (n = 113).1 Participants were only-children (n = 455). They did not finish all scales (116 participants finished only the demographic variables items; 950 participants completed the demographic variables and the first scale, 103 participants completed the demographic variables and the first two scales, 58 participants completed the demographic variables and the second scale) (n = 1,227). They did not finish all scales (116 participants finished only the demographic variables items; 950 participants completed the demographic variables and the first scale, 103 participants completed the demographic variables and the first two scales, 58 participants completed the demographic variables and the second scale) (n = 1,227). They did not answer the attention check items correctly. These items consisted of three statements with obvious answers: “For this item, please select the option ‘Somewhat’ ”; “For this item, please directly select ‘Completely true’ ”; “What is the capital of China?” (n = 640; n = 360; n = 47). They did not answer the attention check items correctly. These items consisted of three statements with obvious answers: “For this item, please select the option ‘Somewhat’ ”; “For this item, please directly select ‘Completely true’ ”; “What is the capital of China?” (n = 640; n = 360; n = 47). Participants chose C, D or E in the self-reported attentiveness check item: “What is your attitude toward completing this questionnaire? A = Very conscientious, B = Relatively conscientious, C = Neutral, D = Relatively unconscientious, E = Very unconscientious.” (n = 113).1 Participants chose C, D or E in the self-reported attentiveness check item: “What is your attitude toward completing this questionnaire? A = Very conscientious, B = Relatively conscientious, C = Neutral, D = Relatively unconscientious, E = Very unconscientious.” (n = 113).1 Participants were only-children (n = 455). Participants were only-children (n = 455). Listwise deletion was applied to missing data, considering: A total of 1,124 participants missed two of three study variables, and 103 missed one of three study variables.Little’s missing completely at random (MCAR) test was conducted to examine missingness patterns. Results indicated that the MCAR assumption was violated for depressive symptoms (χ2 = 123.963, p < .05), whereas the assumption held for parental attachment and sibling relationships (both χ2 = 0.213, p = .975). Although the MCAR test was significant for depressive symptoms, this pattern is typical in mental health surveys, as less distressed adolescents are more likely to disengage. Importantly, missingness was completely random for parental attachment and sibling relationships. A total of 1,124 participants missed two of three study variables, and 103 missed one of three study variables. A total of 1,124 participants missed two of three study variables, and 103 missed one of three study variables. Little’s missing completely at random (MCAR) test was conducted to examine missingness patterns. Results indicated that the MCAR assumption was violated for depressive symptoms (χ2 = 123.963, p < .05), whereas the assumption held for parental attachment and sibling relationships (both χ2 = 0.213, p = .975). Although the MCAR test was significant for depressive symptoms, this pattern is typical in mental health surveys, as less distressed adolescents are more likely to disengage. Importantly, missingness was completely random for parental attachment and sibling relationships. Little’s missing completely at random (MCAR) test was conducted to examine missingness patterns. Results indicated that the MCAR assumption was violated for depressive symptoms (χ2 = 123.963, p < .05), whereas the assumption held for parental attachment and sibling relationships (both χ2 = 0.213, p = .975). Although the MCAR test was significant for depressive symptoms, this pattern is typical in mental health surveys, as less distressed adolescents are more likely to disengage. Importantly, missingness was completely random for parental attachment and sibling relationships. Measures The inventory of parent and peer attachment Parental attachment security was assessed using the Chinese version of the Inventory of Parent and Peer Attachment (IPPA), originally developed by Armsden and Greenberg [29, 30], and was adapted and validated in Chinese by Zhang et al., with adequate reliability and validity indices reported [31]. The questionnaire consists of three sections that separately assess the perceived quality (security) of attachment to mothers, fathers, and peers, yielding three attachment