Skip to content
problem genome

One of 592 unsolved problems in the Problem Genome, a collection written up for student design and engineering teams.

social · health

surrounded, and still alone

After a spouse dies, the loneliness that tracks grief is the kind more company does not easily fix — and a traditional bereavement group in the one trial that measured both kinds moved neither

Source: Vedder et al., “Emotional vs. social loneliness and prolonged grief: a…” Open ↗source: primaryLast updated: 2026-11-12

Problem statement

Loneliness research has long distinguished two kinds. In Weiss's (1973) terms — quoted here as Vedder et al. (2025) quote them; Weiss's book itself was not read — emotional loneliness is "a sense of utter aloneness, whether or not the companionship of others is in fact accessible," and social loneliness is "the absence of an engaging social network." Support that works by adding contact — visits, groups, outings, befriending — speaks to the second kind. After the death of a spouse, one cohort points to the first kind as the one tied to grief: in Danish spousal-bereaved adults aged 65–81 followed from 6 to 48 months after the loss (n = 206 at 6 months), emotional loneliness was linked within-person to prolonged grief symptoms over time, while social loneliness was not (in their model, grief symptoms predicted later emotional loneliness, not the reverse; both kinds were measured with single items). The authors conclude that "loneliness following spousal loss is not easily alleviated by increasing social contacts," and that "so far, no studies have explicitly explored emotional loneliness in bereaved individuals, leaving the question of how best to provide support unanswered." The unsolved problem is that support built around company does not easily reach the loneliness most tied to grief, which is not a shortage of company.

Why this matters

Loneliness is more frequent and intense among the bereaved than the non-bereaved (as Vedder et al. summarise earlier studies), and the stakes are clinical: prolonged grief disorder is now in both DSM-5-TR and ICD-11, and the DSM-5-TR (but not ICD-11) lists loneliness as a symptom. Vedder et al. note that emotional loneliness rises significantly from before to after the loss and that most bereaved individuals experience persistently high levels for years, and that the two kinds share only about 19% of their variance — so a service that measures "loneliness" as one number, or counts contacts, can look successful while missing the kind that tracks grief. Holt-Lunstad's 2024 review makes the general version of this point: assessments that capture only one component of social connection "will likely result in risk assessment errors," and "currently there is no measure that is multi-factorial, validated, and feasible" across settings.

What’s been tried and why it hasn’t worked

Bereavement groups are one organised response. The one trial Vedder et al. identify as including both emotional and social loneliness as outcomes is a cluster-randomised trial in 12 community centres serving older people (215 widowed older adults contacted, 125 eligible participants assessed before, after, and at 16-week follow-up). The control arm was a conventional loss-oriented bereavement group (seven weekly 2-hour sessions plus a 4-hour outing): it improved grief, depression and social support, but participants did not report reduced emotional or social loneliness. Only the experimental dual-process group — grounded in the Dual Process Model — reduced both, and the trial's authors, as the cohort paper reports them, suggest the social-loneliness reduction may be attributed to the group setting and the emotional-loneliness reduction to restoration-oriented coping with the loss (a suggestion, not a tested mechanism; the trial was read at abstract level only). One-to-one support through home visits has been tried for this population too: a Dutch randomised trial offered lonely widowed older people 10–12 home visits by a trained, themselves-widowed volunteer (110 assigned to the visiting service vs 106 controls); the visiting service was not more effective than the control condition in the overall sample on the outcomes its abstract names (mental health and quality of life), though socially lonely, physically ill or low-educated respondents improved more. Whether that trial reported emotional and social loneliness separately as outcomes is not stated in what was read. Beyond these, the cohort authors note that the latest meta-analyses "suggest that CBT interventions may have beneficial effects in addressing both loneliness and prolonged grief symptoms" but also "highlight the overall inconsistent success of interventions in mitigating loneliness and prolonged grief symptoms." The binding constraints are therefore (1) a support model that equates loneliness with lack of contact, (2) almost no trials that separate the two kinds of loneliness as outcomes, and (3) measurement in practice — the Danish cohort itself used single items, which the authors list as a limitation.

What would unlock progress

A support design aimed at emotional loneliness specifically, and a way to tell which kind a bereaved person has. The sources suggest leads: restoration-oriented coping within groups (the dual-process model); recalibrating expectations of relationships, which De Jong-Gierveld et al. and Bouwman and van Tilburg are cited as linking to lower emotional loneliness; CBT-based approaches to the "merged identity" with the deceased; and integrating information on emotional loneliness into the psychoeducation already common in grief interventions. All of these are proposals in the sources, not established treatments. On measurement, validated instruments that separate the two kinds exist (the De Jong Gierveld scale has emotional and social subscales), so the gap the sources point to is in their use in bereavement research and services (Vedder et al. call replication with the established De Jong Gierveld scale perhaps the most important next step), not in their invention.

How we’d know it’s being solved

Emotional and social loneliness reported separately, using a validated multi-item instrument with both subscales, at referral and at follow-up of at least several months, alongside a prolonged-grief measure. A bereavement service would be making progress if emotional-loneliness scores fell, not only social-loneliness or contact counts — and a trial would show it against a conventional bereavement-group comparator, as Chow et al. did. Beware the proxy: attendance at groups, number of visits, and size of support network measure the social kind; they can all rise while emotional loneliness stays put.

