The prevalence of heavy drinking among people aged 65 and above (10.7%) and those aged 75 and above (7.3%) as reported in our study is much higher than those reported by other studies using similar cut off points (21 drinks per week for men and 14 for women) for heavy drinking. Primary care studies, using a similar cut off point, reported a prevalence of 4.6% among those aged 60 and above in USA (Adams et al., 1996
) and 3.4% among those aged 75 and above in UK (Hajat et al., 2004
). Our finding is similar to the highest prevalence found in an urban multi-site study conducted in Latin America (Kim et al., 2007
) which reported a wide range in the prevalence of daily drinking among older adults (from 1.5% in Mexico City to 10% in Buenos Aires). One interesting observation in our study, not directly related to our hypotheses but probably necessary in the interpretation of the findings, is the difference in proportion of educated people in the participants and the co-residents. The higher proportion of educated people in the younger co-residents as compared to the older participants is most likely a reflection of the trend of increasing literacy levels in the Dominican Republic over the years (UNESCO, 2007
Our finding of higher likelihood of psychological morbidity in co-residents of heavy drinkers compared to co-residents of abstainers or occasional drinkers extends these findings from young populations to older adults living in developing countries. Heavy drinking is likely to increase the disability associated with comorbid chronic health conditions which are common among older adults thus increasing the burden on the co-residents. Although in our study there was an independent effect of disability on co-resident psychological morbidity it did not mediate the association between heavy drinking and co-resident psychological morbidity. Behavioural and psychological symptoms have been linked to higher levels of distress in caregivers and this is further exacerbated by problem alcohol use in older adults (Sattar et al., 2007
). In our study, the older adult's severity of behavioural and psychological symptoms had an independent effect on co-resident psychological morbidity and also explained 29.1% of the total effect of heavy drinking among the elderly on co-resident psychological morbidity. Our main association was partially explained by the severity of participant's psychological and behavioural symptoms and not by disability. Some of the other mechanisms that account for psychological morbidity in co-residents of younger heavy drinkers, include non-random pairing of similar individuals (Crow and Felsenstein, 1968
), failure on the part of alcoholic family member to participate in everyday family events and their inability to relate to family members in a non-argumentative manner (Zweben, 1986
), accumulated negative life events (Homish et al., 2006
), poorer health and psychosocial functioning (Dunne, 1994; Graham and Schmidt, 1999
) and the increased risk of alcohol related violence (Cunradi et al., 1999
). Future research needs to explore these other potential mechanisms among the older adult population.
The strengths of our study lie in the large community sample of older adults, the good response rate and the use of cross-culturally validated assessments. However, the cross-sectional design of the study makes it difficult to make conclusions about the temporality of association between heavy alcohol use among older adults and psychological morbidity among their co-residents. Self-reports of alcohol consumption may not be accurate because of memory problems and difficulties in mental averaging among older persons, however, we did not find major changes in our findings when we repeated the analysis after excluding participants with dementia. We have defined heavy drinking based on ‘safe’ drinking recommendations made for younger age groups. It is quite possible that we have underestimated prevalence and estimations would be much higher if we had applied the American Geriatrics Society (Moos et al., 2004
) definition of at-risk alcohol use for over 65 year olds as, on average, more than 1 drink per day or more than 7 drinks per week. However, even if were to use this definition of heavy alcohol use, it would still be difficult to compare with previous studies considering the wide variability in measurement of drinking patterns as an outcome. Another limitation is that information about participants’ behavioural problems has been obtained from the co-resident. This introduces a potential bias as psychological morbidity (especially depression) could influence the co-residents perception and report of participants’ behavioural symptoms. We need to be cautious in interpreting our findings regarding this variable as this potential information bias might have overestimated the role of participant's behavioural symptoms on the main association. We also believe that we have improved reliability of other information obtained from co-residents by excluding co-residents with major cognitive impairments.
It is highly possible that the stress of living with and caring for elderly alcoholics is going to be further magnified in Latin American countries where there is no social security in terms of social insurance and formal social assistance (Dethier, 2007
) and the burden of caring for the elderly with alcohol related disorders is likely to fall on the shoulders of their families. This can also increase the burden on the primary health care services as relatives in these circumstances show high rates of attendance to health care services (Roberts and Brent, 1982
). Alcohol problems among older adults are under recognized and measures to increase detection, especially in primary care settings, are necessary. Simple help for problem drinking has been shown to be efficient for older adults (Fleming et al., 1999
) and should be made available in primary care settings. Early detection and intervention will not only improve outcomes among the elderly heavy drinkers, but will also reduce the burden on the relatives and the health care system.