Loneliness is more than an uncomfortable feeling—research links it to measurable health risks, including increased likelihood of earlier mortality. Yet here’s what makes addressing loneliness particularly challenging: Not all loneliness is the same.

When Generic Advice Doesn’t Work

Consider two groups particularly vulnerable to loneliness, but for entirely different reasons.

University students arriving from high school are often surrounded by people—dorm neighbors, classmates, study groups, and endless social activities. Yet many feel profoundly alone. They’ve just left behind their entire social network at home. They’re forming new identities, questioning who they are, and navigating independence for the first time. The calendar is full, but something crucial is missing: that person who truly knows them, who they can be vulnerable with at 2 a.m. when the facade of having it all together crumbles.

Older adults face the opposite challenge. After decades of marriage, they may have lost their partner. Friends have moved away or passed on. Their social world has contracted dramatically.

They might have one or two close family members who provide deep emotional support, but they’re missing the broader community—people to grab coffee with spontaneously, neighbors to chat with, a sense of belonging to something larger than their household.

Both groups report feeling “lonely,” but telling the student to “deepen existing relationships” or the older adult to “join more clubs” misses what each actually needs. These require fundamentally different solutions.

Loneliness Isn’t One-Size-Fits-All

But this is not only about life stages; this concerns a fundamental challenge in understanding loneliness. What causes loneliness for one person differs dramatically from what causes loneliness for another. The triggers vary, the subjective experience varies, and effective solutions vary just as much.

Researchers have tried to capture this complexity through measurement. A 2022 review by Maes and colleagues of the eight most commonly used loneliness scales found something sobering: All of them included items that don’t actually measure loneliness. On many scales, about half the items measured related but distinct experiences, such as depression or social support, rather than loneliness itself.

In our own systematic review work, we are trying to quantify just how varied the broader concept of social connection really is. By comparing different measurement approaches, in our registered report, we are predicting that categories of social connection (function, quality, and structure) that are often considered to be well understood might not be consistent with what researchers mean when they study “connection” or “loneliness.” The predicted fragmentation makes it harder to build cumulative knowledge about what actually helps.

Starting With Two Dimensions

One approach that has gained traction distinguishes between emotional loneliness (missing intimate attachment) and social loneliness (lacking a broader network). The De Jong Gierveld Loneliness Scale (DJGS-6), developed by Dutch researcher Jenny de Jong Gierveld and her colleagues, is an abbreviated version of the longer scale, using six questions to identify which pattern you’re experiencing—or whether you’re experiencing both simultaneously.

A 2022 EU-wide survey of over 25,000 people across all 27 European Union countries used the questionnaire to measure loneliness, and we have been evaluating the measure’s properties. Our upcoming analysis of this same EU data revealed important limitations in how well the DJGS-6 actually measures what it intends to measure. The scale showed strong relationships with emotional states like depression and happiness across all countries. However, it showed weaker relationships with indicators of social activities and attitudes—the very things we’d expect a loneliness measure to strongly relate to. In addition, across the European Union, the way that users interpret the questions means we cannot compare between countries. This means that we still have important shortcomings in how we understand disconnection.

The Limits of Current Measurement

Let’s be transparent about what we don’t yet measure well. The DJGS-6 has several limitations:

It was developed in the Netherlands (a higher-income country). If it were developed elsewhere, would we measure loneliness differently?
Its content validity scores are modest at best.
If you compare across countries, you can’t compare scores.
Different loneliness scales produce different prevalence estimates—there’s no universally agreed score that defines “lonely.”
It doesn’t specify a timeframe (loneliness can fluctuate daily or persist chronically).
It doesn’t capture relationship quality versus quantity or distinguish online from in-person connections.

Think of current assessment tools as spotlights illuminating specific parts of a dark room. The DJGS-6 focuses on two areas: intimate connections and broader social networks. Other aspects remain in shadow. The tool won’t capture everything and may not be able to compare you to others, but it can reveal which specific type of disconnection you’re experiencing—useful information for deciding where to focus your energy.