The Depression That Hides a Bipolar Trajectory: Disordered Eating as an Early Marker
- Across 338 patients with major depressive disorder recruited at seven Russian centres, higher disordered-eating scores tracked with an earlier age of depression onset and a longer worst depressive episode.
- Disordered-eating severity correlated positively with Hypomania Checklist scores, a recognised signal that a depression may belong to the bipolar spectrum rather than be purely unipolar.
- Patients with more disordered eating also reported more suicidal ideation in the previous month and higher trait anxiety, marking a heavier overall clinical burden.
- The authors read this cluster of earlier onset, longer episodes and hypomanic features as latent bipolarity, that is, an elevated risk that the diagnosis will eventually shift to bipolar disorder.
A patient arrives with what looks like ordinary depression. The clinician's central, often unanswerable question is whether this is unipolar major depression or the depressive face of a bipolar disorder that has not yet declared itself through a manic or hypomanic episode. Getting this wrong is costly: an antidepressant given without a mood stabiliser can destabilise an unrecognised bipolar course. This study, led by a team at the V. M. Bekhterev National Medical Research Center in Saint Petersburg, asks whether something as mundane as eating behaviour can serve as a flag for that hidden bipolar trajectory.
The design is a cross-sectional clinical survey. Researchers recruited 338 adults (216 women, 122 men) with a confirmed diagnosis of major depressive disorder from both outpatient and inpatient settings at seven centres across Russia, capturing patients in exacerbation and in remission alike. Each diagnosis was anchored with the Mini International Neuropsychiatric Interview. The instruments are familiar to any mood clinic: the Eating Attitudes Test (EAT-26) for disordered eating, the Montgomery-Asberg scale for depression severity, the Hypomania Checklist for soft bipolar signs, the Snaith-Hamilton scale for anhedonia, the Columbia scale for suicide risk, and standard anxiety inventories.
The pattern that emerged is clinically legible. Patients with higher EAT-26 scores tended to have become ill younger, to have endured a longer worst depressive episode, and to drift toward hyperphagia rather than appetite loss during their low periods. Disordered eating rose in step with Hypomania Checklist scores (ρ = 0.346, p < 0.001), with recent suicidal ideation (ρ = 0.146, p = 0.008), and with trait anxiety (ρ = 0.198, p = 0.003). None of these correlations is large in isolation. Taken together, though, they sketch a recognisable phenotype: the atypical, early-onset, agitated depression that long clinical tradition has linked to the bipolar end of the mood spectrum.
The conceptual move worth dwelling on is the framing of latent bipolarity. The authors are not claiming that disordered eating causes bipolar disorder, nor that EAT-26 is a diagnostic test for it. They are arguing that within a population labelled unipolar, a subgroup carries the clinical fingerprints of a bipolar course that has not yet surfaced, and that disordered eating is one accessible window onto that subgroup. For the practitioner, the actionable inference is modest but real: when a depressed patient also reports a turbulent relationship with food, especially with overeating in low phases, it is worth widening the assessment to hypomanic history and family history before reaching for an antidepressant alone.
Why eating behaviour belongs in a mood workup
Appetite and eating are not peripheral to mood disorders; they are written into the diagnostic criteria. What this study adds is that the direction and texture of disordered eating may carry diagnostic information beyond mere presence or absence. Hyperphagic, weight-gaining depression has a different prognostic weight than the classic anorexic, insomniac picture, and this cohort ties the former to soft bipolar features.
What a correlation can and cannot tell us
Because the design is cross-sectional, it photographs associations at a single moment and cannot show that any of these patients will actually convert to bipolar disorder. The correlations are statistically reliable but modest in magnitude, which means disordered eating is at best a prompt to look harder, never a verdict. The value here is hypothesis-sharpening, not prediction.
When a depressed patient's relationship with food is as turbulent as their mood, the diagnosis may not yet be telling the whole story.
The study is cross-sectional, so it cannot establish that any patient will progress to bipolar disorder. The reported correlations, while statistically significant, are modest in strength and rest on self-report instruments. There is no longitudinal follow-up to confirm that the proposed latent bipolarity actually emerges over time.