Get access

A cross-sectional study of the association between circulating TSH level and lipid profile in a large Spanish population

Authors

  • Silvia Santos-Palacios,

    1. Department of Endocrinology and Nutrition, Clínica Universidad de Navarra, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Antonio Brugos-Larumbe,

    1. Department of Health Sciences, Public University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Francisco Guillén-Grima,

    1. Department of Preventive Medicine, Clínica Universidad de Navarra, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Juan C. Galofré

    Corresponding author
    1. Department of Endocrinology and Nutrition, Clínica Universidad de Navarra, University of Navarra, Pamplona, Spain
    • Correspondence: Dr Juan C. Galofré, Department of Endocrinology and Nutrition, Clínica Universidad de Navarra, Pio XII 36, 31080 Pamplona, Spain. Tel.: +34 948 255 400; Fax: +34 948 296 500; E-mail: jcgalofre@unav.es

    Search for more papers by this author

Summary

Objective

Some evidence suggests that high serum TSH levels are associated with an adverse lipid profile, but this association is not clear when plasma TSH is within the reference range. Nevertheless, these studies have never been conducted in Spain, a country with a strong adherence to the Mediterranean diet. The study aim was to analyse the association between blood TSH levels and circulating lipids in a large Spanish population and set up a TSH reference range in different age, gender and Body Mass Index (BMI) subpopulations from our cohort.

Design

Cross-sectional study on 20 783 subjects.

Patients

We analysed circulating levels of total cholesterol (TC), low-density lipoprotein cholesterol (LDLc), high-density lipoprotein cholesterol (HDLc) and triglycerides (TG) and compared them with TSH serum levels. Discriminant function analysis was used to determine the TSH cut-off level from where hyperlipidaemia developed.

Results

In individuals free of thyroid dysfunction, the mean and the 95% TSH (mU/l) reference limits were 2·20 and 0·72–4·43, respectively. We observed a sex-related difference in TSH concentration (men, 2·07 and 0·72–4·29; women, 2·29 and 0·72–4·49; < 0·01). We also observed a weight-related difference in TSH concentration (BMI < 30 kg/m2, 2·16 and 0·72–4·39; BMI ≥ 30 kg/m2, 2·28 and 0·71–4·47; < 0·01). TSH was positively associated with TC, TG and LDLc levels and negatively with HDLc.

Conclusion

We found an association between TSH and lipids in that as TSH increased, the lipid profile became less favourable, even within the normal range. Additionally, TSH reference ranges varied according to gender, age and BMI.

Ancillary