Three routinely collected laboratory biomarkers were associated with subsequent stress-related disorder diagnosis in a retrospective cohort study published in BMC Medicine.

Researchers analyzed electronic health records from Central Finland Wellbeing Services County and found that higher hemoglobin and potassium levels were associated with lower risk of stress-related disorder diagnosis, whereas higher low-density lipoprotein cholesterol levels were associated with higher risk. The model showed only modest discrimination, limiting its potential clinical usefulness.

Study Design and Sample

The study included 73,909 patients aged 34 to 92 years: 6,758 diagnosed with a stress-related disorder and 67,151 frequency-matched controls (by sex and birth year). Patients were followed for a mean of 4.6 years.

Stress-related disorders included acute stress reaction, post-traumatic stress disorder, adjustment disorder, other or unspecified reactions to severe stress, burnout, and stress not elsewhere classified. Acute stress reaction and adjustment disorder were the most common first recorded diagnoses among cases.

Biomarkers Examined

The researchers examined 10 biomarkers collected during routine clinical care: C-reactive protein, hemoglobin, fasting glucose, glycated hemoglobin, triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol (LDL-C), creatinine, sodium, and potassium. Measurements were obtained during routine care visits.

Analytical Approach

The researchers evaluated the most recent laboratory measurement obtained within 1 year before diagnosis (or the assigned index date for controls). Analyses adjusted for demographics, comorbidity, medication use, and health care utilization.

Of the 41,514 patients with at least one biomarker measurement in the one-year lookback window, 13,928 had complete data for all exposure variables and were included in the primary multivariable analysis.

Results

In the multivariable model, three biomarkers were independently associated with stress-related disorder diagnosis. Higher hemoglobin and higher potassium levels were each associated with reduced risk, while higher LDL-C levels were associated with increased risk. Hemoglobin showed the strongest association, followed by LDL-C and potassium.

The multivariable model concordance index was 0.67, above the level expected by chance but below the 0.70 threshold often considered clinically actionable.

Temporal Trends

 Mean hemoglobin, potassium, and sodium levels declined in the months preceding diagnosis, whereas lipid markers showed smaller differences over longer periods. These pooled trends represent population-level patterns and should not be interpreted as individual patient trajectories.

Sensitivity and Subtype Analyses

Exploratory analyses of individual stress-related disorder subtypes generally showed similar patterns, although some associations were no longer statistically significant because of smaller subgroup sizes. In time-varying analyses using all available laboratory measurements, only hemoglobin remained significantly associated with stress-related disorder risk.

Proposed Mechanisms

The researchers proposed several candidate mechanisms, while emphasizing that mechanistic interpretations remain speculative in the context of an observational study.

For potassium, the authors suggested that chronic stress-related activation of the hypothalamic-pituitary-adrenal axis may stimulate aldosterone secretion via the renin-angiotensin-ldosterone system, promoting renal potassium excretion. They noted that this interpretation is consistent with prior evidence linking renin-angiotensin-aldosterone system activity to trauma exposure and post-traumatic stress disorder.

Glucocorticoid-mediated changes in lipid metabolism may explain the short-term association between LDL-C and diagnosis.

Hemoglobin, which showed the strongest and most persistent association, may serve as a broad indicator of overall health status because it is influenced by inflammation, iron metabolism, nutrition, and kidney function. However, the authors noted that the lack of an association with C-reactive protein complicates a straightforward inflammatory explanation.

Limitations

Several limitations bear on clinical interpretation. The data came from a single Finnish region and included only patients with health care contacts that produced laboratory measurements. Cases also had laboratory measurements obtained closer to the index date than controls, a pattern that may reflect increased clinical monitoring before diagnosis. The researchers identified this as informed presence bias, in which individuals with more health care contact have a higher probability of both laboratory testing and formal diagnosis. Although they adjusted for care visit frequency to mitigate this effect, residual bias cannot be excluded.

The researchers could not account for lifestyle and social factors, including diet, body composition, smoking, alcohol use, or socioeconomic position. Structured electronic health records also did not capture stressful or traumatic events, symptom onset, or the timing between stress exposure, symptoms, care-seeking, and recorded diagnosis. As a result, the observed associations may reflect preexisting physiologic vulnerability, an undiagnosed symptomatic period before formal diagnosis, health care utilization patterns, or some combination of these factors.

Given the study’s observational design, modest model discrimination, and lack of external validation, the findings are best interpreted as hypothesis-generating rather than practice-changing. The study does not establish that hemoglobin, potassium, or LDL-C can predict, diagnose, or screen for stress-related disorders in individual patients, nor does it establish that changing these biomarker levels would alter stress-related disorder risk.

Conclusions

The findings do not indicate that specific hemoglobin, potassium, or LDL-C thresholds should prompt stress-related disorder screening, nor do they support changes in laboratory ordering or interpretation. Instead, the researchers said the results may help identify physiological domains for future investigation into the somatic dimensions of stress-related disorders, including hematologic status, electrolyte and renin-angiotensin-aldosterone system regulation, and lipid metabolism.

The study was led by Anna A. Peltola, of the Faculty of Medicine and Health Technology at Tampere University, and colleagues.

Funding was provided by the Juho Vainio Foundation, Yrjö Jahnsson Foundation, Research Council of Finland, Samfundet Folkhälsan, Signe and Ane Gyllenberg Foundation, Instrumentarium Foundation, and Sigrid Jusélius Foundation. Open-access funding was provided by Karolinska Institutet. The researchers reported no competing interests.

Source: BMC Medicine