A number of studies (Agner and Hansen, 1983 ; Wingard et al., 1984 ; Reed et al., 1986 ; Trialists, 2010 ; Cui et al., 2012 ; Tanaka and Okamura, 2012 ; Bombelli et al., 2013 ; LaRosa et al., 2013 ; Liu et al., 2013 ; Sahebzamani et al., 2013 ; Singh et al., 2013 ; Whelton et al., 2013 ; Nishikura et al., 2014 ; Tamosiunas et al., 2014 ; Beydoun et al., 2015 ; Benn et al., 2016 ; Besseling et al., 2016 ; Ravnskov et al., 2016) in a variety of settings have demonstrated an all-cause mortality relationship to serum cholesterol. Most primary prevention trials were only designed to detect overall differences in the major fatal events. It has been theorized that low serum cholesterol (<160 mg/dl) is related to the excess deaths for noncardiovascular disease mortality, particularly cancers. In the Whitehall study (Smith et al., 1992), lung cancers showed the most consistent relationship to low serum cholesterol. Another study, by International Collaborative Group (1982), presented pooled data on cancer mortality and serum cholesterol and concluded that pre-existing disease was most likely causing a metabolic effect on serum cholesterol, rather than low serum cholesterol directly enhancing cancer risk. In addition, acute leukemic cells show high affinity to low-density lipoproteins, lowering blood cholesterol concentrations (Iqbal et al., 2016 ; Stefanko et al., 2017). Patients with metastatic prostate cancer indicate a rapid catabolism of low-density lipoprotein compared with patients without metastasis and normal controls (Freeman and Solomon, 2004). Cholesterol, however, is an essential component of the cell membrane and is involved with many aspects of cellular functions (Sabine, 1977), including the enhancement of immune functions such as T-lymphocyte cytotoxicity (Heiniger et al., 1978). Moreover, it is possible that the inverse association between serum cholesterol and cancer is due to the existence of factors associated with low blood cholesterol levels, which accelerate or inhibit the development of cancer. Investigators proposed that an inverse association between serum total cholesterol and cancer is explained in part by the inverse association between serum retinol and cancer. Retinols act as intracellular antioxidants and reduce carcinogenesis. The inverse association between serum cholesterol and cancer was much reduced when the retinol level was controlled for, but the reverse was not apparent. Another study, however, indicated that the inverse association between serum cholesterol and cancer mortality remained significant even after controlling for the serum retinol (Knekt et al., 1988). The current analysis extensively examined the possible relationship between the effect of low serum cholesterol (<160 mg/dl) and outcome of mortality from cancer.
Materials and methods
This pooling epidemiological observational cohort study data from Relative Risk for Life Expectancy (RIFLE) study include 19 of 52 different large scales population studies started in Italy for 9-year period and which focused on cardiovascular diseases or other chronic conditions. This study included men and women in the age range of 20–69 years.
The study sampling was a random sampling frame. The study sample comprised 30 179 male and 26 005 female participants between 20 and 69 years of age. Underlying causes of death were coded according to the International Classification of Diseases, 10th revision (ICD-10) code (WHO, 2010). The average follow-up for participants was ∼9 years as the endpoint (Fig. 1).
Originally, total serum cholesterol was measured in blood samples drawn from the antecubital vein after a 12-h fast. Several automated enzymatic methods were used in the varying studies, but all of the laboratories involved were under direct or indirect control standardized from the WHO Lipid Reference Center in Prague (The RIFLE Research Group, 1993). Blood pressure was measured in a sitting position, after a 4-min rest, in the right arm using a calibrated sphygmomanometer. Observers were trained and tested following rules given in the WHO manual (Rose and Blackburn, 1968) and using the cassettes developed by the London School of Hygiene (Rose, 1965) and later by the Laboratory of Physiologic Hygiene, University of Minnesota (Prineas, 1978). Systolic and diastolic-2 (fifth phase) levels were used for analysis, although diastolic-1 was also recorded in most studies. Weight was measured in light underwear, rounded off to the nearest kilogram, and height was measured without shoes, expressed in centimeters, following the rules suggested in the WHO manual (Rose and Blackburn, 1968). Smoking status was assessed using a questionnaire directly derived from the WHO cardiovascular survey method manual (Rose and Blackburn, 1968). Age was measured, using the difference between the year of examination and the year of birth, accepting an average error of ±6 months. Relative BMI was calculated as a percent deviation of actual weight to standard weight based on mean of body weight distributions by height.
Mortality rates were calculated, based upon decile cholesterol levels in equal intervals from less than 160 mg/day (<4.4 mmol) to more than 276 mg/dl (≥7.14 mmol). Cause-specific mortality was determined according to the ICD-10 code (WHO, 2010) (cancer: pp. 140–239, noncancer liver dysfunction: pp. 570–573).
To achieve reliability and validity, the original data (RIFLE studies) were reviewed in an attempt to ensure that the data collected (for those variables utilized in this project) were as complete as possible. In an effort to obtain complete information related to cholesterol levels and mortality and to verify the information, all variables that were available and in some way related to cholesterol level were selected for the purpose of controlling confounding or interaction effects. It is assumed that any significant event that occurred during the original follow-up studies was recorded in a data file. Moreover, it is assumed that laboratory records of test results performed at an Italian laboratory, and/or interpretation of results were precise. Clinical, anthropometric, and sociodemographic information also was assumed to be properly recorded. An attempt was made to assure that all test results and other information is sufficient for obtaining scientific results in data analysis. However, given the differences in utilized data and in record keeping, in few instances, there may be some data missing or misclassified with possible underestimated results in the study.
