This study highlights a significant association between parental income and mental health outcomes among adolescents, revealing that young individuals from lower-income families experience higher odds of poor mental well-being compared to their peers from higher-income backgrounds after adjusting for country of origin, living arrangements, parental educational level, and illness or disability of the parents. Notably, this relationship remains consistent across genders. Furthermore, our findings indicate a nonlinear relationship between parental income and the odds of poor mental well-being in youth. Specifically, we identified a threshold range of parental income between approximately 500,000 DKK (≈ 67,000 €) and 1,500,000 DKK (≈ 200,000 €). Within this range, increases in parental income correlate more strongly with improvements in mental health. This suggests that while financial resources are crucial, their effect may diminish at higher income levels. This hypothesis aligns with theories suggesting that income and mental health is associated only until a certain threshold. According to this view, income influences mental health primarily by enabling people to meet their basic needs [14].

We were unable to find any previous studies that specifically investigated the relationship between parental income and adolescent mental well-being. However, several studies have explored related outcomes such as psychiatric disorders and mental health problems. For instance, a discordant sibling study by Sariaslan et al. [15] concluded that there was no association between parental income and psychiatric disorders. However, the methodology used in their study to address the research question has been questioned [16,17,18,19]. Their conclusion not only contrasts with our results but also with previous research, including a systematic review that identified an inverse relationship between socioeconomic status and mental health problems in children and adolescents [4]. This review also indicated that there were no significant gender differences in this relationship, which aligns with our findings. Consistent with our results, both Kinge et al. [20] and Hynek et al. [21] found an inverse relationship between parental income and the risk of mental disorders. Kinge et al. further observed that this relationship was non-linear. However, while Hynek et al. [21] found no gender differences in the association, consistent with our findings, Kinge et al. [20] reported gender specific patterns. This difference may be due to the fact that Kinge et al. explored gender-specific disorders (e.g., ADHD in boys, anxiety and depression in girls), while Hynek et al. and our study explored mental health, in general, which may have obscured gender-specific findings for particular diagnoses.

The main strengths of our study include the large sample size and the ability to link to national registers. This linkage allowed us to gather information on parental income, country of origin, and parental education levels. This use of registry-based avoids the biases that may arise from self-reported information. For instance, parental income was based on documented data, increasing the reliability of our findings.

This study also has limitations. The cross-sectional design prevents us from establishing causal relationships. While we observed an association between parental income and poor mental well-being, the direction of this relationship remains unclear, and reverse causation could be at play. For example, an adolescent with low mental well-being facing difficulties in life may need more care and attention from their parents, which may limit their ability to work. In this way, poor mental well-being in young people could lead to lower parental income, by reducing the parents’ ability to work.

Additionally, while DNYS 2014 had a high participation rate of 85%, the 15% who did not participate could still introduce bias. A study by Pisinger et al. [8] found that participants in DNYS 2014 are more likely to have parents with higher incomes than non-participants. If the non-participants also have lower mental health, as our findings suggest, it could influence the results.

A total of 1,918 individuals were excluded due to missing data on one or more key variables. Previous research by Pisinger and colleagues has shown that non-respondents within the Danish DNYS 2014 were more likely to be male, of non-Danish origin, and have parents with lower incomes and educational levels [8]. This suggests that excluded individuals may have been more socioeconomically disadvantaged, potentially biasing our findings toward a more conservative estimate of the association. Our findings may not fully reflect the mental well-being of all Danish high school students, as they may miss those from more vulnerable backgrounds. Another potential source of bias is the reliance on self-reported data in the DNYS 2014 survey for variables such as living situations, parental illness or disability, and mental well-being. Self-reporting may lead to recall bias or social desirability bias [22, 23]. For instance, respondents might overestimate their mental well-being or underreport parental illness due to stigma. Such biases could result in an underestimation of poor mental well-being in this study.

Additionally, since this study relies on pre-collected data, we were unable to include certain potentially important confounding variables. For instance, Kinge et al. [20] included the number of household members in their analysis, explaining some of the association between parental income and mental disorders. Household size could also influence both parental income and poor mental well-being in young people in this study. Having more siblings has been associated with better well-being [24], while larger family sizes can also lead to lower parental income due to reduced working hours. Since we did not have data on household size, we were unable to account for this potential confounder. If included, it would likely strengthen the observed association between parental income and low mental well-being, resulting in a higher odds ratio. Another approach to account for this would have been to use equivalized disposable income, which accounts for household size [9], but this information was not available. Another thing to notice is that although we refer to our income measure as “parental income” it is technically a measure of total household income and may include income from children and adolescents within the household. We were also unable to identify whether some individuals were siblings. This means we could not account for potential clustering within families, which may violate the assumption of independence in the logistic regression models.

Moreover, examining the role of community resources, such as access to out-of-school activities, mental health services, or supportive school environments, may be essential in mitigating the effects of low parental income on adolescent mental health. These contextual factors could serve as protective buffers, particularly for adolescents from lower-income households, and could be integrated into future studies on the topic.