The study, conducted by a team at Jilin University’s School of Public Health and published in Social Indicators Research, analyzed 2,251 working-age participants from the Longitudinal Internet Studies for the Social Sciences panel, a probability-based internet survey in the Netherlands widely regarded as a methodological gold standard. Six waves of measurement spanning seven years allowed the researchers to model not just whether people felt lonely or anxious, but how those feelings changed over time and whether the pace of change in one predicted the course of the other. The answer, in short, was yes, and in both directions.

Technically, the team relied on a family of structural equation models known as latent growth curve models. Unconditional versions first mapped the raw trajectories of loneliness and internalizing difficulties separately, estimating each person’s starting level and their rate of change across the seven annual intervals. Parallel process versions then modeled both trajectories simultaneously, allowing the researchers to ask whether someone’s baseline loneliness shaped the slope of their anxiety and depression symptoms, and vice versa. Finally, conditional models and a parallel process latent class growth analysis grouped participants into distinct co-developmental patterns and tested how membership in each group related to subjective well-being, the scientific shorthand for how positively people evaluate their own lives.

The first headline finding is quietly optimistic: across the seven years, both loneliness and internalizing difficulties declined on average in the working-age population. This contradicts a common cultural assumption that loneliness accumulates relentlessly through adulthood. Instead, the average trajectory bends downward, echoing recent coordinated analyses of nine longitudinal studies suggesting that loneliness does not inevitably climb with age. But averages conceal as much as they reveal, and the deeper story of this study lies in the variability hiding beneath that downward trend.

That variability is where the reciprocal dynamics emerge. Higher initial levels of loneliness predicted steeper or more adverse subsequent trajectories of internalizing difficulties, and the reverse held as well: people who began the study with more depressive and anxious symptoms tended to become lonelier faster. This bidirectional coupling is consistent with theoretical accounts in which loneliness sharpens attention to social threat, making social interactions feel riskier and more taxing, which in turn deepens withdrawal and low mood. It also aligns with a growing body of cross-lagged research showing that loneliness and common mental disorders feed each other over time, creating feedback loops that can trap individuals in a slow spiral of isolation and distress. The present study extends those findings by demonstrating the coupling at the level of whole trajectories rather than single consecutive waves.

The consequences of this spiral show up clearly in subjective well-being. Participants who began the study with higher baseline loneliness or higher internalizing difficulties reported systematically lower well-being throughout. More strikingly, the rate of change mattered: people whose internalizing difficulties declined more steeply over the seven years enjoyed greater well-being than those whose symptoms lingered. In other words, how quickly someone sheds anxiety and depressive symptoms is not merely a clinical curiosity, it is a statistical predictor of how satisfied and happy they will feel, independent of where they started. This suggests that recovery pace, not just initial severity, deserves a central place in how researchers and clinicians think about adult mental health.

The latent class growth analysis added a third layer of insight by sorting participants into three distinct co-developmental profiles. A High Level-Declining Group started with elevated loneliness and internalizing difficulties but improved over time. A Rising Internalizing Difficulties Group, the most concerning of the three, began with comparatively fewer symptoms but grew worse across the seven years. A Low Level-Declining Group started low and stayed low, representing the resilient majority. Well-being differed significantly across these groups, with the rising-symptom class showing markedly poorer outcomes. The existence of this worsening subgroup is perhaps the study’s most practically important discovery, because conventional screening approaches that focus only on current symptom levels would overlook people whose trajectory is quietly heading in the wrong direction.

Why does this matter for the workforce in particular? The working-age population carries the double burden of occupational stress and the social expectations of family and community life, and previous research has linked loneliness among employed adults to reduced productivity, higher healthcare use, and poorer mental health service engagement. International reviews have documented loneliness prevalence across 113 countries, and dozens of national governments have now launched formal loneliness strategies. Yet most interventions are designed for children or retirees. By showing that loneliness and internalizing symptoms dynamically co-develop during the prime working years, and that their early levels seed later trajectories, the study argues for prevention rather than reaction: monitoring both conditions together, identifying individuals on adverse co-developmental paths early, and intervening before reciprocal reinforcement takes hold.

The technical sophistication of the analysis also deserves note. Working with ordered categorical survey measures across six waves, the authors verified longitudinal measurement invariance to ensure that the scales meant the same thing at every time point, a step that many longitudinal studies skip and one that, if ignored, can distort growth parameters. Model fit was evaluated against conventional structural equation modeling criteria, and the use of a probability-based panel mitigates the selective non-response biases that plague convenience samples. The data themselves are openly available through the LISS infrastructure, allowing other researchers to replicate or extend the findings.

Still, the study has limits that temper its conclusions. The sample is Dutch, and loneliness norms, social support structures, and labor market conditions differ across cultures, so the specific trajectory shapes may not generalize universally. Self-report measures capture perceived loneliness and symptoms rather than clinical diagnoses. And while parallel process growth models reveal how trajectories relate, they cannot on their own prove causation; unmeasured factors, from personality to economic shocks, could partially drive both paths. Even so, the reciprocal pattern is difficult to explain without some genuine mutual influence, and it converges with experimental and neuroimaging work showing that perceived social isolation alters cognition and threat sensitivity in ways that plausibly feed internalizing symptoms.

The broader takeaway is a reframing of two familiar problems into one intertwined process. Loneliness is not simply an unhappy byproduct of poor mental health, and depression and anxiety are not merely consequences of being alone. Over years of adult life, the two develop as a coupled system, each shaping the slope of the other, and together they determine how positively people judge their own lives. For a society grappling with an aging workforce, post-pandemic social disruption, and rising policy attention to social connection, the message from seven years of Dutch panel data is clear: to protect well-being during the working years, treat loneliness and internalizing difficulties as two faces of a single, trackable, and treatable trajectory.

Subject of Research: Longitudinal co-development of loneliness and internalizing difficulties and their impact on subjective well-being in the working-age population

Article Title: How Do Loneliness and Internalizing Difficulties Change Subjective Well-Being?—A Trajectory Study of the Working-Age Population

Article References: Zhang, C., Yuan, T., Cui, Y., Zheng, C., Qing, Y., Liu, K., Pan, L., Zhang, L., Su, Y., Ding, F., & Mei, S. (2026). How Do Loneliness and Internalizing Difficulties Change Subjective Well-Being?—A Trajectory Study of the Working-Age Population. Social Indicators Research, 184(2), Article 43. https://doi.org/10.1007/s11205-026-03925-6

Image Credits: AI Generated

DOI: 10.1007/s11205-026-03925-6

Keywords: loneliness, internalizing difficulties, subjective well-being, latent growth curve models, LISS panel, working-age population, depression, anxiety, developmental trajectories, mental health, longitudinal study, Netherlands