Advertisement

Effect of Cholesterol on Mortality and Quality of Life up to a 46-Year Follow-Up

      The effect of cholesterol level on the health of older people is a matter of debate, probably because of the bidirectional association. We investigated this paradox in a long-term study. The baseline assessments of the Helsinki Businessmen Study (a cohort of mainly business executives, born 1919 to 1934) included the total cholesterol value and other cardiovascular risk factors from 1964 to 1973. These men were followed up for ≤46 years (through January 2010). During the follow-up period, the cholesterol value was assessed by self-report in 2000 (n = 1,292). Mortality was ascertained from the national registers, symptoms, and health-related quality of life with RAND-36 from questionnaires in 2000. A total of 3,277 healthy men without chronic diseases at baseline were included in the analyses. The median total cholesterol concentration at baseline was 6.5 mmol/L (251 mg/dl) (interquartile range 5.8 to 7.3 mmol/L, 224 to 282 mg/dl) and, in 2000, was 5.2 mmol/L (201 mg/dl) (interquartile range 4.6 to 5.9 mmol/L, 178 to 228 mg/dl). During the follow-up period, 1,773 men (54%) died. A strong and graded relation was found between the cholesterol level and total mortality, with the men with a cholesterol level ≤4 mmol/L (154 mg/dl) having the lowest mortality. In all, the men with the lowest cholesterol gained the most life years. However, no association was found with the cholesterol level in 2000 (when 16% were using statins) and subsequent mortality. The lowest (≤4 mmol/L) cholesterol value in midlife also predicted a higher score in the physical functioning scale of RAND-36 in old age. In conclusion, a low total cholesterol value in midlife predicts both better survival and better physical functioning in old age.
      The effect of a low serum cholesterol value on total mortality, especially at an older age, has been a matter of debate and controversy.
      • Räihä I.
      • Marniemi J.
      • Puukka P.
      • Toikka T.
      • Ehnholm C.
      • Sourander L.
      Effect of serum lipids, lipoproteins, and apolipoproteins on vascular and nonvascular mortality in the elderly.
      • Schatz I.J.
      • Masaki K.
      • Yano K.
      • Chen R.
      • Rodriguez B.L.
      • Curb J.D.
      Cholesterol and all-cause mortality in elderly people from the Honolulu Heart Program: a cohort study.
      An elevated serum cholesterol level is a well-known risk factor for coronary heart disease in midlife and early old age; however, in late life, this association seems to disappear or even reverse.
      • Räihä I.
      • Marniemi J.
      • Puukka P.
      • Toikka T.
      • Ehnholm C.
      • Sourander L.
      Effect of serum lipids, lipoproteins, and apolipoproteins on vascular and nonvascular mortality in the elderly.
      Furthermore, low or lowered total cholesterol might adversely affect mental functioning and health-related quality of life (HRQOL).
      • Äijänseppä S.
      • Kivinen P.
      • Helkala E.L.
      • Kivelä S.L.
      • Tuomilehto J.
      • Nissinen A.
      Serum cholesterol and depressive symptoms in elderly Finnish men.
      • Muldoon M.F.
      • Barger S.D.
      • Ryan C.M.
      • Flory J.D.
      • Lehoczky J.P.
      • Matthews K.A.
      • Manuck S.B.
      Effects of lovastatin on cognitive function and psychological well-being.
      Because cholesterol was measured in early midlife in the Helsinki Businessmen Study during the 1960s,
      • Miettinen T.A.
      • Huttunen J.K.
      • Naukkarinen V.
      • Strandberg T.
      • Mattila S.
      • Kumlin T.
      • Sarna S.
      Multifactorial primary prevention of cardiovascular diseases in middle-aged men Risk factor changes, incidence, and mortality.
      we had an opportunity to test the associations among cholesterol, life years gained, and HRQOL in old age for ≤46 years of follow-up. We earlier reported the mortality and HRQOL data (RAND-36/Medical Outcomes Study Short-Form 36 physical and mental component summary scores) of this cohort up to 2002, showing a favorable effect of a low cholesterol level using 5 mmol/L (193 mg/dl) as a cutoff point.
      • Strandberg T.E.
      • Strandberg A.
      • Rantanen K.
      • Salomaa V.V.
      • Pitkälä K.
      • Miettinen T.A.
