Atrial fibrillation is associated with cognitive decline in stroke-free subjects: the Tromsø Study.

BACKGROUND AND PURPOSE: Previous studies have shown associations between atrial fibrillation (AF) and cognitive decline. We investigated this association in a prospective population study, focusing on whether stroke risk factors modulated this association in stroke-free women and men. METHODS: We included 4983 participants (57% women) from the fifth survey of the Tromsø Study (Tromsø 5, 2001), of whom 2491 also participated in the sixth survey (Tromsø 6, 2007-2008). Information about age, education, blood pressure, body mass index, lipids, smoking, coffee consumption, physical activity, depression, coronary and valvular heart disease, heart failure and diabetes was obtained at baseline. AF status was based on hospital records. The outcome was change in cognitive score from Tromsø 5 to Tromsø 6, measured by the verbal memory test, the digit-symbol coding test and the tapping test. RESULTS: Mean age at baseline was 65.4 years. The mean reduction in the tapping test scores was significantly larger in participants with AF (5.3 taps/10 s; 95% CI: 3.9, 6.7) compared with those without AF (3.8 taps/10 s; 95% CI: 3.5, 4.1). These estimates were unchanged when adjusted for other risk factors and were similar for both sexes. AF was not associated with change in the digit-symbol coding or the verbal memory tests. CONCLUSION: Atrial fibrillation in stroke-free participants was independently associated with cognitive decline as measured with the tapping test.

Serum TSH is positively associated with BMI.

OBJECTIVE: To investigate whether there is an association between serum thyroid-stimulating hormone (TSH) within the normal range and body mass index (BMI). DESIGN AND SUBJECTS: The study was performed in 6164 subjects (2813 males) who attended the fifth Tromsø study in 2001, and in 1867 subjects (873 males) that attended both the fourth Tromsø study in 1994/1995 as well as the fifth Tromsø study. MEASUREMENTS: Height, weight, and serum TSH were measured in all subjects, and smoking status was recorded. RESULTS: Smokers and nonsmokers were analyzed separately. In the fifth Tromsø study, serum TSH was positively and significantly associated with BMI in the nonsmokers. Within the normal TSH range (defined as the 2.5-97.5 percentile), nonsmoking males in the highest TSH quartile had a mean BMI 0.4 kg/m(2) higher compared to those in the lower quartile, whereas the difference for nonsmoking women was 1.4 kg/m(2). Similarly, in nonsmokers in the longitudinal study, there was a significant and positive association between delta serum TSH (serum TSH in 2001 minus serum TSH in 1994) and delta BMI in those with serum TSH within the normal range both in 1994 and 2001. In these subjects, the quartile with the highest delta serum TSH had a mean increase in BMI from 1994 to 2001 that was 0.3 kg/m(2) higher compared to those in the quartile with the lowest delta serum TSH. For the smokers, relations between serum TSH and BMI were not statistically significant. CONCLUSION: In nonsmokers there is a positive association between serum TSH within the normal range and BMI.