Loneliness refers to an individual’s subjective sense of isolation resulting from unmet interpersonal relationship needs, whether in terms of quality or quantity [6]. Loneliness has various adverse effects on individuals, including accelerated aging [7,8] and increased mortality [9] in terms of physical health. It also has predictive negative effects on mental health, such as personality disorders [10] and mental disorders [11]. A recent report on global adolescent, loneliness reveals a significant surge among adolescents worldwide since 2010, rendering it a severe global concern [12]. Consequently, it is crucial to investigate the origins of loneliness. Many factors contributing to loneliness can be traced back to early life environments [13]. Father’s love, as important interactions in an individual’s early environment, plays a crucial role because father-love absence can result in the deficiency of essential emotional bonds between father and children. Moreover, it may impede the establishment of intimate relationships, hinder the development of fundamental trust. Thus, it is hard for individual to develop the comprehension of intricate emotions, perceptions, and social behaviors as individuals mature. Consequently, individuals experiencing father-love absence have a higher propensity to experience loneliness [14,15].

The association between father-love absence and loneliness is also supported by the social needs theory. The social needs theory is an important theory focusing on the mechanisms of loneliness generation, regarding that loneliness can emerge when adolescents are dissatisfied with their social needs or experience a lack of interpersonal relationships [16]. It is essentially a subjective experience of one’s own social circumstances [17]. Weiss found that individuals possess six inherent social needs, including the needs for attachment, social integration, and nurturing [18]. These needs can only be satisfied within particular social relationship contexts, and the absence of such needs’ fulfillment or the absence of specific social relationships can lead to the emergence of loneliness among the individual. In the early environment of an individual, the father, being a significant relationship [19], plays a crucial role in individual’s growth. Father-love absence results in the individual’s inability to meet attachment and nurturing needs, leading to the development and increase of loneliness. As a result, this study proposes H1: father-love absence positively predicts adolescents’ loneliness.

Gratitude, as an emotion or personality trait, recognizes good deeds and affective experience of appreciation and pleasure that develops when an individual is favored [20]. It includes a complex array of subjective feelings, including curiosity, appreciation, and appreciation of life [21]. As a positive emotion trait, gratitude directs an individual’s attention towards the positive and valuable attributes of others, resulting in a more favorable memory bias [22]. Consequently, it contributes to the establishment and sustenance of relationships while fostering greater closeness with others [23]. Thus, gratitude diminishes the likelihood of an individual experiencing loneliness [24]. Furthermore, fathers play a crucial role in shaping the development of adolescents’ gratitude [25] because fathers offer direct guidance to adolescents regarding gratitude [26]. Simultaneously, fathers’ perceptions of emotional experiences, including gratitude, have an indirect yet significant influence on children’s development of gratitude [27,28]. Therefore, father-love absence may negatively predict individual levels of gratitude. Accordingly, we propose H2: father-love absence influences adolescents to develop loneliness through the mediating effect of gratitude.

Peer relationships encompass interpersonal connections formed and cultivated through shared activities and mutual support during interactions among individuals of the same age and similar levels of psychological development [29]. Peer relationships, functioning as microsystems influencing individual development, have a direct and profound influence on an individual’s loneliness [30]. This subjective experience of lacking peer relationships can evoke feelings of loneliness in individuals. As individuals enter adolescence, their focus on seeking intimacy and emotional support gradually shifts from parents to peers [31], and peer relationships become increasingly vital [32]. Consequently, adolescents with inadequate peer relationships face challenges in attaining social satisfaction and positive emotions from their peers, potentially resulting in loneliness. Similarly, the evidence with regard to the lower the quality of peer relationships, the greater the loneliness experienced by adolescents [33,34]. Regarding the relationship between father-love absence and peer relationships, adolescents experiencing father-love absence often exhibit a diminished sense of security in their interpersonal interactions. So, they face challenges in forming positive relationships with others [35]. In summary, this study proposes H3: father-love absence leads to loneliness through the mediating influence of peer relationships.

