Thyroid dysfunction and its association with dyslipidaemia and dysglycaemia in adults: a cross-sectional study of primary-care laboratory records (2019-2025).

Abstract

To determine the prevalence of thyroid dysfunction and its association with dyslipidaemia and dysglycaemia in adults. Cross-sectional study. Primary-care laboratory records, Antalya, Türkiye, 2019-2025. Adults aged ≥18 years with a serum thyrotropin (TSH) measurement; the first eligible record per person was analysed (n=1 424 031). Thyroid function status, classified from TSH and free thyroxine as euthyroid, subclinical or overt hypothyroidism and subclinical or overt hyperthyroidism; dyslipidaemia and dysglycaemia and their association with thyroid status, assessed by χ² test, OR and multivariable logistic regression adjusted for age and sex, with E-value, calendar-year and interaction sensitivity analyses. Among 1 316 090 adults with a complete thyroid assessment, 11.1% had thyroid dysfunction: subclinical hypothyroidism 7.97%, overt hypothyroidism 0.78%, subclinical hyperthyroidism 1.85% and overt hyperthyroidism 0.49%. Dysfunction was more frequent in women and increased with age. Correlations between TSH and lipid or glucose parameters were statistically significant but negligible in magnitude (Spearman rho 0.021 to 0.046 for lipids; -0.055 for glycated haemoglobin). Dyslipidaemia prevalence rose across hypothyroid categories, from 72.8% in euthyroid to 84.8% in overt hypothyroidism. Hypothyroidism was associated with dyslipidaemia (univariable OR 1.31 (1.29-1.33); age- and sex-adjusted OR 1.29 (1.27-1.31); p<0.001) and, more weakly, with dysglycaemia (adjusted OR 1.06 (1.04-1.07)). The adjusted association was unchanged after additional adjustment for calendar year, was consistent across years, and had an E-value of 1.52. Thyroid dysfunction-predominantly subclinical hypothyroidism-was common in this tested primary-care population and was associated with dyslipidaemia independent of age and sex. The association was modest, and residual confounding by unmeasured factors such as body mass index, thyroid medication and comorbidity cannot be excluded; the findings support attention to lipid status in adults with hypothyroidism but do not establish causality.