I would be surprised if any correlation were spurious (assuming that the correlation is highly statistically significant and thus not a product of chance).