Is the study racist in and of itself? Of course not.
Were the people who completed the poll racist? Perhaps. It's hard to tell. In the absense of any further information, it's feasible that white women scored higher simply because more white men took part in the poll, or that the samples voted on were skewed.
Are his conclusions racist? Perhaps. He forms conclusions as to why he's got the results he has, and we can't tell solely from the BBC article as to whether that's anything other than concjecture. He's essentially saying that black women are fatter and more masculine than white women, which is a pretty inflammatory thing to be claiming without some serious evidence to back it up, and I'd perhaps argue that an average score of 3.5 against one of 3.7 isn't all that statistically significant (assuming that it's a well-conducted study at all, which I've not bothered to look into).