      Low cholesterol, mortality, and quality of life in old age during a 39-year follow-up.
      We have extended this follow-up period to 2010 to report the life years gained and have conducted a more detailed HRQOL analysis, in particular, examining the long-term effect of the lowest cholesterol value (≤4 mmol/L or 154 mg/dl). We also report on the effect of the cholesterol level measured in old age on subsequent mortality, as well as the effect of cholesterol-lowering medication started in old age, among this cohort.

      Methods

      In a long-term outcome study, a cohort of 3,490 healthy Finnish businessmen, born from 1919 to 1934 (the Helsinki Businessmen Study), aged 30 to 45 years at baseline, has been prospectively followed up since the 1960s to the present for cardiovascular risk factors, mortality, and HRQOL.
      • Miettinen T.A.
      • Huttunen J.K.
      • Naukkarinen V.
      • Strandberg T.
      • Mattila S.
      • Kumlin T.
      • Sarna S.
      Multifactorial primary prevention of cardiovascular diseases in middle-aged men Risk factor changes, incidence, and mortality.
      • Strandberg T.E.
      • Strandberg A.
      • Rantanen K.
      • Salomaa V.V.
      • Pitkälä K.
      • Miettinen T.A.
      Low cholesterol, mortality, and quality of life in old age during a 39-year follow-up.
      Initially, the study cohort was a pragmatic health examination cohort conducted at the Institute of Occupational Health in Helsinki. From 1964 to 1973, these men participated in health checkup procedures that included clinical examinations and laboratory tests. Only 1 baseline measurement of serum total cholesterol was available. The Ethics Committee of the Helsinki University Hospital, Department of Medicine approved the follow-up procedures (the most recent application by the code HUS 429/13/03/01/09).
      All participants were white men from the highest social class and were either businessmen or management executives with similar socioeconomic and employment status. The study population and examinations have been previously described in detail.
      • Miettinen T.A.
      • Huttunen J.K.
      • Naukkarinen V.
      • Strandberg T.
      • Mattila S.
      • Kumlin T.
      • Sarna S.
      Multifactorial primary prevention of cardiovascular diseases in middle-aged men Risk factor changes, incidence, and mortality.
      • Mattila S.
      Correlation Between ECG Abnormalities Observed in Health Examinations and Coronary Heart Disease in Middle-Aged Men.
      • Salomaa V.
      Long-Term Effect of Primary Prevention Measures of Coronary Heart Disease on the Risk Factor Levels in Middle-Aged Men.
      • Strandberg T.E.
      • Salomaa V.V.
      • Vanhanen H.T.
      • Naukkarinen V.A.
      • Sarna S.J.
      • Miettinen T.A.
      Mortality in participants and non-participants of a multifactorial prevention study of cardiovascular diseases: a 28-year follow-up of the Helsinki Businessmen Study.
      The present study on mortality and cholesterol included the 3,277 men (94% of the total cohort) for whom the baseline serum total cholesterol level was available. The exclusion of the 612 men who participated in the 5-year intervention study during the 1970s
      • Miettinen T.A.
      • Huttunen J.K.
      • Naukkarinen V.
      • Strandberg T.
      • Mattila S.
      • Kumlin T.
      • Sarna S.
      Multifactorial primary prevention of cardiovascular diseases in middle-aged men Risk factor changes, incidence, and mortality.
      did not alter the conclusions of the present study; thus, the results of the whole original cohort are presented.
      In 2000, we sent a questionnaire to all survivors (n = 2,251), with 1 remailing for nonrespondents, and 1,820 (81%) responded. This questionnaire included questions about current diseases, medications, and the current cholesterol value by self-report. The Finnish version of the RAND-36 Item Health Survey, version 1.0 (practically identical to the Medical Outcomes Study Short-Form 36-item health survey
      • Aalto A.-M.
      • Aro A.R.
      • Teperi J.
      RAND-36 as a measure of health-related quality of life: reliability, construct validity, and reference values in the Finnish general population.
      • Hays R.D.
      • Morales L.S.
      The RAND-36 measure of health-related quality of life.
      ) was embedded in that questionnaire. The 8 scales of the RAND-36 (physical functioning, role limitations caused by physical health problems, bodily pain, general health, vitality, social functioning, role limitations caused by emotional problems, and mental health) evaluate different aspects of HRQOL. The RAND-36 as a postal questionnaire has been validated in the Finnish general population.
