When Sir Walter Scott wrote "Marmion" in 1808, he told a story of love, lust, honor, and deception in the setting of war between the kingdoms of England and Scotland. After reading the work by Flythe et al.(1) entitled "Disentangling the ultrafiltration rate-mortality association: The respective roles of session length and weight gain," one cannot help but recall the poem's most memorable verse: "Oh! What a tangled web we weave." Although few would consider ultrafiltration with hemodialysis as complex as the affairs of the heart, the factors that contribute to optimal ultrafiltration are more intricate than the plot lines of most gothic romantic novels or modern soap operas.In their article, Flythe et al.(1) entertain the reader, as did Scott, with a carefully crafted story. Flythe and colleagues note that attempting to disentangle related factors (i.e., dialysis session length and interdialytic weight gain) can be difficult or arguably, nearly impossible, when these factors are highly correlated. Therefore, they analyzed data on 14,643 randomly selected prevalent adult patients on three-times-per-week hemodialysis and examined associations between dialysis session length, interdialytic weight gain, and mortality in parallel. In other words, for analyses examining the association between dialysis session length and mortality, patients were matched on a narrow range of interdialytic weight gain (60.25 kg) along with other key covariates. For analyses examining the association between interdialytic weight gain and mortality, patients were similarly matched on the same covariates (age, sex, vascular access, and postdialysis weight) along with dialysis session length rounded to the nearest 5-minute increment. Matched pairs were very well balanced on the matched factors (as expected); on other factors, including race, diabetes mellitus, and heart failure, there were important imbalances.In their primary analyses, Flythe et al.(1) reported a 32% increase in the relative mortality hazard (95% confidence interval, 3% to 69%) associated with dialysis session lengths less than 240 minutes relative to sessions 240 minutes or more and a 29% increase in the relative mortality hazard (95% confidence interval, 1% to 65%) associated with interdialytic weight gains of more than 3 kg relative to 3 kg or less. Unlike previous reports that either examined one of two parameters or examined both simultaneously in multivariable regression models, the matching approach used by Flythe et al.(1) allows us to more confidently conclude that dialysis session length and interdialytic weight gain are independently associated with mortality. We agree with the contention of Flythe and colleagues that an association between shorter dialysis session length and mortality is "highly plausible" (1). As session lengths shorten, removal of plasma water will, at some point, exceed the rate at which water can be recruited from the interstitial and/or intracellular spaces. Hypotension may develop, or BP may be sustained by heightened sympathetic and othe...