INTERACT Study Maps Loneliness Patterns Across England

The Measuring Loneliness in England (INTERACT) Study has released initial descriptive findings from a large-scale, cross-sectional survey of over 135,000 adults. Published in BMC Global and Public Health, the research aims to characterize the distribution, intensity, and sociodemographic patterning of loneliness across diverse population groups in England. Unlike previous large-scale studies that often focused on specific age cohorts or localized areas, the INTERACT study integrates validated loneliness measures with neighborhood-level social capital indicators and geospatial mapping. The primary objective of this first phase is to provide a detailed account of the study design, survey instrument, and sample characteristics, alongside initial descriptive findings from the analytic cohort collected between March and July 2023.

Loneliness is increasingly recognized as a major public health challenge with significant implications for mental, physical, and social wellbeing. It is distinct from social isolation, which refers to an objective lack of social contact, whereas loneliness is a subjective, distressing experience resulting from a discrepancy between desired and actual social relationships. The World Health Organization has identified social isolation and loneliness as critical determinants of health, highlighting the need for evidence-based interventions. In the United Kingdom, the economic burden is substantial, with estimates suggesting loneliness costs employers up to £3.7 billion annually. Despite this, large-scale studies capturing loneliness across the full adult life course, alongside contextual factors such as social capital and geography, have remained limited.

The INTERACT study employed a multi-methods quantitative approach, combining survey data with geospatial analysis. Participants were recruited via National Healthcare Service Primary Care Networks, voluntary and community sector organizations, and the National Institute for Health and Care Research Be Part of Research Network. Targeted recruitment was also conducted via social media platforms to reach younger adults and digitally engaged populations. The survey was open from October 2021 to June 2024, but the present analysis reports findings from a defined subset of respondents who completed the survey between 1 March and 31 July 2023. A total of 135,725 adults participated, all aged 16 years or older and residing in England.

The study utilized two validated instruments to measure loneliness: the University of California Los Angeles (UCLA) 3-item Loneliness Scale and the Office for National Statistics Direct Measure of Loneliness (DMOL). The UCLA scale assesses how often participants feel they lack companionship, feel left out, or feel isolated from others, with scores ranging from 3 to 9. The DMOL asks how often participants feel lonely, with response options ranging from never to often or always. Additionally, the survey incorporated social capital and community trust indicators adapted from validated scales, assessing neighborhood trust, reciprocity, cohesion, and perceived social support. These items were dichotomized and summed to generate an overall social capital score ranging from 0 to 7.

The sample was predominantly female, with 61.6% of respondents identifying as women, and 82.6% identified as White. The largest age group consisted of those over 65 years old, accounting for 32.4% of the cohort, followed by those aged 56–65 (24.5%) and 46–55 (16.2%). Educational attainment was varied, with 42.7% holding a university degree or higher. The sample over-represented highly educated individuals and under-represented minority ethnic groups compared to national census data. Regarding employment, 33.18% were retired, 28.28% were employed full-time, and 10.8% were employed part-time. Additionally, 18.4% reported having a disability and 45.7% had a long-term health condition.

Results indicated that 16.5% of participants reported feeling lonely “often or always” on the DMOL, while 26.4% felt lonely some of the time. On the UCLA-3 scale, 21% to 22.5% of participants frequently felt a lack of companionship, left out, or isolated. The median UCLA-3 score was 6, with an interquartile range of 3 to 7. This right-shifted distribution suggests that at least half of the respondents scored at or above 6, a threshold sometimes used as an indicative marker of loneliness. However, the authors caution that such cut-offs are not universally standardised and should be interpreted with caution within non-probability samples. The median represents the distribution of responses within the sample rather than a population prevalence estimate.

Perceptions of social capital were mixed, with a median score of 4 and an interquartile range of 2 to 6. While a majority of participants agreed that neighbors were willing to help (51.7%) and could be trusted (58.9%), responses were more evenly distributed for indicators of close-knit communities and emergency support. The impact of the COVID-19 pandemic was also assessed, with 44.0% of participants reporting increased loneliness and 48.2% reporting increased isolation during the pandemic period. These responses are based on retrospective self-report and may be influenced by recall and acquiescence bias, given the cross-sectional design of the study.

Geospatial analysis using kernel density estimation at the Lower Super Output Area level indicated clustering of participants reporting higher loneliness scores in major urban areas, including London, Birmingham, and Manchester. These spatial patterns reflect the distribution of responses within the sample rather than underlying population-level prevalence. The authors note that these findings are consistent with literature describing the “urban paradox,” whereby higher population density does not necessarily translate into greater social connectedness. However, given the non-probability nature of the sample, these patterns should be interpreted as reflecting response distribution rather than geographic differences in loneliness across the wider population.

The principal limitation of the study is selection bias arising from the volunteer-based, non-probability sampling design. Recruitment through NHS and community channels may have preferentially included individuals with poorer health or greater engagement with services, which could plausibly elevate observed levels of loneliness. The online-only survey format may also underrepresent individuals with limited digital access or literacy, including older adults, rural populations, and people with disabilities. Consequently, the findings are descriptive of the responding cohort and are not intended to estimate population prevalence. The large sample size does not compensate for the lack of representativeness, and estimates derived from this dataset should not be interpreted as representing the general population.

Despite these limitations, the INTERACT Study provides a comprehensive national dataset on loneliness and social disconnection in England. The scale, methodological rigour, and spatial granularity of the study offer indicative insights that may inform the design of targeted, place-based interventions. The authors emphasize that the findings are hypothesis-generating and require confirmatory analyses in representative samples. Future work within the INTERACT programme will explore longitudinal patterns, potential determinants, and intervention-relevant outcomes, including additional data collection among NHS staff and qualitative work on lived experiences of loneliness among selected groups.

Subject of Research: Descriptive analysis of loneliness distribution and social capital in a large volunteer sample in England

Article Title: Measuring loneliness at a national scale: The measuring loneliness in England (INTERACT) study’s approach and initial findings

Article References: El-Osta, A., Alaa, A., Al-Ammouri, M., Altalib, S., Tristán-López, A., & Majeed, A. (2026). Measuring loneliness at a national scale: The measuring loneliness in England (INTERACT) study’s approach and initial findings. BMC Global and Public Health, 4(1), Article 69. https://doi.org/10.1186/s44263-026-00296-5

Image Credits: AI Generated

DOI: 10.1186/s44263-026-00296-5

Keywords: loneliness, public health, INTERACT study, social capital, geospatial analysis, England, survey research, mental health, Measuring, national, scale, INTERACT