      • Aalto A.-M.
      • Aro A.R.
      • Teperi J.
      RAND-36 as a measure of health-related quality of life: reliability, construct validity, and reference values in the Finnish general population.
      The physical functioning score of this instrument has been shown to be a valid measure of mobility disability in epidemiologic studies of older people.
      • Syddall H.E.
      • Martin H.J.
      • Harwood R.H.
      • Cooper C.
      • Aihie Sayer A.
      The SF-36: a simple, effective measure of mobility-disability for epidemiological studies.
      The participating men came to the laboratory between 9 and 11 a.m. after a 12-hour fast. During the 1960s and up to 1974, the serum cholesterol concentration was determined using the method of Huang et al
      • Huang T.C.
      • Chen C.P.
      • Wefler V.
      • Raftery A.
      A stable reagent for the Liebermann-Burchard reaction: application to rapid serum cholesterol determination.
      ; subsequently, routine laboratory analyses with enzymatic methods have been used. According to our measurements,
      • Salomaa V.
      Long-Term Effect of Primary Prevention Measures of Coronary Heart Disease on the Risk Factor Levels in Middle-Aged Men.
      the cholesterol levels measured with older methods resulted in 8.3% greater values. Accordingly, a value of 5.0 mmol/L (193 mg/dl) in the 1960s would correspond to 4.6 mmol/L (178 mg/dl) measured with an enzymatic method. We used the corrected values in the present analyses; the conclusions nevertheless remained the same when the original values were used.
      In addition to the baseline cholesterol level, we have the self-reported cholesterol values in 2000 for 1,292 men (71% of the respondents in 2000), and in 2002 to 2003, serum samples were obtained from a random sample of 665 men, with these measured in Professor Miettinen's Laboratory, Helsinki University. Both the self-reported values in 2000 and the laboratory measured values in 2002 to 2003 displayed a Gaussian distribution and were significantly correlated (r = 0.52, p <0.001) despite the more prevalent use of statins in 2002 to 2003 than in 2000.
      The total mortality of the study cohort up to January 2010 was retrieved from the Central Population Register, which keeps a record of all Finnish citizens. The cause of death (available for 95% of all deaths) through December 2007 was determined from the countrywide computerized Cause-of-Death Register of Statistics Finland in which trained nosologists (unaware of the present study) coded the cause of death. The causes were categorized into 5 groups: cardiovascular, cancer, accidents, suicides, and other.
      The baseline serum total cholesterol levels were divided into 7 groups at 1 mmol/L intervals as follows: ≤4.0 mmol/L (154 mg/dl) (n = 22), 4.1 to 5.0 mmol/L (155 to 193 mg/dl) (n = 202), 5.1 to 6.0 mmol/L (194 to 232 mg/dl) (n = 803), 6.1 to 7.0 mmol/L (233 to 270 mg/dl) (n = 1,170), 7.1 to 8.0 mmol/L (271 to 309 mg/dl) (n = 720), 8.1 to 9.0 mmol/L (310 to 347 mg/dl) (n = 255), and >9.0 mmol/L (347 mg/dl) (n = 105). The cholesterol level from 2000 was divided into 5 groups (few men had the highest cholesterol levels), with the lowest and highest groups being ≤4 mmol/L (154 mg/dl) and >7 mmol/L (270 mg/dl), respectively. The cholesterol-specific cumulative mortality was calculated using the Kaplan-Meier method, and the log-rank test was used to estimate the significance of differences between the groups. We examined mortality both from the year of the first visit (median 1968, range 1964 to 1973) and from 2000. The follow-up time extended until death or the end of the follow-up period (January 2010). Hazard ratios with their 95% confidence intervals for mortality were calculated using Cox's proportional hazards regression. Cholesterol groups were compared for survival time, and RAND-36 scales using analysis of covariance. Statistical analyses were performed using NCSS (2004) (NCSS, Kaysville, Utah). Significance was defined as 2-sided p <0.05.
      For the present study, we investigated (1) the effect of the baseline cholesterol level on mortality and life years gained through January 2010; (2) the effect of the cholesterol level and cholesterol-lowering drug use in 2000 on subsequent mortality from 2000 to 2009; and (3) the effect of midlife cholesterol level on detailed HRQOL in old age in 2000, with a high response rate. The median follow-up period was 39 years (interquartile range 30 to 42, maximum 46), denoting 113,425 person-years.