scores. Each section includes 25 items, such as “My mother respects my feelings,” and is broken down into three dimensions (communication, trust, and alienation). Participants rated the items on a 5-point scale (1 = Almost never true to 5 = Almost always true). The total score for each section is communication + trust − alienation, with scores ranging from 25 to 125. Higher scores reflect greater perceived attachment (high attachment security) than lower scores. This study measured adolescents’ attachment to their mothers and fathers. Total scores for attachment to the mother and father were used in our analysis. In the present study, the Cronbach’s alpha for mother and father attachment were 0.81 and 0.80, respectively. The sibling relationship questionnaire Sibling relationships were assessed using the Chinese version of the Sibling Relationship Questionnaire (SRQ), originally developed by Furman and Buhrmester [32], and was adapted and validated in Chinese by Dai et al., with adequate reliability and validity indices reported [33]. The SRQ includes 36 items such as “How often do you and this sibling do nice things for each other?” and consists of three dimensions: warmth/closeness (21 items), conflict (9 items), and rivalry (6 items). For warmth and conflict, participants responded to each item on a five-point scale ranging from 1 = hardly at all to 5 = all the time. The rivalry dimension was scored in two ways. When assessing the degree of overall parental favoritism, items were scored as follows: Options 1 and 5 were assigned 2 points, Options 2 and 4 were assigned 1 point, and Option 3 was assigned 0 points. This scoring procedure was recommended by the author to facilitate comparisons with other studies. In contrast, when examining which child the parent favored more specifically, the items were scored in a similar manner to the other two dimensions. The subscores of warmth and conflict range from 1 to 5, while the subscore of rivalry ranges from 0 to 2. The subscores of the three dimensions were used in the present study, while the total SRQ score was not computed [32]. In the present study, the Cronbach’s alpha for warmth/closeness, conflict, and rivalry were 0.95, 0.83, and 0.89, respectively. The patient health questionnaire Depression was assessed using the Chinese version of the Patient Health Questionnaire (PHQ-9), originally developed by Löwe, Kroenke, Herzog and Gräfe [34], and was adapted and validated in Chinese by Hu et al., with adequate reliability and validity indices reported [35]. This questionnaire includes nine items corresponding to nine depressive symptoms. Participants assess how often, over the past two weeks, they have been affected by each depressive symptom. Response options are “not at all,” “several days,” “more than half the days,” and “nearly every day,” scored as 0, 1, 2 and 3, respectively. The PHQ-9 score ranges from 0 to 27, with scores of 0–4, 5–9, 10–14, 15–19, and 20–27 representing typically minimal, mild, moderate, moderately severe, and severe levels of depression, respectively. In the present study, the Cronbach’s alpha for the PHQ-9 was 0.91. Procedure This research project was approved by the Ethics Committee of the authors’ institution. Permission to collect the data was obtained from the school’s principal, and written informed consent was obtained from parents or legal guardians and from students before participation. Data were collected by computer in the classroom. Teachers administered the survey and explained the questionnaire procedure using standard instructions. Participants were informed that they could withdraw from the study at any time and that the survey was anonymous, with the collected data used only for research purposes. Results Preliminary analysis Descriptive statistics showed that skewness values ranged from 0.52 to 1.99, which is consistent with typical distributions of mental health indicators. Kurtosis values ranged from − 0.51 to 1.06 for all variables except rivalry. Rivalry showed a slightly higher kurtosis (3.38), reflecting a mildly leptokurtic distribution that is acceptable given the large sample size. After controlling for sex, age, number of children, and birth order, pearson correlations were calculated (Table 1) and the results revealed that attachment to the mother (AttMo) and father (AttFa) were both positively correlated with warmth and negatively linked to conflict, rivalry, and depression. Depression was negatively associated with warmth and positively related to conflict and rivalry. Table 1Pearson correlations among attachment, sibling relationship, and