Randomised controlled trials (RCTs) or prospective observational studies that reported morbidity or mortality data for comparisons of thyroid hormone therapy were eligible for inclusion. Studies used either triiodothyronine (T3) or thyroxine (T4). Therapy needed to be administered for at least 24 hours and compared to placebo in euthyroid adults who were either obese and undergoing calorific deprivation (<1,000kcal per day) or diagnosed with acute or chronic non-thyroidal illnesses.
Studies were excluded for use of before-and during-treatment study designs or if they were of adults who received thyroid hormone therapy perioperatively for less than 24 hours. Also excluded were studies of patients with primary diagnoses of infertility, premenstrual syndrome, psychiatric disorders, hyperlipidaemia, chronic urticaria and angioedaema, supraventricular tachycardia, bronchial asthma, hypertension or Raynaud's phenomenon.
Most studies assessed T3 in obese adults (as defined by study authors; further details were reported in the review). Other participants included those diagnosed with acute renal failure, burns, congestive heart failure or patients in intensive care units or who received coronary artery bypass grafts, which were mainly treated by T4. Patient age varied from 17 to 69 years. Studies included male and female participants. The synthetic thyroid hormones T3 and T4 were used at a range of doses in the included trials. For all included patients, serum thyroid stimulating hormone (TSH) and TSH response after T3 or T4 administration were evaluated.
Additional outcomes measured in obese patients were total weight loss, urinary 3-methylhistidine excretion, nitrogen balance, resting metabolic rate, and pulse rate arrival time. For patients with non-thyroidal illnesses, outcomes assessed were heart rate, cardiac output, systemic vascular resistance, morbidity and mortality.
Three reviewers reviewed studies for inclusion. Any disagreements were resolved by discussion.