A natural, but non-causal, multicollinearity between brand sales and the economy.  As a brand matures its growth typically slows.  Recent deterioration in the economy will often have a correlation with product sales that might have slowed anyway, regardless of economic trends.

Change in the economy over the past few years is not much different than a simple trend line, thus limiting the data variance needed to develop a good analytic model.  But the analysis should be focused on the recent trends because any long-term historical relationship between sales and the economy has probably changed in the evolving marketplace.

Time series models create autocorrelation issues that bias the estimated impact of the economy.

Many other causal factors, such as FDA decisions, sales force promotional effort, seasonality, price, etc., simultaneously impact sales and need to be separated from the effect of the economy.