Failure to control hypercholesterolaemia in the Irish adult population: cross-sectional analysis of the baseline wave of The Irish Longitudinal Study on Ageing (TILDA)

The study reports a weighted mean TC of 5.1, 4.8 mmol/L in men and 5.3 mmol/L in women and a weighted mean LDL-C of 2.9, 2.7 mmol/L in men and 3.0 mmol/L in women in a nationally representative sample of adults aged 50–79 years in the Irish population during the period 2009–2011. One-third of adults in the age group were on lipid-lowering medication. In a subgroup aged 50–64 years, the prevalence of hypercholesterolaemia was 73%. Despite European and National policy recommendations and guidelines regarding the risks associated with high cholesterol [9, 16], control of LDL-C in those at highest risk for a CVD event is lower than expected with 57% of those with known CVD and 49% of diabetics achieving guideline targets. Lack of awareness of hypercholesterolaemia was high across all primary prevention risk groups.

These results compare with findings from a nationally representative cohort of non-institutionalised adults aged 20 years and older in the National Health and Nutrition Examination Surveys (NHANES) in the United States which found an overall TC of 196 mg/dL (equivalent to 5.0 mmol/L) and LDL-C of 116 mg/dL (equivalent to 3.0 mmol/L) during the time period 2007–2010 [17]. The Health Survey for England 2011 conducted in a community living sample of adults aged 16 years and older found a mean TC of 5.1 mmol/L in men and 5.2 mmol/L in women [18]. Differences in the age profile of participants in other studies creates difficulty in making direct comparisons, however, certain patterns in the data are similar to those found elsewhere. Women had higher cholesterol levels than men at all ages beyond 50 years in our study. This is similar to other studies of younger cohorts which found that up to the age of 50 years men had higher cholesterol levels than women, however, once over the age of 50 years women have higher cholesterol levels [19]. The finding that cholesterol levels peak higher and later in women compared to men is consistent with findings from representative surveys of the French population in 2006/2007 [20] and the Health Survey for England 2011 [19]. This unique rise in cholesterol in peri-menopausal women is also documented in the Study of Women’s Health across the Nation (SWAN) in the United States which highlights the importance of monitoring lipids in older women [21].

Previously in Ireland, the Kilkenny health project conducted TC examinations at baseline on an intervention and reference group (aged 35–64 years) in 1985–1986. Mean total cholesterol in men was 6.0 mmol/L in the intervention and in the reference group and 6.0 mmol/L in women in the intervention group and 5.9 mmol/L in the reference group [22]. While our study examines TC in an older cohort (50–79 years) it identifies a substantial decline in TC of approximately 0.9 mmol/L in the population over a 25 year period and may reflect a survivorship effect. This decline exceeds the 0.2 mmol/L decline in TC per decade in the European region as outlined by Farzadfar et al. [7] and is similar to results from the northern Sweden MONICA study between 1994 and 2014, where cholesterol levels decreased by 0.7 mmol/L in the 20 year period [23]. Previous research on a sample of 1207 adults aged 45 years and older taken from a nationally representative sample of adults in Ireland found that 18% of the population had a TC <5 mmol/L and were not on lipid-lowering medication [24]. Our study found an increase in this proportion to 22%. This may be partly due to our larger sample size, slightly older age profile or an improvement in dietary and lifestyle factors in the population during the intervening time period.

The use of cholesterol-lowering medication in this cohort (33%) was slightly higher than that reported by the National Health and Nutrition Examination Survey (NHANES) during 2011–2012 when 28% of adults aged 40 years and older were reported using prescription cholesterol-lowering medication in the previous 30 days [25].

The findings in the subgroup aged 50–64 years present an opportunity to examine interventions aimed at lowering cholesterol according to an individual’s CVD risk profile. In individuals at high risk of future CVD events i.e. those with existing CVD or diabetes, opportunities are being missed to achieve LDL-C control as recommended in clinical guidelines. These findings corroborate those found in an observational cohort of Irish patients during the same time period which found persistent dyslipidaemia despite statin therapy [26]. Our findings also reveal a low level of treatment and control and lack of awareness in those classified as very high risk (SCORE ≥10%) or high risk (SCORE ≥5 and <10%). Providing risk information has been shown to change high risk patients perception of risk and physician's prescribing habits [27] yet recent research from Ireland found that only a third of general practitioners (GPs) frequently use a CVD risk assessment tool in practice [28]. Our data is limited in that we collected information on current medication use and therefore cannot ascertain prescription of lipid-lowering medication and the extent to which non adherence to treatment is a feature of our data. Individuals at lower absolute CVD risk are required to weigh up the smaller benefits to them individually of taking medication and the adoption of lifestyle modification in consultation with their general practitioner (GP). In one UK study following the widening of the risk threshold for offering statin therapy one in four patients took up the opportunity to discuss starting statins with their GP and only one in ten chose to start a statin [29]. Poor adherence to statins is reported in primary prevention populations with low perceived risk a strong predictor of poor adherence and many displaying a preference for dietary control [30]. Our findings reveal a lack of awareness of hypercholesterolaemia in the primary prevention population including those at low, moderate and high absolute risk. The largest number of future cases of CVD will arise in the moderate risk group due to its relative size. Our findings suggest that opportunities to promote dietary and lifestyle modification at a population level to reduce future CVD are being missed.

The strengths of this study include the large nationally representative sample which allows us to generalise our findings to the older adult population in Ireland. Additional strengths include the use of standardised protocols for the interviews, health assessments and laboratory methods. The TC and LDL-C levels reported may underestimate the cholesterol profile in this population as those included in the analysis had more favourable risk profiles than respondents who did not take part in the health assessment and thus were excluded from the analysis. This selection bias was addressed at the design stage by the inclusion of home and health centre based assessments and at the analysis stage by the utilisation of calibration weights. Information on cholesterol testing and awareness of hypercholesterolaemia was based on self-report which may be subject to misclassification. In this study we used the SCORE risk assessment tool and associated guideline LDL-C target thresholds. However, a large number of CVD risk assessment tools are in use in primary care [27, 31]. While most recommend prediction models based on simple risk factor data for decision making or further investigation and management, there is no consensus on the strategy for screening, recommended target populations, screening tests or treatment thresholds [31]. We acknowledge the uncertainty this creates in using a single risk assessment tool in our analysis.

Our findings show a decrease in TC and LDL-C in line with global trends which are likely driven by improvements in diet and pharmacological interventions which reduce cholesterol. Despite these improved results at a population level, our analysis of the levels of awareness, treatment and control of hypercholesterolaemia within CVD risk categories demonstrates an ongoing need for targeted interventions to reduce LDL-C especially in those at highest risk of CVD in the older Irish population.