      Results

      The baseline characteristics of the study group in midlife and in 2000 (time of the HRQOL assessment and the start of old age mortality follow-up) are listed in Table 1. During the follow-up period, a duration of ≤46 years, 1,773 men (54%) died. A graded, highly significant association was seen between both total mortality and cholesterol; thus, the lowest mortality was observed in men with baseline cholesterol ≤4.0 mmol/L (154 mg/dl) (Figure 1). The age-adjusted mortality increased consistently with the baseline cholesterol, with the difference in the mean survival 5.5 years between the lowest and highest baseline cholesterol group (Figure 2).
      Table 1Comparison of study cohort in midlife and old age
      VariableMidlife (1964–1973) (n = 3,277)Old Age in 2000 (n = 1,820)
      Age (years)
       Median4274
       Interquartile range38–4570–76
      Body mass index (kg/m2)
       Median25.725.5
       Interquartile range24.0–27.523.6–27.5
      Based on self-report.
      Blood pressure (mm Hg)
       Systolic135 (125–140)
       Diastolic85 (80–90)
      Total cholesterol (mmol/L) [mg/dl]
       Median6.5 [251]5.2 [201]
       Interquartile range5.8–7.3 [224–282]4.6–5.9
      Based on self-report.
       [178–228]
      Total cholesterol grouping
      Data are presented as the mean (number of patients).
       ≤4.0 [154]3.7 [143] (22)3.6 [139] (127)
       4.1–5.0 [155–193]4.7 [181] (202)4.6 [178] (442)
       5.1–6.0 [194–232]5.6 [216] (803)5.5 [212] (478)
       6.1–7.0 [233–270]6.5 [251] (1,170)6.5 [251] (204)
       7.1–8.0 [271–309]7.5 [290] (720)7.6 [293] (41)
       8.1–9.0 [310–347]8.4 [324] (255)
       >9.0 [347]9.8 [378] (105)
      One-hour glucose (mmol/L) [mg/dl]
      Measured in 2,750 men.
       Median6.0 [108]
       Interquartile range4.9–7.4 [88–133]
      Triglycerides (mmol/L) [mg/dl]
      Measured in 1,415 men.
       Median1.3 [115]
       Interquartile range1.0–1.8 [89–159]
      Smoker41%
      Based on self-report.
      9%
      Based on self-report.
      low asterisk Based on self-report.
      Data are presented as the mean (number of patients).
      Measured in 2,750 men.
      § Measured in 1,415 men.
      Figure thumbnail gr1
      Figure 1Survival curves of all 7 baseline cholesterol groups (n = 3,277): I = <4.0 mmol/L [154 mg/dl] (n = 22) (light green), II = 4.1 to 5.0 mmol/L [155 to 193 mg/dl] (n = 202) (dark green), III = 5.1 to 6.0 mmol/L [194 to 232 mg/dl] (n = 803) (yellow), IV = 6.1 to 7.0 mmol/L [233 to 270 mg/dl] (n = 1,170) (light blue), V = 7.1 to 8.0 mmol/L [271 to 309 mg/dl] (n = 720) (dark blue), VI = 8.1 to 9.0 mmol/L [310 to 347 mg/dl] (n = 255) (pink), VII = >9.0 mmol/L [347 mg/dl] (n = 105) (red).
      Figure thumbnail gr2
      Figure 2Age-adjusted mean life-years gained according to baseline cholesterol group (i.e., subdivided into 7 subgroups: ≤4 mmol/L [154 mg/dl] [dark green], 4.1 to 5.0 mmol/L [155 to 193 mg/dl] [light green], 5.1 to 6.0 mmol/L [194 to 232 mg/dl] [yellow], 6.1 to 7.0 mmol/L [233 to 270 mg/dl] [light blue], 7.1 to 8.0 mmol/L [271 to 309 mg/dl] [dark blue], 8.1 to 9.0 mmol/L [310 to 347 mg/dl] [pink], >9 mmol/L [347 mg/dl] [red]). p Value: age-adjusted analysis of covariance.