depressionM(SD)12345671 AttMo97.74 (20.09)12 AttFa93.56 (20.74)0.75***13 Warmth3.57 (0.83)0.47***0.49***14 Conflict1.99 (0.69)− 0.32***− 0.35***− 0.34**15 Rivalry0.28 (0.50)− 0.30***− 0.31***− 0.29***0.26***16 Depression5.56 (5.92)− 0.53***− 0.55***− 0.33***0.31***0.26***1N = 3,054. ***p<.001, **p<.01 Table 1Pearson correlations among attachment, sibling relationship, and depressionM(SD)12345671 AttMo97.74 (20.09)12 AttFa93.56 (20.74)0.75***13 Warmth3.57 (0.83)0.47***0.49***14 Conflict1.99 (0.69)− 0.32***− 0.35***− 0.34**15 Rivalry0.28 (0.50)− 0.30***− 0.31***− 0.29***0.26***16 Depression5.56 (5.92)− 0.53***− 0.55***− 0.33***0.31***0.26***1N = 3,054. ***p<.001, **p<.01 Pearson correlations among attachment, sibling relationship, and depression N = 3,054. ***p<.001, **p<.01 The mean score of the PHQ-9 was broken down into 1,614 (52.8%) participants with typically minimal, 753 (24.7%) with mild depression, 421 (13.8%) with moderate depression, 155 (5.1%) with moderately severe depression, and 111 (3.6%) with severe depression. Mediation analysis The mediating role of sibling relationships was tested using the PROCESS Macro Model 4 in SPSS with 5,000 bootstrapped samples [36]. Attachment was the independent variable, depression the dependent variable, and the three dimensions (warmth, conflict, and rivalry) of sibling relationships the mediating variables. All variables were standardized. Sex, age, number of children, and birth order were controlled for. AttMo and AttFa models were run separately. The results showed that AttMo was negatively linked to depression (β = − 0.53, p < .001; R²=0.30, F = 261.78, p < .001). AttMo was positively linked to warmth (β = 0.47, p < .001; R²=0.22, F = 174.14, p < .001) and negatively linked to conflict (β = − 0.32, p < .001; R²=0.12, F = 80.38, p < .001) and rivalry (β = − 0.30, p < .001; R²=0.09, F = 61.39, p < .001). When we added warmth, conflict, and rivalry to the model, AttMo was still negatively linked to depression (β = − 0.44, p < .001), warmth was negatively linked to depression (β = − 0.06, p < .001), and conflict and rivalry were positively linked to depression (β = 0.13, p < .001; β = 0.07, p < .001), R²=0.33, F = 187.69, p < .001. Thus, sibling relationships mediated the effect of AttMo on depression. The total, direct, and indirect effects are presented in Table 2. Table 2Mediation effect of sibling relationships in the association between AttMo and depressionPathEffect95% CIRatioDirect effectAttMo-depression−0.44−0.47, − 0.400.83Indirect effectAttMo-warmth-depression−0.03−0.05, − 0.010.05Indirect effectAttMo-conflict-depression−0.04−0.05, − 0.030.08Indirect effectAttMo-rivalry-depression−0.02−0.03, − 0.010.04Total indirect effect−0.09−0.11, − 0.070.17Total effect−0.53−0.56, − 0.501.00 The results also revealed that AttFa was negatively associated with depression (β = − 0.55, p < .001; R²=0.32, F = 292.12, p < .001). AttFa was positively associated with warmth (β = 0.49, p < .001; R²= 0.24, F = 190.80, p < .001) and negatively associated with conflict (β = −0.35, p < .001; R²= 0.14, F = 97.04, p < .001) and rivalry (β = −0.31, p < .001; R²= 0.10, F = 65.94, p < .001). When we added warmth, conflict, and rivalry to the model, AttFa was still negatively associated with depression (β = −0.47, p < .001). Warmth was negatively associated with depression (β = −0.05, p < .001), whereas conflict and rivalry were positively associated with depression (β = 0.11, p < .001; β = 0.07, p < .001), R = 0.59; R² = 0.35, F = 200.62, p < .001. Thus, sibling relationships mediated the effect of AttFa on depression. The total, direct, and indirect effects are presented in Table 3. Table 2Mediation effect of sibling relationships in the association between AttMo and depressionPathEffect95% CIRatioDirect effectAttMo-depression−0.44−0.47, − 0.400.83Indirect effectAttMo-warmth-depression−0.03−0.05, − 0.010.05Indirect effectAttMo-conflict-depression−0.04−0.05, − 0.030.08Indirect effectAttMo-rivalry-depression−0.02−0.03, − 0.010.04Total indirect effect−0.09−0.11, − 0.070.17Total effect−0.53−0.56, − 0.501.00 Mediation effect of sibling relationships in the association between AttMo and depression Table 3Mediation effect of sibling relationships in the association between AttFa and depressionPathEffect95% CIRatioDirect effectAttFa-depression−0.47−0.50, − 0.440.85Indirect effectAttFa-warmth-depression−0.02−0.04, 0.000.04Indirect effectAttFa-conflict-depression−0.04−0.05, − 0.030.07Indirect effectAttFa-rivalry-depression−0.02−0.03, − 0.010.04Total indirect effect−0.08−0.11, − 0.060.15Total effect−0.55−0.58, − 0.521.00AttMo