We found that lower cholesterol levels in middle age predicted lower total mortality and better physical quality of life in old age after 39 years of follow-up. No difference was seen in the mental component of quality of life. These findings support current guidelines for cholesterol goals and show that a low risk-factor level in midlife can affect long-term mortality and morbidity, postpone physical disability, and improve quality of life in old age.
Much of the illness and disability in elderly patients is related to risk factors in midlife, such as smoking, hypertension, weight gain, diabetes, and dyslipidemia. It is hoped that reducing these factors will not only postpone death but also reduce disability and increase the quality of life in old age. The compression-of-morbidity hypothesis suggests that it may be possible to reduce cumulative lifetime morbidity.1 If the delay in the onset of disability in persons with a healthier lifestyle is greater than the postponement of death, a greater number of years without disability results, with a respective improvement in the quality of life. If this theory holds true, then better functioning in old age should be seen in those seniors who had tighter control of cardiovascular risk factors when they were young.
Low or lowered cholesterol level has been shown in numerous studies to reduce cardiovascular morbidity and mortality.2 There have been concerns regarding the possible adverse effects of long-term low (or lowered) serum cholesterol level, however, especially on mental functioning and quality of life.3-5 We evaluated the effect of baseline cholesterol levels of initially healthy men on mortality rates and quality of life in old age over a follow-up period of 39 years.
Patients and methods
In a long-term outcome study, a cohort of 3,277 healthy Finnish businessmen, aged 30 to 45 years at baseline, was prospectively followed from the 1960s to 2002 for total cholesterol levels and mortality. All participants were white men from the highest social class (businessmen or executives). They were healthy at baseline and took no medication for chronic diseases. Their cholesterol levels were measured at the start of the study, in 1974, and in 1985.
A questionnaire was sent to all survivors of the original sample in 2000, and a response was received from 1,820 of 2,251 participants (81%). Subjects were asked about their lifestyle habits, including their most recent serum cholesterol level, physical activity, alcohol consumption, and smoking. The quality of life was measured with the RAND-36 health survey,6 which is equivalent to the Short Form (SF)-36 health survey. RAND-36 as a mailed questionnaire has been validated in the Finnish population. It is divided into two summary scores reflecting the physical and mental components of quality of life. We also accounted for deaths using the method introduced by Diehr and Patrick7 because, in a long-term study, only survivors are studied and thus important information may be left out.
Data regarding total mortality during the follow-up period were ob­tained from the reliable Finnish Central Pop­ulation Register up to December 31, 2002. Survival was analyzed by comparing baseline cholesterol values divided into six groups: 5.0 mmol/L or lower, 5.1—6.0 mmol/L, 6.1–7.0 mmol/L, 7.1–8.0 mmol/L, 8.1–9.0 mmol/L, and above 9.0 mmol/L. Quality of life was compared between the baseline low-cholesterol group and the other cholesterol groups combined.
The average age of the sample at the beginning of the study was 38 (standard deviation [SD], 4) years. At baseline, all men were healthy. None of the participants had diabetes or cardiovascular disease. Nearly half (45%) were smokers. The average serum cholesterol concentration was 6.6 (SD, 1.2) mmol/L, and 224 men (6.8%) had a serum cholesterol level of 5.0 mmol/L or lower. The average age of the re­spondents was 73 (SD, 4) years. Those alive in 2000 reported their most recent cholesterol value in the 2000 survey, showing that baseline differences of cholesterol values had basically stayed the same over the years, with a general regression to the mean, except in the lowest cholesterol group. A total of 16% of responders reported that they were taking medication to lower cholesterol.
After a follow-up period of 39 years, 1,173 deaths from all causes (35.8%) occurred. Total 39-year mortality increased in a graded manner based on the baseline serum cholesterol level (Figure 1). Multivariate analyses showed that the relative hazard of death rose by 11% (95% confidence interval [CI], 6%—17%) for every increase of 1 mmol/L of cholesterol (adjusted for age, body mass index [BMI], and the year of first cholesterol measurement). The relative hazard was insensitive to more baseline risk factors in the model. With similar adjustments, total mortality was 25% lower (95% CI, 2%–43%) in the low-cholesterol (5 mmol/L or less) group compared with the higher-cholesterol groups combined.
