As a post-doc I am working on some data that I did not collect myself. The central question I'm trying to answer: is there a difference in the fat mass of neonates born to mothers with gestational diabetes mellitus (GDM), versus mothers with normal glucose tolerance?

I thus want to create a model where fat mass is the dependent variable, and GDM (coded 0 for mothers without GDM, and 1 for those with) is the independent variable. I also need to control for other variables such as gestational weight gain, pre-pregnancy BMI, gestational age, etc., as these could potentially confound the relationship between neonatal fat mass and GDM.

Sample size is 72 for GDM mothers/babies and 211 for non-GDM mothers/babies. I'd also like to run the model taking out all missing data for comparison, which would leave me with 68 GDM mothers/babies and 112 non-GDM mothers/babies.

I know I need to consider the sample size when determining how many IVs to add to the model, and I've searched online and found some general rules of thumb for how many IVs are appropriate in multivariable regression with a given sample size. It appears that the data set I'm working on is too small to accommodate the number of potential confounders, especially for the smaller diabetes sub-sample.

Therefore, I'm stuck, as I cannot do what I would like to do - collect more data - and I have about 10 IVs that I think should be added into the model based on their potential to confound the association of interest.

Another thing to note: I recently read a paper discussing how it is inappropriate to test bivariable associations of potential IVs with your outcome of interest in order to determine whether they should be added into the model (based on their bivariable significance). So I've already ruled that out as an option for determining which IVs should make it into the model.

Does anyone have any advice on how to determine which/how many IVs to add that goes beyond what I've already considered?

Thanks so much.


I would say your sample isn't that small and can definitely accommodate 10 predictors if entered without interactions with the treatment. If this really concerns you, you can estimate a propensity score as the predicted probability of GMD given the predictors and simply fit a model regressing the outcome on GMD and the propensity score (though it would be important to allow this model to be flexibly estimated).

If you're worried some of your predictors are not true confounders, you can use the group lasso with doubly robust estimation described in Koch, Vock, & Wolfson (2018), which estimates a doubly robust treatment effect (robust to misspecification of either the propensity score model or outcome model) and uses lasso to select only the relevant covariates (so the models aren't overly saturated). This was specifically designed to select only the covariates from a large pool of potential confounders that are helpful in estimating the treatment effect.

For missing data, it would inappropriate to drop cases. You should use multiple imputation or FIML estimation to retain your original sample size. It does not significantly complicate the problem to use either of these methods (though FIML cannot be used with propensity scores).

  • $\begingroup$ +1 Just want to emphasize the importance of not dropping cases with missing data in this circumstance. That would omit nearly half of the non-GDM cases, and I suspect that the missing data might be from tests that aren't routinely done in uncomplicated pregnancies. You might want to add a link to FIML, as that might not be so broadly appreciated as multiple imputation. $\endgroup$
    – EdM
    Apr 12 '19 at 19:34
  • $\begingroup$ Thanks for this! Following Stef van Buuren's book on multiple imputation, I was able to carry this out, and ran my fat mass ~ GDM + confounders model, getting very similar results to 1) the regression run on original data set with missing values and 2) data set following list-wise deletion. $\endgroup$
    – Meghan.S
    Apr 15 '19 at 21:08
  • $\begingroup$ I would like to report the results of the multiply imputed dataset in the paper, but I can only see how to run the regression model with the imputed data - is there a way to use the imputed data set to do the other tests I'd like to report, e.g. t-tests and Pearson correlations? $\endgroup$
    – Meghan.S
    Apr 15 '19 at 21:10
  • $\begingroup$ T-tests and Pearson correlations are just regression. A t-test is a regression of a continuous variable on a binary variable. A Pearson correlation is just a standardized regression coefficient from a single regression. $\endgroup$
    – Noah
    Apr 16 '19 at 4:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.