Drinking alcohol to excess is one of the largest modifiable causes of morbidity in middle-aged adults with a third of drinkers aged 45-64 years drinking above the UK low-risk guidelines (14+ units per week). Social-ecological frameworks suggest that alcohol behaviour is driven by both individual factors and wider level contexts. Evidence gaps for understanding risk alcohol use in this age group were identified for trends in intake, individual, transitional and regional associations and an age-relevant understanding of income and alcohol expenditure. The current middle-aged literature was found to be limited by inconsistent risk alcohol intake thresholds, a lack of late middle-aged adults in birth cohorts, and lack of adjustment for the wider determinants of health. This thesis addressed these research gaps using three independent but linked studies. The first study used cross-sectional data from the Health Survey for England (HSE) from 1998 (20,871 n pooled) to consider trends and demographic, socioeconomic and co-risk factors associated with exceeding the former UK guidelines (21/14+ units per week) and binge drinking (8/6+ units on heaviest day). The second study used two waves from Understanding Society (USoc), a UK panel study (12,737 n), to determine if transitions in socioeconomic or social status between waves were associated with binge drinking patterns. The final study used USoc longitudinally over 7 waves (17,407 households) to assess the relationship with income and household composition factors associated with expenditure on alcohol. Each study used multilevel regression modelling to account for shared characteristics of individuals within regional living areas, as Government Office Regions (GOR), and account for the nonindependence and correlation in repeated measures panel data. Non-drinkers were excluded from each study to avoid abstainer bias. Data from the HSE found a gender convergence in exceeding the former weekly guidelines with binge drinking increasing in both sexes since 1998. Exceeding the guidelines was associated with smoking, higher income and education, retirement, cohabiting marital status and having no children in the household for both sexes. Binge drinking was associated with smoking, higher body-mass index, higher income, divorced marital status, no religious belonging and urban residence in both sexes with associations for having friends and no educational qualifications in men only. Entering or leaving a relationship were both associated with maintaining binge drinking over time in men compared to a stable relationship status but not in women. Moving to a rural area in both sexes, changing to no reported friends in men only, becoming a non-smoker in women only and entering employment in women only, were all negatively associated with maintaining binge drinking. Living in the North East and North West of England contributed to increased binge drinking in men independently of individual characteristics and was associated with higher household expenditures on alcohol. A £100 increase in income was associated with a £0.40 increase in alcohol expenditure in middle-aged headed households, accounting for household composition factors including household size, child status, sources of income and housing tenure. Overall, this thesis contributes to the literature on excessive alcohol use by providing age and sex-specific analyses of a previously under-researched age group, discovering multiple risk factors associated with acute and chronic risk from alcohol use and spending on alcohol. These findings make use of multilevel modelling to account for regional contexts and individual trajectories of binge drinking and household spending over time, demonstrating the utility of secondary data to explore changing health behaviours.