      The association between cholesterol measured in old age and subsequent mortality was different. From 2000 to 2010, 575 men died (32%), and, of those with self-reported cholesterol levels, 384 men (30%) died. No relation was found between the cholesterol level and mortality (p = 0.74).
      In 2000, 277 men (16% of respondents) reported using cholesterol-lowering medication (mostly statins); 84 of them (30%) died during follow-up through January 2010. The age-adjusted mortality among the drug users was not significantly different from that of the nonusers (hazard ratio 0.87, 95% confidence interval 0.69 to 1.09; p = 0.23). Because long-term hypercholesterolemia and co-morbidity were more prevalent among the statin users, we further adjusted the analyses for the cholesterol level during the 1960s and self-rated health in 2000. After these adjustments, statin use was associated with a significant 26.5% lower mortality (hazard ratio 0.74, 95% confidence interval 0.57 to 0.94; p = 0.02).
      In the 2000 questionnaire survey, the response rate was 81% overall, 94% in the lowest cholesterol group versus 79% to 83% in the other groups. The detailed and age-adjusted RAND-36 scores in old age according to cholesterol groups in midlife are shown in Figure 3. The mean scores were in a graded fashion greater among those with the lowest serum cholesterol in midlife; however, the global difference among the cholesterol groups was significant only for the physical functioning score (p = 0.03). It is important to note that a low cholesterol level was not associated with poorer scores for the scales of RAND-36 assessing mental function.
      Figure thumbnail gr3
      Figure 3Baseline cholesterol and health-related quality of life (RAND-36, adjusted for age, baseline BMI, and smoking) in old age in 2000. Baseline cholesterol groups: 1 = ≤4 mmol/L [154 mg/dl] (dark green), 2 = 4.1 to 5.0 mmol/L [155 to 193 mg/dl] (light green), 3 = 5.1 to 6.0 mmol/L [194 to 232 mg/dl] (yellow), 4 = 6.1 to 7.0 mmol/L [233 to 270 mg/dl] (light blue), 5 = 7.1 to 8.0 mmol/L [271 to 309 mg/dl] (dark blue), 6 = 8.1 to 9.0 mmol/L [310 to 347 mg/dl] (pink), 7 = >9 mmol/L [347 mg/dl] (red). In all RAND-36 scales, score of 100 is best possible. Abbreviations for RAND-36 scales: PF = physical functioning, RP = role limitations owing to physical health; BP = bodily pain; GH = general health; VT = energy/vitality; SF = social functioning; RE = role limitations due to emotional problems; MH = mental health/emotional well being. Global p values for overall difference between groups depicted over bars.

      Discussion

      Our results show that the lowest cholesterol level, ≤4 mmol/L (154 mg/dl), in midlife is associated with longer survival and better physical functioning according to RAND-36 in old age. In our previous report,
      • Strandberg T.E.
      • Strandberg A.
      • Rantanen K.
      • Salomaa V.V.
      • Pitkälä K.
      • Miettinen T.A.
      Low cholesterol, mortality, and quality of life in old age during a 39-year follow-up.
      when the lowest cholesterol group ≤5 mmol/L (193 mg/dl) was divided in half (<4.7 mmol/L [181 mg/dl], n = 114 and 4.7 to 5.0 mmol/L [181 to 193 mg/dl], n = 110), no additional reduction in mortality risk was found (mortality 27.2% vs 24.5%, respectively; p = 0.76). In contrast, this extended follow-up period found no evidence for a J-shape; in fact, a steady, graded relation was found between the life-years gained and the baseline cholesterol value. As far as we are aware, this is 1 of the longest follow-up studies of the effect of the cholesterol value in midlife on subsequent mortality.
      The loss of function mutation R46L in the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene has been associated with a reduction of 12% to 15% in the mean low-density lipoprotein cholesterol level, but the reduction in the risk of coronary heart disease has been 28% to 47% in different studies.
      • Cohen J.C.
      • Boerwinkle E.
      • Mosley Jr, T.H.
      • Hobbs H.H.
      Sequence variation in PCSK9, low LDL, and protection against coronary heart disease.
      • Benn M.
      • Nordestgaard B.G.
      • Grande P.
      • Schnohr P.
      • Tybjaerg-Hansen A.
      PCSK9 R46L, low-density lipoprotein cholesterol levels, and risk of ischemic heart disease: 3 independent studies and meta-analyses.
