Reference for max vars in a model

I used 7 predictor variables in a regression model with a sample size of 103. Someone who reviewed my manuscript said that was too many predictors.

I thought the rule of thumb was that you should use no more than 1 predictor variable for every 10 observations. For example, no more than 5 predictors where N=50.

I have to defend my model to the reviewer. Does anyone know of a book or article I can reference regarding the appropriate number of predictors to use relative to sample size??



Ambassador to the humans
It's more of a rule of thumb. Really what you want is a decent number of observations per predictor. I typically try to have around 20 observations per predictor but this clearly isn't always possible. The reviewer sounds to be more in this camp though.
How else can I defend my model?

I had 7 predictors that were p<0.20 in bivariate analysis; after multivariate adjustment, only 3 were significant at the p<0.05 level.

R2 = 0.025
F = 4.59 (p<0.001)

I'm not sure how to interpret the R2 & F statistic. Do they support the validity of the model?

Thanks again!!