Attachment to Mother, AttFa Attachment to Father Table 3Mediation effect of sibling relationships in the association between AttFa and depressionPathEffect95% CIRatioDirect effectAttFa-depression−0.47−0.50, − 0.440.85Indirect effectAttFa-warmth-depression−0.02−0.04, 0.000.04Indirect effectAttFa-conflict-depression−0.04−0.05, − 0.030.07Indirect effectAttFa-rivalry-depression−0.02−0.03, − 0.010.04Total indirect effect−0.08−0.11, − 0.060.15Total effect−0.55−0.58, − 0.521.00AttMo Attachment to Mother, AttFa Attachment to Father Mediation effect of sibling relationships in the association between AttFa and depression AttMo Attachment to Mother, AttFa Attachment to Father In summary, parental attachment security was negatively associated with adolescent depression, and this association was mediated by sibling relationships. Therefore, Hypothesis 1 was supported. Moderation analysis To test the moderating role of sibling relationships in the association between attachment and depression, we used the PROCESS macro (Model 1) in SPSS [36] with 5,000 bootstrap samples. Attachment was the independent variable, depression the dependent variable, and sibling warmth the moderating variable. Sibling conflict and sibling rivalry were not included in the model because pre-analysis showed that neither the interaction of AttMo × conflict/rivalry nor the interaction of AttFa × conflict/rivalry predicted depression (all β > −0.03, p > .0125).2 All variables were standardized. Sex, age, number of children, and birth order were controlled for. AttMo and AttFa models were run separately. Regarding AttMo, the overall model was statistically significant, explaining 31.28% of the variance in depression (R2 = 0.31, F = 198.06, p < .001). The results showed that AttMo and warmth were negatively correlated with depression (β = −0.47, p < .001; β = −0.10, p < .001), and the interaction of AttMo × warmth was positively correlated with depression (β = 0.06, p < .001; ΔR²= 0.004, F = 16.40, p < .001). Regarding AttFa, the overall model was statistically significant, explaining 31.28% of the variance in depression (R2 = 0.33, F = 214.42, p < .001). Results showed that AttFa and warmth were negatively correlated with depression (β = −0.51, p < .001; β = −0.08, p < .001), and the interaction of AttFa × warmth was unrelated to depression (β = 0.03, p = .032; ΔR²= 0.001, F = 4.58, p = .032).2 This suggests that the relationship between maternal (but not paternal) attachment security and adolescent depression was moderated by sibling warmth. Moderation analysis were performed by testing the conditional effect of attachment on depression at three moderator value points: at 1 SD above the sample mean (high warmth), at the sample mean (medium warmth), and at 1 SD below the sample mean (low warmth). Detailed results are presented in Table 4; Fig. 1, which show that warmth weakened the negative association between attachment and depression. In other words, the stronger the warmth, the weaker the association between attachment and depression. Table 4Conditional effects of attachment on depression at three levels of warmthWarmthEffect95% CI p AttMo−1.000 (low)−0.5285−0.5701, − 0.4870< 0.0010.0000 (mean)−0.4727−0.5062, − 0.4391< 0.0011.0000 (high)−0.4168−0.4614, − 0.3722< 0.001 Table 4Conditional effects of attachment on depression at three levels of warmthWarmthEffect95% CI p AttMo−1.000 (low)−0.5285−0.5701, − 0.4870< 0.0010.0000 (mean)−0.4727−0.5062, − 0.4391< 0.0011.0000 (high)−0.4168−0.4614, − 0.3722< 0.001 Conditional effects of attachment on depression at three levels of warmth Fig. 1The moderating role of sibling relationships in the association between parental attachment and adolescent depression. Note. ATTMo = Attachment to Mother; ATTFa = Attachment to Father The moderating role of sibling relationships in the association between parental attachment and adolescent depression. Note. ATTMo = Attachment to Mother; ATTFa = Attachment to Father Discussion The present study examined how parent−child and sibling relationships jointly influence adolescent depression by testing the mediating and moderating roles of sibling relationships in the association between parental attachment and adolescent depression. The results revealed that parental attachment security was negatively associated with adolescent depression. The negative effect of parental attachment security on adolescent depression was mediated and moderated by sibling relationships. However, we just found the moderating effect of sibling warmth (but not sibling conflict and rivalry), and this effect was significant only for maternal (but not paternal) attachment. That is, parental attachment and sibling relationships jointly have a negative effect on adolescent depression. Parental attachment affects adolescent depression through the mediating role of sibling relationships; only sibling warmth buffers the negative effect of insecure maternal–child (not paternal−child) attachment on adolescent depression. The mediating role of sibling relationships The findings confirmed that the level of parental attachment security was negatively correlated with adolescent depression, consistent with the results of previous studies [11, 12]. These results support the plausibility that insecure parental attachment increases depressive symptoms. According to attachment theory [37–39], caregivers’ insensitive responses to children shape negative parent–child connections, which makes children form insecure (less secure) attachments to parents and negative IWM. Specifically, if primary caregivers are insensitive to children, children may internalize the belief that the self is unlovable and valueless (negative self-model) and/or that others are untrustworthy and unreliable (negative other-model). These negative beliefs are gradually strengthened by caregivers’ insensitive responses and develop into negative mental representations, and even negative thinking patterns. These representations function as templates guiding expectations and behaviors in subsequent close relationships. Feelings of worthlessness and negative mental representations are closely linked to depression. Thus, these negative cognitive representations mediate the link between parental attachment and subsequent depressive symptoms. Our results revealed that parental attachment security was positively correlated with sibling warmth and negatively correlated with sibling conflict and rivalry. These findings indicate that the quality of parental attachment is linked to the quality of sibling relationships, consistent with previous studies [40–42]. The concept of the IWM in attachment theory provides possible explanations for our findings. Negative parent–child interaction experiences in early life form negative IWMs. The IWM provides an internal schema used to cope with stress and interactions in other relationships throughout life [43, 44]. In the IWM, a negative other-model causes individuals to distrust others, such as siblings because the parent–child interaction model is replicated in the interaction between siblings. Therefore, individuals with negative IWMs are more likely to develop negative sibling relationships. This explains why attachment quality was positively linked to sibling warmth and negatively linked to sibling conflict and sibling rivalry in the present study. We also found that the quality of sibling relationships was negatively associated with depression, consistent with the findings of previous studies [14]. This result can be explained by attachment and social comparison theories. From an attachment perspective, hostile and conflictual sibling relationships reinforce negative internal working models, by intensifying feelings of low self-worth and consolidating maladaptive cognitive schemas, and then increase depressive vulnerability. Within the stress–coping framework, negative sibling interactions constitute a chronic interpersonal stressor, which may deplete coping resources and heightens emotional exhaustion, thereby increasing depressive symptoms over time. Furthermore, siblings naturally compare themselves with each other because they not only share but also compete for family resources and parental love. According to social comparison theory [45], individuals compare themselves with others to maintain high self-esteem. Siblings develop low self-esteem when they perceive their siblings are stronger or gain more family resources, which causes conflict and rivalry. In fact, conflict and rivalry are external expressions of low self-esteem, which is one of the main causes of depression. This explains why conflict and rivalry were positively associated with depression in the present study. In summary, the quality of parental attachment was negatively associated with adolescent depression through the mediating role of sibling relationships. These results support the proposed mediation pathway. The moderating role of sibling relationships We also detected that sibling warmth moderated the relationship between maternal attachment security and adolescent depression, consistent with results found among emerging adults [46]. For adolescents with higher sibling warmth, the negative link between maternal