      This larger than expected reduction in the occurrence of coronary heart disease is believed to be 1 consequence of a lifelong reduction of low-density lipoprotein levels, conferring greater benefit than would a similar reduction in later life. This proposal is in line with our findings.
      Not only was survival better, the quality of those life-years gained was also better. In our previous report,
      • Strandberg T.E.
      • Strandberg A.
      • Rantanen K.
      • Salomaa V.V.
      • Pitkälä K.
      • Miettinen T.A.
      Low cholesterol, mortality, and quality of life in old age during a 39-year follow-up.
      we reported the physical and mental component summary scores of RAND-36. The physical component summary was significantly better in the lowest cholesterol group of ≤5 mmol/L (193 mg/dl) compared to all other cholesterol groups (47.6 vs 45.4, p = 0.02). In contrast, the mental component summary was similar (52.3 vs 52.9, p = 0.51). In the present study, we report the detailed and adjusted scores in 8 different domains and have compared scores through the range of the lowest ≤4 mmol/L (154 mg/dl) to the highest >9 mmol/L (347 mg/dl) value. It was striking that the mean scores were systematically greatest in the lowest cholesterol group; however, statistical significance was achieved only for physical functioning. It was reassuring that mental functioning was not poorer among those with lower cholesterol values.
      The association between cholesterol values in old age and subsequent mortality was no longer significant. Most of these men (70.8% of respondents in 2000) reported their cholesterol value in 2000 and after a follow-up of 9 years (January 2010), an inverse relation showed a trend between cholesterol and mortality, although this was not statistically significant. In our previous report,
      • Strandberg T.E.
      • Strandberg A.
      • Rantanen K.
      • Salomaa V.V.
      • Pitkälä K.
      • Miettinen T.A.
      Low cholesterol, mortality, and quality of life in old age during a 39-year follow-up.
      we showed that quite normal cholesterol levels in old age could have been substantially greater in the same men in midlife. Thus, the cholesterol level present in old age does not necessarily represent the lifetime exposure. A community-based study with a follow-up of 917 persons (aged 50 to 93 years at baseline) for 8 years, showed that the total cholesterol, low-density lipoprotein cholesterol and high-density lipoprotein cholesterol levels decreased with advancing age.
      • Ferrara A.
      • Barrett-Connor E.
      • Shan J.
      Total LDL and HDL cholesterol decrease with age in older men and women: the Rancho Bernardo Study 1984–1994.
      In the follow-up of the Helsinki Aging Study (aged ≥75 years), those who had lowered cholesterol synthesis and reduced cholesterol absorption had the greatest mortality risk, implying that their low cholesterol levels was a consequence of other diseases and deteriorating health.
      • Tilvis R.S.
      • Strandberg T.E.
      • Miettinen T.A.
      Letter in response to cholesterol and all-cause mortality in Honolulu.
      • Tilvis R.S.
      • Valvanne J.N.
      • Strandberg T.E.
      • Miettinen T.A.
      Prognostic significance of serum cholesterol, lathosterol, and sitosterol in old age; a 17-year population study.
      Furthermore, it has been shown in several longitudinal studies that the cholesterol levels in those who subsequently developed Alzheimer's disease had decreased many years before the clinical diagnosis of dementia.
      • Stewart R.
      • White L.R.
      • Xue Q.-L.
      • Launer J.L.
      Twenty-six-year change in total cholesterol levels and incident dementia: the Honolulu-Asia Aging Study.
      • Notkola I.L.
      • Sulkava R.
      • Pekkanen J.
      • Erkinjuntti T.
      • Ehnholm C.
      • Kivinen P.
      • Tuomilehto J.
      • Nissinen A.
      Serum total cholesterol, apolipoprotein E epsilon4 allele, and Alzheimer's disease.
      • Solomon A.
      • Kåreholt I.
      • Ngandu T.
      • Winblad B.
      • Nissinen A.
      • Tuomilehto J.
      • Soininen H.
      • Kivipelto M.
      Serum cholesterol changes after midlife and late-life cognition: twenty-one-year follow-up study.
      Thus, the association of cholesterol level in old age and mortality is a complex entity, influenced by many confounding factors.
      • Kaplan G.A.
      • Haan M.N.
      • Wallace R.B.