In 2000, the reported prevalence of coronary artery disease and cerebro­vascular disease was significantly lower (P = .002 and P = .04, respectively) in the lowest baseline cholesterol group than in the combined higher-cholesterol groups. The proportions of re­ported diabetes, cancer, and mental illness were not statistically different be­tween the low- and higher-cholesterol groups.
Quality of life
Quality of life was assessed with the RAND-36 questionnaire6 in the 2000 survey. The physical component score (adjusted for blood pressure, BMI, baseline smoking, and age) showed a markedly better quality-of-life score in the group with the lowest baseline cholesterol level compared with the combined higher-cholesterol group (47.6 versus 45.4; P = .02; Figure 2). There was no difference in the mental component score of quality of life between the low-cholesterol and combined higher-cholesterol groups (52.3 versus 52.9).
If deaths during the follow-up period were accounted for, the difference in the physical component of quality of life became more prominent and highly significant (P = .005), and the mental component tended to be better in the lowest cholesterol group (Figure 2).
This prospective all-male cohort study of nearly 40 years’ duration provides evidence that cholesterol levels in young and middle-aged men have a long-term effect not only on cardiovascular health but also on total mortality and the quality of life in old age.
Other published reports, such as the results of the Multiple Risk Factor In­tervention Trial (MRFIT),8 have shown that cholesterol levels are an indicator of total and cardiovascular mortality. The benefits of very low cholesterol, however, have been in dispute. The Honolulu Heart Program showed that low blood cholesterol levels predicted a greater total mortality risk in the elderly.9 The explanation given for this J-shaped curve of cholesterol levels and mortality has been that fatal diseases such as cancer, for example, may lower cholesterol in the terminal stage, giving the J-shaped association if only the final years are examined. In all, the cholesterol levels in old age were lower than in middle age. This trend has also been observed in the Finnish population since the 1970s, and it has been attributed to dietary changes in the population.10
The use of cholesterol-lowering me­d­ication became more common only after the introduction of HMG-CoA reductase inhibitors (statins) in the 1980s and after subsequent large randomized trials, such as the Scan­din­avian Simvastatin Survival Study (4S), showed the benefit of statins. The men in our study were healthy at baseline, and cholesterol-lowering drugs were still rarely used at the time of the 1985 survey, when only 3% of respondents reported taking these drugs. A total of 16% of men said that they were taking cholesterol-lowering medication, mainly statins, in the 2000 survey, and this had an effect on the trend of lower cholesterol values in the 2000 questionnaire. Still, the benefits of low cholesterol in our cohort mostly reflect the effect of naturally modified low cholesterol levels, and thus the results cannot be extrapolated to the effect achieved by the use of statins, for example. We know from the extension study of the 4S,11 however, that medication-lowered cholesterol is also as­sociated with long-term benefits.
Quality of life is a sensitive indicator of disability in the elderly because it is affected by all physical, mental, and social forms of disability, such as impairment of mobility, hearing, vision, sleep, or depression. Regarding the compression-of-morbidity theory, it can be calculated that each 1-point difference of the physical component score would mean a postponement of disability by 1 year.6 The 2-point difference seen in our study population between the cholesterol groups therefore means that low cholesterol levels in midlife result in a postponement of disability by 2 years in old age. In addition to the personal gain, delaying geriatric care by 2 years would significantly reduce the financial burden on the health care system.
It has been debated whether low or lowered cholesterol could have a negative effect on mental or cognitive function. We found no such trend in the mental component of quality of life in our study. On the contrary, the mental component was not affected and even tended to be better in the low-cholesterol group if deaths were accounted for. In a long-term study, only survivors are studied, and thus important information may be left out. In addition, the association between better physical quality of life and low cholesterol levels was strengthened further when deaths were accounted for in the calculations.
The sample in our study was obviously selective: All subjects were white men with high social status. This is a strength of the study because it re­moves the confounding effects of social status. However, caution should be used when generalizing the findings to the general population, especially women.
We prospectively followed a co­hort of 3,277 middle-aged men for 39 years to determine whether midlife cholesterol levels affect total mortality and quality of life in old age. We found that men with naturally low cholesterol levels had better long-term survival and better quality of life in old age. The results support the current cholesterol guidelines and highlight the importance of actively lowering cholesterol levels in the mid­dle-age population.