When statistics are “adjusted” it means they are considering different variables or confounders that might have an impact on the results (male age, female race, BMI, grade of the embryo, sperm origin, etc.). These variables are controlled for and the statistical analysis is repeated to see if the same results stand.
For example, let’s say we’re looking at the impact of ICSI vs conventional IVF and live birth outcomes among women who are aged 30-45. We find that ICSI has worse live birth rates overall, but more older women did ICSI compared to younger women. Since older women have a lower chance for live birth in general, maybe this is what’s lowering the statistic for ICSI. So we need to control for age.
We repeat the statistical analysis, but this time instead of just looking at ICSI vs IVF across all ages, we look at ICSI vs IVF in younger women, and ICSI vs IVF in older women. By splitting the data up into two age groups and then performing the statistical analysis, we’re controlling for age or “adjusting” the statistical analysis to control for age. If there still is a decrease in the older group, then we know that older women who perform ICSI vs IVF show a decrease in live birth rates with ICSI. But let’s say there is no difference — then there’s no difference in ICSI vs IVF for both age groups!
It’s important to do these statistical adjustments to control for differences in the baseline characteristics, especially ones that are known to have an effect. In this example I’ve provided, the initial result was that ICSI lowered live birth rates, but after controlling for age we see that isn’t the case.