Strengths and limitations
Our study is not without limitations. This is a retrospective convenience sample and prone to confounding bias, while every effort was made to control for potential confounding factors by implementing multivariable logistic regression analyses to estimate the association between gestational weight loss and pregnancy outcomes, there may be residual variables for which we did not have information. Furthermore, given the study design, we acknowledge that some outcomes are underpowered and that a higher number of study participants might reveal stronger associations.
Another study limitation is the procedure for assessment of patient’s weight. At our centres, equipment is calibrated on a regular basis. Due to the retrospective nature of the study it is difficult to guarantee the standardisation of the weighing process.
Moreover, we do not know the reasons for weight loss in this population. It is possible that some women lost weight because of factors that placed them at risk of adverse outcomes, but we are unable to fully explore this. We assumed that women who lost/gained insufficient weight likely made lifestyle changes subsequent to their GDM diagnosis. We acknowledge that further investigation is required to clarify the reasons for weight loss in this population. In addition, due to the observational nature of this data, we cannot state a causal relationship between gestational weight loss and adverse outcomes. Since it is not possible to perform a randomized controlled study of weight loss/insufficient GWG versus adequate GWG (as recommended by the IOM), we believe that the current analysis provides valuable information that may be useful in caring for this high-risk population.
Lastly, we used HbA1c as a measure of glycemic control as individual readings were recorded in a paper-based diary retained by the study participant and thus not available to the research team.