Latest / This Week In Wellbeing Measurement / Can Life’s Essential 10 predict healthspan? Plus: flood wellbeing

Transcript

Can Life’s Essential 10 predict healthspan? Plus: flood wellbeing · This Week In Wellbeing Measurement

• 0:00
  Jenny: When something bad happens in your community, where do you think the first signs of stress actually show up?
• 0:06
  Davis: Honestly, before the clinic or the council meeting, I think it shows up in tiny routines breaking, like school runs, group chats, and whether people still trust the next update.
• 0:16
  Jenny: That's what makes me intrigued and nervous about using the internet as a kind of thermometer, because public posts can show distress after floods, but they can also miss everyone who's offline, private, or just too tired to type.
• 0:30
  Davis: Right, but if the water's rising, even an imperfect early signal can tell responders where disruption and alienation are building, so today we're asking what gets counted when wellbeing becomes measurable...welcome to This Week In Wellbeing Measurement on paperboy.fm.
• 1:01
  Jenny: Quick map for this week: the search pulled one thousand two hundred thirty-nine hits, we shortlisted two hundred, and ninety-one papers made the cut. Those ninety-one papers came from about four hundred unique authors across twenty-seven countries, so it's a solid pool, but not a globe-spanning one.
• 1:20
  Davis: And the first twist is volume versus fit. Query hits rose from one thousand one hundred seventy-one to one thousand two hundred thirty-nine, up sixty-eight, or about six percent, while qualified papers fell from ninety-seven to ninety-one. So the net got wider, but the usable catch got smaller.
• 1:39
  Jenny: That makes me ask what kind of material flooded in. The top methods were qualitative studies at nineteen, surveys at fifteen, and quantitative papers at eleven, with only six RCTs, meaning randomized trials where people are assigned to different conditions. So this week leans toward measuring experience in context, not clean lab-style testing.
• 2:02
  Davis: The author base also narrowed. We had four hundred seven unique authors, down thirty-six from last time, an eight percent drop. Inside that, seventy-two were first-time authors, meaning first-ever paper, not just new to our feed; one hundred forty-seven were emerging, and one hundred eighty-eight were experienced.
• 2:22
  Jenny: The country drop is the one I'd flag hardest: fifty-one last episode to twenty-seven this week, down twenty-four countries, or forty-seven percent, even with ninety-one qualified papers. Is that a real concentration in where wellbeing measurement is being published this week, or partly a metadata problem, since cities are zero and institutions show up as just one?
• 2:45
  Davis: Theme-wise, the center of gravity fits the episode: wellbeing appears nine times, mental health eight, then emotion regulation and older adults at three each. That's the through-line in miniature: better measurement tools are showing up, but every tool is also choosing which lives, settings, and tradeoffs count.
• 3:05
  Jenny: Alright, let's get into the papers with A population-level assessment framework for flood-related wellbeing from public discourse in Ireland. Róisín Luo and K. Morrissey are asking whether public posts can reveal what floods do to people, not just what floods do to roads, homes, and insurance claims.
• 3:24
  Jenny: The plain idea is that flood wellbeing has more than one layer. They score distress, meaning fear or emotional strain, functional disruption, meaning daily life getting knocked off course, and institutional alienation, meaning people feeling ignored or let down by authorities, using about two hundred twenty-four thousand Ireland posts from twenty twelve to twenty sixteen that mentioned flood, rain, or storm.
• 3:50
  Davis: How do we know this is measuring flood wellbeing, not just who happens to post online during bad weather?
• 3:56
  Jenny: They pull the posts through Meta Content Library, then use a reasoning model called Wellbeing-Former, which is basically a tool trained to read a post, assign evidence-based scores, and explain why it saw distress, disruption, or alienation. That's a real step beyond counting angry words, but the big limit stays big: social media can be timely and rich, but it isn't the same as a representative population survey.
• 4:23
  Davis: So the useful version is an early warning layer, not a replacement for asking people directly. Two hundred twenty-four thousand posts is enough smoke to check for fire, but it still misses the person with no account, no signal, or no spare minute because they're bailing water out of the kitchen, which is exactly the digital measurement tradeoff this episode keeps circling.
• 4:43