Where to start

Facility-free door: design a brief screening and triage tool for bereavement volunteers and services that distinguishes emotional from social loneliness, built on the published De Jong Gierveld subscales, with plain-language scripts for what to offer in each case; test comprehension and acceptability with volunteers or older adults recruited informally (no clinical data). Research door: code the published bereavement-intervention literature (the trials inside the meta-analyses the sources cite) for whether loneliness was measured at all, whether the two kinds were separated, and which intervention components were present — a systematic map that tests the sources' claim that the distinction "has been largely overlooked in grief research." Start from the measurement review Vedder's group published in 2021: it covered 63 articles on 51 datasets, mostly cross-sectional and longitudinal studies, found that about half used one of two validated scales and the rest only single- or few-item measures, and counted the De Jong Gierveld scale in 8 of the 51. It included only five intervention studies, so it is not yet the map of intervention trials this door asks for. Partner door (design the trial, don't run it): with a hospice bereavement service or charity, write a protocol for adding a restoration-oriented module to an existing loss-oriented group, with emotional loneliness as the primary outcome and a pre-registered analysis plan, and hand it over. Access note: the Danish cohort data are not public; none of the doors needs them. Relevant skills: psychology, service design, measurement, systematic-review methods.

Genes

Related problems

Stuck the same way, in other fields.

field: the built worldthe battery eats the cargo spaceHeavy-duty long-haul truck battery weight displaces cargo capacitybuilt without the people who'd live with itsource: secondaryfield: making & materialssolar glass no furnace wantsSixty to seventy percent of a solar panel is glass, and the antimony added to make it clearer keeps that glass out of float furnaces — so "recycled" glass goes to lower-quality usesbuilt without the people who'd live with itsource: secondaryfield: the built worldcounts of unknown accuracyHighway decisions rest on traffic counts whose accuracy is rarely known, checked by methods that vary agency to agency, and may not hold after installationbuilt without the people who'd live with itsource: primaryfield: the built worldthe accident reports that never comeRoughly half of recent airline accidents have no published final report, even though the treaty rules ask States to publish one within a year where possiblebuilt without the people who'd live with itsource: primary

Sources

  • Vedder A, O'Connor M, Boelen PA. "Emotional vs. social loneliness and prolonged grief: a random-intercept cross-lagged panel model." European Journal of Psychotraumatology 16(1):2488101, 2025, doi:10.1080/20008066.2025.2488101, accessed 2026-10-22. Link ↗
  • Chow AYM, Caserta M, Lund D, Suen MHP, Xiu D, Chan IKN, Chu KSM. "Dual-Process Bereavement Group Intervention (DPBGI) for Widowed Older Adults." The Gerontologist 59(5):983–994, 2019, doi:10.1093/geront/gny095 (abstract), accessed 2026-10-22. Link ↗
  • Holt-Lunstad J. "Social connection as a critical factor for mental and physical health: evidence, trends, challenges, and future implications." World Psychiatry 23(3):312–332, 2024, doi:10.1002/wps.21224, accessed 2026-10-22. Link ↗
  • Onrust S, Willemse G, van den Bout J, Cuijpers P. "Effects of a visiting service for older widowed individuals: a randomized clinical trial." Death Studies 34(9):777–803, 2010, doi:10.1080/07481181003761252 (abstract), accessed 2026-10-22. Link ↗
  • Onrust S, Smit F, Willemse G, van den Bout J, Cuijpers P. "Cost-utility of a visiting service for older widowed individuals: randomised trial." BMC Health Services Research 8:128, 2008, doi:10.1186/1472-6963-8-128, accessed 2026-10-22. Link ↗
  • Vedder A, Stroebe MS, Schut HAW, Boerner K, Stokes JE, Boelen PA. "Loneliness in Bereavement: Measurement Matters." Frontiers in Psychology 12:741762, 2021, doi:10.3389/fpsyg.2021.741762, accessed 2026-11-12. Link ↗
Revision notes

This brief gets checked against its sources and contains the following revisions. Future revisions get added here.

Worth knowing

  • the cohort linked grief to emotional loneliness within each person over time; between people, the same paper does report a link for social loneliness
  • the Dutch home-visiting trial offered peer support with grief content, not just company

Full check

Sources read

  • Vedder et al., grief and emotional and social loneliness after loss (European Journal of Psychotraumatology, 2025): full text
  • Onrust et al., a visiting service for widowed older people: trial design (BMC Health Services Research, 2008): full text
  • Onrust et al., visiting service trial results (Death Studies, 2010): abstract only
  • Chow et al., Dual-Process Bereavement Group Intervention trial (The Gerontologist, 2019): abstract only
  • Holt-Lunstad, social connection as a critical factor for health (World Psychiatry, 2024): full text
  • De Jong Gierveld & Van Tilburg, 6-item loneliness scale (Research on Aging, 2006): record only
  • Vedder et al., Loneliness in Bereavement: Measurement Matters (Frontiers in Psychology, 2021): full text
  • Weiss, Loneliness (MIT Press, 1973): record only (quoted as the 2025 cohort paper quotes it)

What changed

  • Not found in the sources, replaced with what they say: the Dutch trial's null result could be read as a loneliness result; the page now names the outcomes its abstract gives, mental health and quality of life.
  • Contradicted by the source, corrected: the Dutch home-visiting trial was described as contact on its own; the trial paper describes one-to-one support from trained widowed volunteers.
  • Numbers: 110 people were said to be visited; 110 were assigned to the visiting service, and the sources do not say how many were visited.

Left open

  • the bereavement-group trial's figures and results (abstract only)
  • whether the Dutch visiting trial reported the two kinds of loneliness separately as outcomes (abstract only)
  • the 1973 definitions of the two kinds of loneliness (quoted second-hand; the book was not read)

Separate check of specific claims: A quotation from the cohort paper now appears in full, and a published review of how loneliness after loss has been measured was added to a starting point.

Earlier full check: Three claims were softened to match the sources, including how the bereavement-group programme was described.

Separate check of specific claims: The titles were toned down to match the source, a Dutch home-visiting trial was added, and statements no source made were cut or narrowed.