Gratitude is a positive emotion that contributes to the establishment of interpersonal relationships, which reduces an individual’s sense of isolation, diminishes the feelings of insecurity and instability [36]. Consequently, it can enhance an individual’s interaction with the external environment, fostering more positive peer social relationships [37–39]. On the other hand, gratitude also aids in repairing interpersonal relationships; individuals who exhibit gratitude can develop more positive perceptions of their partners, display a greater willingness to sustain long term relationship with others [40], and cultivate closer connections among their peers [41]. Therefore, individuals with high levels of gratitude enjoy more harmonious peer relationships.

According to social functionalist theory, emotions are formed in dynamic social interactions that are influenced by relationships [42]. At the same time, emotions can affect individual’s connections with others [43]. Keltner et al. further extended this theory, finding that emotions can also affect one’s interpersonal relationships [44]. Interactions with fathers, as an important interpersonal relationship for individuals, affect the individual’s emotion of gratitude. The emotion of gratitude affects one’s peer relationships. Conversely, connections and interactions with peers may further lead to the individual’s loneliness. Therefore, individuals with father-love absence may present less gratitude emotions, have weaker peer relationships, and possess more loneliness. Building on this premise, we advance H4: father-love absence influences adolescences’ loneliness through the chain mediation involving gratitude to peer relationships.

Prior research has demonstrated significant gender disparities in the consequences of father-love absence on mental health among adolescents. For instance, Sears et al. reported in their study that father-love absence was a stronger effect of aggressive behaviors in males [45]. McLanahan et al. indicated in their research that father-love absence had a more pronounced impact on the emotional and mental health development of males [46]. Hetherington et al. suggested in their study that women with father-love absence are less likely to form intimate relationships compared to men, which may result in heightened feelings of loneliness [47]. Therefore, building upon these previous findings and aligning with the preceding hypotheses, we introduce H5: Gender differences may exist in the mechanisms through which father-love absence affects loneliness between gratitude and peer relationships.

The purpose of the current study is to investigate the impact of father-love absence on adolescents’ loneliness, with a focus on the mediating roles played by gratitude and peer relationships. While some prior research has assessed the influence of early adverse environments on the mental health of adolescents [48–50], much of the existing literature has focused primarily on the effects of childhood maltreatment on adolescent loneliness. However, how father-love absence specifically affects feelings of loneliness in children and the underlying mechanisms has not been directly explored in previous research. At the same time, owing to the traditional division of family roles in China, the role and status of the father within the Chinese family have been consistently due to the distinctiveness of gender role. Resulting in a father-child relationship that diverges from those in other cultures. Chinese society confronts various challenges, including the high prevalence of parental divorce and the shift from the traditional extended family model to the nuclear family model. This could potentially contribute to an increasing absence of father in the Chinese society. Therefore, we recruited participants from China. In summary, utilizing the social functionalist theory, the social needs theory, and a review of relevant literature, this study aims to assess the link between father-love absence and feelings of loneliness.

In this study, questionnaire was administered to 2,148 students from a junior high school and a senior high school in Hunan Province, China. Some participants were excluded from the analyses, because neither those complete all the questionnaires, nor they filled out questionnaires indiscriminately. Thus, a total of 1,795 valid samples were collected, and the questionnaire completion rate was 83.56%. Among them, 319 students in the third year of junior high school, 568 were senior one school students, 462 were senior two school students, and 446 were senior three school students. The sample consisted of 774 males with a mean age of 16.23 ± 1.32 and 1021 females with a mean age of 16.33 ± 1.17. All participants were free from any mental health condition. They all provided handwritten informed consent prior to completing the questionnaire. All subjects signed an informed consent form before participating in the formal study. At the same time, we obtained full consent from teachers and parents before distributing the research questionnaire to the students. Detailed instructions were given to fill out the questionnaire. The study was approved and consented by the ethics approval of the authors’ institution.

First, a common method deviation analysis was performed on the data by Harman’s one-way test using validated factor analysis incorporating all variables into the same latent variable. The results revealed a very poor model fit index [ Δχ2 = 32.514, p < 0.001; RMSEA = 0.133; SRMR = 0.129; CFI = 0.436]. Therefore, there is no serious common method bias in this study.

