I'm running a growth mixture model using four waves of data (one year in between each wave) to model different patterns of sexual intercourse frequency over time among couples in a stable relationship, that is these couples stayed together across all four waves. At each wave respondents (N = approximately 1,300) were asked how long they have been together with their partner (relationship duration ranges from 1 month to 31 years). I'm wondering if there is a way to model intercourse frequency by relationship duration instead of wave. I have tried to reorganize the data into an accelerated longitudinal design (i.e., instead of intercourse frequency data at each wave for each respondent, I have frequency at four consecutive relationship duration years for each respondent; e.g., respondent A has data for 1, 2, 3, 4 years into a relationship vs. respondent B has data on 9, 10, 11, 12 years into a relationship depending on how long they had been in their relationship at Wave 1) but that leads to too much missing data because of the wide range of relationship duration. I hope my description make sense to you. I would very much appreciate your input. Thank you.