  Davis: That missing person with no account or no spare minute is a good bridge, because this next paper goes back to asking people directly. It's called Factors associated with loneliness and social capital in a large volunteer sample of adults in England, and the sample is huge: one hundred thirty-five thousand seven hundred twenty-five community-dwelling adults, meaning adults living in ordinary homes rather than institutions.
• 5:06
  Davis: The plain finding is that loneliness clustered around younger age, especially sixteen to twenty-five, and around being single, unemployed, or living with disability. The protective number is striking: people reporting nine or more friends had much lower odds of loneliness, with an adjusted odds ratio of zero point zero nine on the UCLA scale and zero point sixteen on the ONS direct measure, where adjusted odds means the model is comparing people while accounting for other measured differences.
• 5:35
  Jenny: But if the UCLA three-question scale and the ONS direct loneliness question don't always tell the same story, which one should policy trust? I noticed university education pointed in opposite directions: higher loneliness on the UCLA scale, lower loneliness on the ONS direct measure.
• 5:54
  Davis: I think the paper's answer is, don't pretend one ruler captures the whole thing. They used both validated loneliness measures, plus a social capital scale built from neighbourhood trust, cohesion, and reciprocity, then ran multivariable ordinal and binary logistic regressions, which are models that estimate associations while holding several factors in view. The big caveat is theirs, not mine: this is a volunteer sample, so even at one hundred thirty-five thousand plus people, it shouldn't be treated as a population estimate for England.
• 6:25
  Jenny: That feels like the right level of confidence: a strong map of who in this sample looks isolated, not a census of national loneliness. For the work, time, support thread, the usable takeaway is pretty concrete: if you're targeting help, don't just ask whether someone feels lonely once; ask in more than one validated way, and pay attention to disability, unemployment, age, and whether the person has anything like nine friends to call.
• 6:52
  Jenny: That “one ruler” problem carries over, but now the ruler is bigger: Associations of Life's Essential 10 with Mortality and Healthspan Free of Cardiovascular Disease and Cancer follows one hundred ten thousand eight hundred six UK Biobank participants and twenty thousand thirty-one NHANES participants in the United States to see whether a ten-part heart-health score predicts death and healthspan, meaning years lived before cardiovascular disease or cancer shows up.
• 7:20
  Jenny: Plainly, people with high scores did much better. Life's Essential ten takes the usual heart-health checklist and adds alcohol use and mental health. In UK Biobank, high cardiovascular health versus low cardiovascular health was linked to lower all-cause mortality, with adjusted hazard ratios of zero point four three for men and zero point three four for women, meaning the death rate during follow-up was less than half after accounting for other factors.
• 7:49
  Jenny: At age fifty, the model projected five point four seven more years of overall life expectancy for men with high cardiovascular health, and five point eight five more years for women. It also projected about four extra years free of cardiovascular disease and cancer.
• 8:07
  Davis: When a score bundles ten things together, do we learn what to change, or mostly identify who already has money, sleep, safe streets, decent food, and a doctor?
• 8:17
  Jenny: They did strengthen the case by using prospective cohorts, so the score came before the outcomes, then Cox regression, which models time until an event, and competing risk models, which account for one outcome blocking another. In UK Biobank, the median follow-up was thirteen point eight years, with six thousand one hundred thirty-seven deaths, thirteen thousand nine hundred sixty-eight cardiovascular events, and eighteen thousand seventy-seven cancer cases, and the pattern was also replicated over seven point two years in NHANES. The big limit is causal: this predicts outcomes, but it doesn't prove that improving someone's score would buy the full five or six years.
• 8:59
  Davis: As a composite health score, it's a strong dashboard, not a repair manual. For a clinic or public-health team, the useful move is to unpack which of the ten parts are actually changeable for which group, especially mental health and alcohol, so a clean prevention score doesn't just rename advantage.
• 9:17
  Paperboy.fm: This is the free version of the podcast. Subscribe at paperboy.fm to access a dozen different paper review podcasts for five dollars a month.