The constructed measurement models were then tested using AMOS 24.0 to examine how well the indicators predicted the underlying variables. The scale assessing father-love absence were split into four factors. The scale assessing gratitude were split into two factors. The question assessing peer relationships were split into three factors. The scale assessing loneliness were split into three factors. According to the suggestion by Hu et al. [57] and Kenny [58] for the model fit index, Δχ2 = χ2 /df ≤ 5, CFI (0.90 or more), TLI (0.90 or more), RMSEA (0.08 or less), and SRMR (0.08 or less) were used as indicators to evaluate the good fitness of the model.

Descriptive statistics were utilized to illustrate the correlation among the variables. Table 1 displays the mean, standard deviation, and correlations among father-love absence, gratitude, peer relationships, and loneliness. Significant correlations exist between these variables. These findings offer preliminary support for the hypothesis.

The potential variables in the measurement model included father-love absence, gratitude, peer relationships, and loneliness. The results indicated that the data fit the measurement model well [ Δχ2 = 0.781, p < 0.001; RMSEA = 0.041; SRMR = 0.032; CFI = 0.987]. In addition, factor loading for all latent variables were significantly correlated with them (p < 0.001), suggesting that latent variables can be a good proxy for observed variables. Table 1 shows the means, standard deviations, and correlations between father-love absence, gratitude, peer relationships, and loneliness, with all variables correlated with each other. The results may provide initial support for the hypothesis.

Without the mediating variables gratitude and peer relationships, father-love absence significantly predicts loneliness. Model 1 was then developed in which father-love absence directly predicts loneliness and also indirectly predicts loneliness through gratitude to peer relationships (see Fig. 1). The results showed a good fit of model 1 [ Δχ2 = 4.006, p < 0.001; RMSEA = 0.041; SRMR = 0.032; CFI = 0.987] (see Table 2). Therefore, we used model 1 as the structural model.

Also, the bootstrap method was used to test the stability of the model mediation effect by randomly sampling 2000 bootstrap samples generated from the original data set (N = 1795). The results showed that gratitude [−0.086, −0.015] and peer relationships [−0.238, −0.122] mediated significantly between father-love absence and loneliness at 95% confidence intervals. Gratitude [0.054, 0.160] played a significant role between father-love absence affecting peer relationships, and peer relationships [0.079, 0.155] between gratitude affecting loneliness (see Table 3).

To test the stability of the structural model, the model was further analyzed for stability across gender. First, the five latent variables were tested for the presence of gender differences. The results showed that there were no significant gender differences in father-love absence [t₍₁₇₉₅₎ = 0.846, p > 0.05], gratitude [t₍₁₇₉₅) = −1.146, p > 0.05], and loneliness [t₍₁₇₉₅₎ = 1.103, p > 0.05]. In contrast, there were significant gender differences in peer relationships [t₍₁₇₉₅₎ = 2.084, p < 0.05] and the male peer relationship scores were higher than the female peer relationship scores.

Based on the gender differences found, further multi-cluster analysis was used to examine the stability of the models. Based on the Byrne’s [59] study, two models were developed in this study, one allowing free estimation of pathways across gender (unconstrained structural pathways) and the other allowing significantly equal path coefficients for both genders (constrained structural pathways), while holding the underlying parameter factor loadings, error variances, and structural covariances constant. The results showed a significant difference between the two models [ Δχ2 = 2.378, p < 0.001], while both models met the criteria for goodness of fit (see Table 4). Therefore, the above multicluster analysis model is acceptable. In addition, because of the susceptibility to reach significant levels due to the large sample size, to verify the cross-sex stability of the models, we further examined the cross-sex stability of the structural models using the critical ratios of differences (CRD) of the standard deviations of the two models as indicators. According to the decision rule, an absolute value of the CRD greater than 1.96 indicates a significant difference between the two parameters [60]. The results indicated that there was a significant difference in the structural path from father-love absence to peer relationships (CRD = −2.764), and none of the other structural paths for the remaining variables (CRD father-love absence to gratitude = 1.266, CRD gratitude to loneliness = 1.574, CRD gratitude to peer relationships = −0.716, CRD father-love absence to loneliness = −0.6, CRD peer relationship to loneliness = −0.202).