      Understanding changing risk factor associations with increasing age in adults.
      We also evaluated the age-adjusted mortality from 2000 to 2009 among statin users and compared it with those of nonusers. When the baseline cholesterol value during the 1960s and at the self-rated health report in 2000 were taken into account, statin use was significantly associated with a 26.5% lower mortality compared to that of nonusers. It has been claimed that selective prescribing leads to an overestimation of the benefits of lipid-lowering medication, because frailty and co-morbidity were associated with a decreased use of these drugs.
      • Glynn R.J.
      • Schneeweiss S.
      • Wang P.S.
      • Levin R.
      • Avorn J.
      Selective prescribing led to overestimation of the benefits of lipid-lowering drugs.
      Our cohort is very homogenous, and among these men, statin use was associated with lower mortality, when co-morbidity was accounted for in the analyses.
      • Charlson M.E.
      • Pompei P.
      • Ales K.L.
      • MacKenzie C.R.
      A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.
      The strengths of our study include the truly long-term follow-up from early midlife to old age—almost approaching a life-course study—and the reliable collection of a relatively large number of end points from national registers. The homogeneity of the cohort (men, high socioeconomic class, no cardiovascular disease, or diabetes at baseline) reduces confounding and helps to clarify the mechanisms. However, our study had several limitations. The number of participants was not particularly large, and all were white men of high socioeconomic status, limiting the generalizability of our study. Only one measured baseline serum cholesterol value was available, and the proportion of men with very low cholesterol levels (≤4 mmol/L, 154 mg/dl) at baseline was quite small (n = 22), because high cholesterol values were very common in Finland in the 1960s. The cholesterol value in 2000 was self-reported by the participants. In contrast, in a large proportion of these men, the cholesterol value was measured in 2002 to 2003, and these 2 sets of values correlated significantly. Finally, the long-term results of the midlife cholesterol value apply to naturally low cholesterol values.

      References

        • Räihä I.
        • Marniemi J.
        • Puukka P.
        • Toikka T.
        • Ehnholm C.
        • Sourander L.
        Effect of serum lipids, lipoproteins, and apolipoproteins on vascular and nonvascular mortality in the elderly.
        Arterioscler Thromb Vasc Biol. 1997; 17: 1224-1232
        • Schatz I.J.
        • Masaki K.
        • Yano K.
        • Chen R.
        • Rodriguez B.L.
        • Curb J.D.
        Cholesterol and all-cause mortality in elderly people from the Honolulu Heart Program: a cohort study.
        Lancet. 2001; 358: 351-355
        • Äijänseppä S.
        • Kivinen P.
        • Helkala E.L.
        • Kivelä S.L.
        • Tuomilehto J.
        • Nissinen A.
        Serum cholesterol and depressive symptoms in elderly Finnish men.
        Int J Geriatr Psychiatry. 2002; 17: 629-634
        • Muldoon M.F.
        • Barger S.D.
        • Ryan C.M.
        • Flory J.D.
        • Lehoczky J.P.
        • Matthews K.A.
        • Manuck S.B.
        Effects of lovastatin on cognitive function and psychological well-being.
        Am J Med. 2000; 108: 538-546
        • Miettinen T.A.
        • Huttunen J.K.
        • Naukkarinen V.
        • Strandberg T.
        • Mattila S.
        • Kumlin T.
        • Sarna S.
        Multifactorial primary prevention of cardiovascular diseases in middle-aged men.
        JAMA. 1985; 254: 2097-2102
        • Strandberg T.E.
        • Strandberg A.
        • Rantanen K.
        • Salomaa V.V.
        • Pitkälä K.
        • Miettinen T.A.
        Low cholesterol, mortality, and quality of life in old age during a 39-year follow-up.
        J Am Coll Cardiol. 2004; 44: 1002-1008
        • Mattila S.
        Correlation Between ECG Abnormalities Observed in Health Examinations and Coronary Heart Disease in Middle-Aged Men.
        ([Dissertation, in Finnish])in: University of Helsinki, Helsinki1980: 1-179
        • Salomaa V.
        Long-Term Effect of Primary Prevention Measures of Coronary Heart Disease on the Risk Factor Levels in Middle-Aged Men.
        ([Academic Dissertation, in Finnish])in: University of Helsinki, Helsinki1988: 1-138
        • Strandberg T.E.
