In this work, we have used a continuous variable as an IV, representing a genetic score comprised of multiple variants. Such a score may be problematic to obtain for exposures with few genetic determinants which are valid IVs. However, we have previously shown that the first-stage R2 is the key parameter influencing power, regardless of what type of IV is used (i.e., single or multiple IVs; dichotomous, ordinal, or continuous IVs). Thus, our findings for a given first-stage R2 will apply to any type of instrument, be it continuous or discrete, or a set of multiple instruments.
