Does C‐reactive protein identify a subclinical metabolic disease in healthy subjects?