Effect of thyroid substitution on hypercholesterolaemia in patients with subclinical hypothyroidism: a reanalysis of intervention studies

OBJECTIVE: The significance of mild hypercholesterolaemia in subclinical hypothyroidism and whether there is beneficial reduction after thyroxine replacement, remain controversial. We aimed to describe the association between hypercholesterolaemia and subclinical hypothyroidism, and to quantify the effect of thyroid substitution therapy by an analysis of previously published intervention studies. DATA SOURCES: Intervention studies cited in the Medline database from January 1976 until January 1995, with index terms cholesterol, hypercholesterolaemia, hyperlipidaemia, thyrotrophin (TSH), hypothyroidism, thyroid and human. A total of 148 studies were reviewed. DATA EXTRACTION: We recorded the year of publication, study design, number of patients enrolled, mean age, duration of thyroid substitution, normal range of TSH levels, TSH levels pre and post-substitution treatment and total cholesterol in plasma before and after treatment. DATA ANALYSIS: (1) Qualitative description of studies on the relationship between hypercholesterolaemia and hypothyroidism, both subclinical and clinical. (2) Precision weighted pooled estimates of the effect of thyroid substitution therapy on the plasma levels of total cholesterol, in patients with subclinical and overt hypothyroidism. RESULTS: Subclinical hypothyroidism was two to three times more frequent in people with an elevated total plasma cholesterol. In addition, the total plasma cholesterol levels were slightly elevated in patients with subclinical dysfunction of the thyroid. Thyroid substitution therapy in patients with subclinical hypothyroidism, restoring the TSH levels to normal, decreased total cholesterol by 0.4 mmol/l (95% confidence interval (Cl) 0.2-0.6 mmol/l) independently of the initial plasma level. The effect of thyroid substitution therapy on HDL-cholesterol in patients with subclinical hypothyroidism was not consistent. The effect of thyroid substitution in patients with overt hypothyroidism was highly dependent on the pretreatment levels of total cholesterol. In these patients substitution therapy decreased total cholesterol by 1.2 mmol/l (95% Cl 0.9-1.5 mmol/l) when the plasma levels were elevated up to 8 mmol/l, and by 3.4 mmol/l (95% Cl 3.0-3.7) when plasma levels were higher than 8 mmol/l. The high density lipoprotein (HDL)-cholesterol level decreased and amounted to 0.16 mmol/l (95% Cl 0.07-0.24). CONCLUSIONS: Thyroid substitution treatment in patients with hypercholesterolaemia and subclinical hypothyroidism decreases total plasma cholesterol by 0.4 mmol/l, but plasma levels remain elevated in most patients. Further treatment with dietary restriction and cholesterol synthesis inhibitors should then be considered.