attachment security and depression was weaker than for those reporting lower sibling warmth. These results show that positive sibling relationships possibly buffer the negative effect of insecure maternal–child relationships on adolescent depression, providing new evidence for this research area. From the perspective of social support theory [24], individuals with warm (more positive) sibling relationships can gain more social support, whereas those with less positive sibling relationships often have less social support. For individuals with insecure parental attachment, social support can compensate for decreased self-value, while for those with secure attachment, social support can strengthen their existing self-value. From the perspective of attachment theory, warm, supportive sibling bonds can provide an alternative source of security, counteracting feelings of worthlessness and reducing depressive risk [47, 48]. Such relationships may also partially modify the negative other-model by offering experiences of reliable, responsive connection with others. From the perspective of stress and coping theory [49], positive sibling relationships may enhance individuals’ capacity to manage stress by providing opportunities for modeling adaptive coping strategies, thereby exerting a buffering effect against maladaptive internal working models and depressive symptoms. Following a correction for multiple testing, the moderating effect involving paternal attachment was no longer statistically significant. One possible explanation concerns traditional Chinese families' operational patterns, where mothers are typically more central to daily caregiving and emotional support, whereas fathers are often perceived as primarily responsible for economic provision and discipline rather than intimate emotional bonding. This cultural division of family roles may lead to maternal attachment having a more consistent and robust association with adolescent mental health and sibling relationship dynamics, while the influence of paternal attachment is comparatively weaker and less stable, rendering its effects non-significant after strict control for Type I error inflation. These findings highlight the importance of considering gendered parenting roles in Chinese culture and suggest that maternal attachment may carry greater weight in shaping sibling relationships. Conclusions and limitations The present study found that parental attachment and sibling relationships jointly influenced adolescent depression, by demonstrating that sibling relationships not only mediate but also moderate the relationship between parental attachment and adolescent depression. We only found a moderating effect of sibling warmth (but not sibling conflict and rivalry), and this effect was significant only for maternal (but not paternal) attachment. That is, sibling warmth buffered the negative association between maternal attachment security and adolescent depression. This finding suggests that negative parent−child and sibling relationships may both have adverse effects on adolescent depression, whereas positive sibling relationships may mitigate this negative effect. These findings provide new scope for interventions adolescent depression. They indicate that siblings can be included in the support system of adolescents with depression, especially when parents are unable or unwilling to provide support. However, these conclusions should be interpreted with caution, considering the following limitations. First, the present study was conducted among adolescents. This process may be different for adults or younger children because sibling relationships change with age [18]. Second, this study was conducted in the Chinese cultural context. These results may differ in other cultures. Third, while the proposed mediation model is theory-driven and consistent with developmental accounts of adolescent mental health, the cross-sectional design limits causal inference. Temporal ordering cannot be confirmed, and alternative or reciprocal pathways remain plausible. Fourth, the single-school sample may limit generalizability. Student clustering within classes and teacher-administered surveys may introduce clustering effects and social desirability bias. Future studies could use multilevel modeling or robust standard errors to address clustering. Fifth, all variables were assessed using self-report measures, which may introduce concerns regarding common method variance and reporter bias. Finally, sibling age spacing, gender composition, and socioeconomic status should be considered in future sibling research.