Cardiovascular disease is the leading cause of death both in Ireland and Europe . Our analysis revealed that 64% of the older Irish population were objectively hypertensive. A prevalence of 64% is somewhat high compared to international prevalence rates when one considers our sample is relatively young (59.7% are in the 50–64 age-group). A review published in 2005 investigated the global burden of objective hypertension through analysis of worldwide data and found that in established market economies the prevalence rate of hypertension for the 50–59 and the 60–69 age groups was 44.8% and 60.3% respectively for men and 42% and 58.7% for women . Crucially this study defined objective hypertension as we did (SBP ≥ 140 mm Hg and/or DBP ≥ 90 mm Hg and/or on antihypertensive medication). This underlines the fact that while it was not feasible to define objective hypertension as per the ESC guidelines ‘at least 2 blood pressure measurements per visit and at least 2 to 3 visits’ , our method was validated and has been used in other international studies estimating hypertension prevalence.
Another key finding of our study was the discrepancy between self-reported and objective prevalence of hypertension (37% and 64%, respectively). One could suggest that the self-reported and the objective measures are not contemporary (i.e. in the past one truly did not have hypertension but in the interval has developed it) and this induces discrepancy. However on analysis 90% of the sample had attended their physician or hospital in the past year.
A study based on data from the Health Survey for England found that objective hypertension (similarly defined as ≥140/90 or being on treatment for blood pressure) was observed in 30% of individuals of which 34% were unaware of the diagnosis . The ‘lack of awareness’ figure is comparable to our findings however we find much higher objective prevalence, the corollary of which is that the impact of such ‘unawareness’ is much more sinister in terms of public health. Whether the crux of the problem is that hypertension is not simply being diagnosed/is suboptimally managed or whether poor patient awareness is the main factor, is not clear. Is there a socioeconomic reason for the discrepancy? Given that previous studies have suggested that hypertension control rates in Europe are low , we attempted to gain clearer understanding by disentangling the proportions of those on antihypertensive medication in the context of socioeconomic gradient.
We discovered that of our entire sample 36% were not hypertensive, 26% were appropriately managed, 19% were not treated to target and finally 19% were untreated. As illustrated in Table 3 and discussed in the results section, we found that individuals who had tertiary/higher education and in the highest tertile of the wealth distribution had a higher probability of being ‘healthy’. Interestingly optimal and suboptimal management of hypertension (the latter which could also equate to non compliance with medication) did not demonstrate a SEG although those in the middle and highest tertiles of the wealth distribution did have a lesser probability of being hypertensive and untreated. In analyses not detailed here, we tested whether the discrepancy in our results between objective hypertension and self-report of hypertension was influenced by a SEG in accuracy of self-report. As per Johnston et al (who also demonstrated large discrepancy) those at the top of the education or wealth ladder were less likely to report false negatively (p < 0.05).
Focusing then on hypercholesterolemia, there was once again a demonstrated discrepancy between the self-report and objective measure, prevalence being 41.1% and 72.1% respectively. In terms of SEG, the wealthier were more likely to self-report elevated cholesterol however this association was not sustained objectively. Commercially there is widespread advertisement of low cholesterol spreads and it is likely this raises public awareness of hypercholesteraemia over hypertension. Interpreting self-report as patient awareness, this suggests perhaps that these individuals have been influenced by the media and are more likely to seek and pay for screening than lower socioeconomic classes.
The more educated did demonstrate a higher probability of having elevated HDL-C, which represents a more ‘positive’ lipid profile. Banks et al. and Muennig et al.  used data from 1999/2002 NHANES in similar studies to ours, the former looking at HDL-C and the latter at both LDL-C and HDL-C. Their results also demonstrated no SEG in LDL-C. Despite no SEG, we found that a high proportion of our population have elevated LDL-C, also a primary target of dyslipidemia therapy to reduce cardiovascular risk.
It is difficult to explain the SEG in HDL-C given that we adjusted for use of cholesterol-lowering medication in separate analyses and this made no difference. One could argue that this was because participants were not fasting. However we have previously demonstrated that there was no statistically significant difference between fasting and non-fasting samples in population studies.
Limitations of this study exist in relation to our cardiovascular measures. Our methodology for diagnosis of objective hypertension, while used in other extant large scale population studies including the Health and Retirement Study (HRS) and the English Longitudinal Study on Ageing (ELSA), is not in adherence with either the American Heart Association, nor the European Cardiology Society guidelines for diagnosis of hypertension, according to which diagnosis should be based on blood pressure measurements taken repeatedly over a period of time. It must be therefore acknowledged that it is possible our prevalence levels may be an overestimation, and include those with ‘white-coat hypertension’ and also those who are on antihypertensive medications for reasons other than hypertension. Nevertheless we do not consider that this nullifies the large discrepancy we have illustrated in prevalence rates between self-report and objective hypertension. In terms of objective lipid profile, the samples were not fasting, which is the gold standard, however as discussed above we do not consider that this altered the measure significantly.
Strengths of the study are the large nationally representative sample size and the availability both self-reported and objective cardiovascular measures. This enabled more insightful investigation into the epidemiology of these prevalent health conditions.