Health Science Hub
Health Science Hub

Evidence-Based Reviews & Longevity Research

Which Health Habits Are Most Strongly Linked to Future Anxiety and Bipolar Disorder?

Health Science Hub Editorial

Evidence review & editorial synthesis

Can the same everyday factors used to judge heart health also tell us something about future mental health? A new study of 242,947 UK Biobank participants suggests they can. Over a median 13.8 years, people belonging to the least healthy of five cardiovascular-health profiles developed anxiety and bipolar disorder substantially more often than people in every other profile. The intriguing part is what happened when researchers looked beyond a single health score: poor metabolic health alone did not appear to explain the pattern.

Eight measures, five very different health profiles

The researchers used the American Heart Association's Life's Essential 8 framework. It combines four behavioral measures, diet, physical activity, nicotine exposure and sleep, with four biological measures, body mass index, blood lipids, blood glucose and blood pressure. Each is modifiable to some degree, although biology, medication, socioeconomic conditions and disease can strongly influence them.

Instead of simply averaging those eight measures into one score, the investigators used latent profile analysis to find recurring combinations in the population.

Five groups emerged. One had generally low scores across the board. Another was dominated by obesity, elevated blood pressure and poor glucose control. A third was distinguished by low physical activity and poor sleep. A fourth had high blood pressure and dyslipidaemia but relatively favorable behavior scores. The final group performed well across nearly all eight measures.

That distinction matters because two people can receive similar total scores for very different reasons. One may exercise, sleep well and avoid smoking but have hypertension and unfavorable cholesterol. Another may have relatively good metabolic measurements while being inactive and sleeping poorly.

The risk differences were large

During follow-up, researchers recorded 11,419 new anxiety diagnoses and 341 bipolar disorder diagnoses.

Compared with the lowest-LE8 group, anxiety hazard was 23% lower in the adiposity, blood-pressure and glucose profile, 22% lower in the low-activity and poor-sleep profile, 32% lower in the high-blood-pressure and dyslipidaemia profile, and 35% lower in the high-LE8 group. The corresponding adjusted hazard ratios were 0.77, 0.78, 0.68 and 0.65.

The separation was even larger for bipolar disorder. Relative to the lowest-health profile, hazard was 49% lower in the adiposity, blood-pressure and glucose group, 59% lower in the low-activity and poor-sleep group, 63% lower in the high-blood-pressure and dyslipidaemia group and 59% lower in the high-LE8 group.

That produces a counterintuitive result. The group with the best metabolic profile was not uniquely protected. People with elevated blood pressure and poor lipid scores still had among the lowest psychiatric risks in the study.

So which unhealthy factors look most important?

This study cannot cleanly rank the eight components because the exposure was a profile, not an experiment that changed one variable at a time. But the contrasts provide useful clues.

The lowest-health group had a mean LE8 score of about 56 and performed poorly across most dimensions. Its nicotine score was especially striking: approximately 0 on a 0-to-100 scale, reflecting near-universal current smoking. By comparison, profiles defined mainly by metabolic problems retained healthier behavior patterns and had much lower psychiatric risk.

That makes it difficult to explain the findings as a simple story about blood pressure, glucose, cholesterol or weight. Preserved behaviors such as not smoking, remaining active and maintaining adequate sleep may be especially relevant, although the current analysis cannot tell us how much each contributes independently.

A separate prospective UK Biobank analysis of roughly 400,000 people points in the same direction. When investigators divided LE8 into health behaviors and health measurements, deterioration in the behavioral domain showed a stronger association with later depression or anxiety than deterioration in the biological health-factor domain.

Physical activity has independent support as well. A meta-analysis of prospective cohorts found physically active adults had lower subsequent odds of anxiety symptoms and anxiety disorders, although the evidence varied in quality and cannot eliminate all confounding.

Sleep may be particularly important around bipolar illness. A systematic review found that disturbed sleep can precede bipolar disorder by months or years and may appear before first mood episodes. That does not mean poor sleep necessarily causes bipolar disorder. In some people, sleep disruption may instead be an early manifestation of vulnerability or emerging illness.

The pattern may matter more than any single habit

The strongest interpretation is therefore not that scientists have found one lifestyle cause of mental illness. It is that mental and physical health appear to cluster much more tightly than traditional organ-by-organ thinking suggests.

The least healthy profile combined several exposures at once. Smoking, poor diet, inactivity, sleep problems and adverse metabolic health can interact through inflammation, vascular function, stress physiology, circadian rhythms and other pathways. They can also share underlying causes such as deprivation, chronic stress, medication effects or early disease.

That complexity is why the profile approach is useful. It asks what real people look like when multiple favorable and unfavorable factors occur together rather than pretending every component contributes equally.

The result also fits broader brain-health research. In a separate prospective study using both UK Biobank and the US All of Us cohort, poor LE8 health was associated with roughly double the adjusted risk of a composite outcome including stroke, dementia and late-life depression compared with optimal health.

What is actually actionable?

The findings do not justify telling people that optimizing eight numbers will prevent bipolar disorder or anxiety. Genetics, early development, trauma, social conditions and many other influences remain central to psychiatric disease.

But they strengthen a more practical idea: sleep, regular movement, avoiding smoking, nutritious food and treatment of hypertension, abnormal glucose and lipids should not be viewed only as heart-health measures. They may be part of a broader brain-and-mental-health strategy too.

And the new study adds an important twist. You may not need perfection across every biomarker before healthier behavior matters. Some groups with substantial metabolic problems still had markedly lower psychiatric risk than the group doing poorly across nearly everything.

The next question is therefore more interesting than whether a high total LE8 score is good. Researchers now need to determine which components are genuinely driving these differences, which are early warning signs rather than causes, and whether improving specific combinations of sleep, activity, nicotine exposure and metabolic health can actually change psychiatric outcomes.

Scientific References & Further Reading

Disclaimer: This article is for informational purposes only and does not constitute medical advice. It is not a substitute for diagnosis, treatment, or individualized medical care.

Related Research