        • Salomaa V.V.
        • Vanhanen H.T.
        • Naukkarinen V.A.
        • Sarna S.J.
        • Miettinen T.A.
        Mortality in participants and non-participants of a multifactorial prevention study of cardiovascular diseases: a 28-year follow-up of the Helsinki Businessmen Study.
        Br Heart J. 1995; 74: 449-454
        • Aalto A.-M.
        • Aro A.R.
        • Teperi J.
        RAND-36 as a measure of health-related quality of life: reliability, construct validity, and reference values in the Finnish general population.
        in: Research Report 101. Stakes, Helsinki1999: 1-78
        • Hays R.D.
        • Morales L.S.
        The RAND-36 measure of health-related quality of life.
        Ann Med. 2001; 33: 350-357
        • Syddall H.E.
        • Martin H.J.
        • Harwood R.H.
        • Cooper C.
        • Aihie Sayer A.
        The SF-36: a simple, effective measure of mobility-disability for epidemiological studies.
        J Nutr Health Aging. 2009; 13: 57-62
        • Huang T.C.
        • Chen C.P.
        • Wefler V.
        • Raftery A.
        A stable reagent for the Liebermann-Burchard reaction: application to rapid serum cholesterol determination.
        Anal Chem. 1961; 33: 1405-1407
        • Cohen J.C.
        • Boerwinkle E.
        • Mosley Jr, T.H.
        • Hobbs H.H.
        Sequence variation in PCSK9, low LDL, and protection against coronary heart disease.
        N Engl J Med. 2006; 354: 1264-1272
        • Benn M.
        • Nordestgaard B.G.
        • Grande P.
        • Schnohr P.
        • Tybjaerg-Hansen A.
        PCSK9 R46L, low-density lipoprotein cholesterol levels, and risk of ischemic heart disease: 3 independent studies and meta-analyses.
        J Am Coll Cardiol. 2010; 55: 2833-2842
        • Ferrara A.
        • Barrett-Connor E.
        • Shan J.
        Total LDL and HDL cholesterol decrease with age in older men and women: the Rancho Bernardo Study 1984–1994.
        Circulation. 1997; 96: 37-43
        • Tilvis R.S.
        • Strandberg T.E.
        • Miettinen T.A.
        Letter in response to cholesterol and all-cause mortality in Honolulu.
        Lancet. 2001; 358: 1904-1905
        • Tilvis R.S.
        • Valvanne J.N.
        • Strandberg T.E.
        • Miettinen T.A.
        Prognostic significance of serum cholesterol, lathosterol, and sitosterol in old age; a 17-year population study.
        Ann Med. 2011 Jan 24; (Epub)
        • Stewart R.
        • White L.R.
        • Xue Q.-L.
        • Launer J.L.
        Twenty-six-year change in total cholesterol levels and incident dementia: the Honolulu-Asia Aging Study.
        Arch Neurol. 2007; 64: 103-107
        • Notkola I.L.
        • Sulkava R.
        • Pekkanen J.
        • Erkinjuntti T.
        • Ehnholm C.
        • Kivinen P.
        • Tuomilehto J.
        • Nissinen A.
        Serum total cholesterol, apolipoprotein E epsilon4 allele, and Alzheimer's disease.
        Neuroepidemiology. 1998; 17: 14-20
        • Solomon A.
        • Kåreholt I.
        • Ngandu T.
        • Winblad B.
        • Nissinen A.
        • Tuomilehto J.
        • Soininen H.
        • Kivipelto M.
        Serum cholesterol changes after midlife and late-life cognition: twenty-one-year follow-up study.
        Neurology. 2007; 68: 751-756
        • Kaplan G.A.
        • Haan M.N.
        • Wallace R.B.
        Understanding changing risk factor associations with increasing age in adults.
        Annu Rev Public Health. 1999; 20: 89-108
        • Glynn R.J.
        • Schneeweiss S.
        • Wang P.S.
        • Levin R.
        • Avorn J.
        Selective prescribing led to overestimation of the benefits of lipid-lowering drugs.
        J Clin Epidemiol. 2006; 59: 819-828
        • Charlson M.E.
        • Pompei P.
        • Ales K.L.
        • MacKenzie C.R.
        A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.
        J Chronic Dis. 1987; 40: 373-383