Most research on speech rhythm has focussed on duration. For example, [1] suggested the normalised Pairwise Variability Index for vocalic intervals (nPVI-V) in order to measure the variability of vocalic durations. This paper argues that speech rhythm research should also take into account other correlates of prominence as well as their interaction. The duration-based nPVI, or nPVI-V(dur), is supplemented by an nPVI-V(avgLoud) that measures variability in average loudness. These two metrics account for variability in duration and loudness, but cannot measure if loudness and duration reinforce each other by varying simultaneously in the same direction. This simultaneous variability is accounted for by the combined nPVI-V(dur+avgLoud), which is higher than the average of the other two measures, if vocalic intervals that are longer than average are also louder than average. The three metrics are subsequently applied to recordings of a reading task performed by 20 speakers of Indian English (IndE) and 10 speakers of British English (BrE). Results indicate that IndE has less variability in duration and less variability in loudness than BrE. In addition, IndE has less simultaneous variability in duration and loudness than BrE. This indicates that duration and loudness are less often used together as cues to prominence in IndE than in BrE.