It is always useful to plot the data before we contemplate models. This sometimes can reveal features which we may not have noticed in the data, and can warn us of potential issues. A scatterplot matrix is a good first choice for multiple regression.

[Minitab:] Select Matrix Plot. . . from the Graph menu, and then select Matrix of plots with the With Smoother option before clicking on OK. 

Enter either c1-c7 or bwt-smoke into the Graph variables text box and click on OK.

[R:] Type
pairs(bw.df, upper.panel = panel.smooth)

You should notice that there appears to be an unusual age value of 99, which we think can reasonably considered a missing value even though it is not mentioned in the data description. We should remove this point from our analysis.
[Minitab:] Hover (move the pointer with the mouse but do not click) over the the far right point in any of the plots in the age column. This should pop up a label telling you that the observation is in Row 401. Select Delete. . . from the Data menu. Enter 401 into the Rows to delete text box and enter bwt-smoke or c1-c7 into the Columns from which to delete these rows text box. Click on OK.

[R:] Type 

bw.df = subset(bw.df, age != 99)