The baseline characteristics were presented as frequencies and mean±SD. The association between total cholesterol and mortality was analyzed using the Cox proportional hazards model. To achieve reliability and validity an attempt was made to ensure that the utilized data were as complete as possible. In an effort to obtain complete information related to cholesterol levels and mortality to verify the information, all variables (e.g. sociodemographic, anthropometric, clinical, laboratory, age, smoking status, BMI, and systolic and diastolic blood pressures) were analyzed for the purpose of controlling confounding or interaction effects. Data analysis was accomplished by utilizing the statistical package for the social sciences software (SPSS; SPSS Inc., Chicago, Illinois, USA) to elaborate on associations between mortality and cholesterol levels.
This data was available by State University New York School of Medicine at Buffalo for used my cardiovascular course project, by permission of professor M. Trevisan as my teacher. Therefore, any related issue was on behalf SUNY at Buffalo.
A total of 64 966 men and women between the ages of 20 and 69 years were reviewed; after deletion of missing values, 30 179 men and 26 005 women remained for analysis of the primary selected variables. Table 1 describes the mean and SD for BMI, age, systolic and diastolic blood pressures, smoking status, and follow-up status in relation to total cholesterol level decile for the male population compared with female population based upon decile cholesterol levels. Generally, mean and SD of all deciles within each variable were similar.
Table 2 indicates mortality rate and age-adjusted, multivariate-adjusted hazard ratio (HR) for mortality of cancer for different cholesterol levels in male and female population. These results showed that cancer death rate drastically decreased in male population by increasing the cholesterol level from the first decile to the last decile. However, this rate fluctuated among female participants. The HR of death and its 95% confidence interval (CI) were calculated with reference to the risk of individuals with cholesterol levels of 276 mg/dl or more. These estimations were adjusted for age and other potential confounding factors (e.g. smoking status, systolic and diastolic blood pressures, BMI, and follow-up status) using the Cox proportional hazards model. In this table, cancer mortality and noncancer liver dysfunction multivariate-adjusted HR, with a 95% CI indicates a risk association with low serum cholesterol for the first decile (<160 mg/dl) for the male population (cancer: HR=1.52, 95% CI: 1.06–2.18; noncancer liver dysfunction: HR=10.73, 95% CI: 3.74–30.18). In the second decile only noncancer liver dysfunction has a significant HR (HR=3.73, 95% CI: 1.16–11.95). For the remaining deciles cancer and noncancer liver dysfunction has some risk or protective association but the results are not statically significant in the male population. In the female population, noncancer liver dysfunction indicates that the risk association, with a 95% CI, is at a significant level for the first decile (HR=25.8, 95% CI: 3.09–217.70). In addition, other deciles of noncancer liver dysfunction, for all cholesterol levels of cancer and other noncancer, there is some association in risk and protective direction, but no significant results exist in the female population.
The primary objective of this study was to determine whether an association exists between low cholesterol (<160 mg/dl) as a risk factor for mortality rate in cancer patients. The study populations were Italians aged 20–69 years, followed up over a 9-year period. Participants selected were similar in health status for most sociodemographic, clinical, anthropometric, and laboratory variables.
During the 9-year study follow-up period, there were 1906 deaths (1439 male and 467 female). Among total 841 mortalities due to cancer, 88 noncancer liver dysfunction accrued in both sex. The total mortality for cancer is almost 10 times (9.6) greater than that for noncancer liver dysfunction and about half the total deaths were due to cancer. This finding is consistent with previous studies that reported that low cholesterol increased risk for different types of cancer mortality (Malvezzi et al., 2013 ; US Cancer Statistics Working Group, 2015). Other studies have found an association between high total cholesterol level cancer mortality in both sexes (Hiatt and Fireman, 1986 ; Knekt et al., 1988 ; Chen et al., 2013 ; Jung et al., 2013 ; Heron and Anderson, 2016).
Our finding showed mortality rate at the lowest cholesterol level belonging to cancer, and noncancer liver dysfunction was high during the follow-up period for the male population and noncancer liver dysfunction in the female population. The finding of a previous study conducted by Mamtani et al. (2016) showed that increased serum cholesterol was independently associated with decreased risk for cancer (Mamtani et al., 2016). This study in the opposite direction supports our study related to issue that, if cholesterol level rises, the risk for cancer will be diminished. Finding of this study indicated an inverse association with lowering cholesterol level and risk for cancer.
Moreover, in a study by Strohmaier et al. (2013), total serum cholesterol concentrations in men in the fifth quintile were significantly borderline associated with decreasing risk for total cancer. In addition, among women, HRs for the fifth quintile were associated with decreasing risk for total cancer (Strohmaier et al., 2013).
Contrary to our findings and previous studies, Marrer et al. (2013) reported that no evident association was found between the lipoprotein(a) levels and the incidence of cancer. Nevertheless, a higher cancer risk seemed to be observed for the highest lipoprotein(a) levels (Marrer et al., 2013). It is debatable that previous study findings were quite different in terms of differences in type of studies. The disagreement might be explained by the difference in study populations and study designs, follow up times, sample sizes and statistical analyses.
The study findings also suggest that different cholesterol levels for different groups in the male population had different biological meanings for those in the low serum cholesterol group. These findings indicate a decrease in cholesterol levels toward the lowest decile (<160 mg/dl) that has a different biological magnitude compared with the other deciles. This information provides indicative properties of cholesterol in the level of the cell membrane. Furthermore, cholesterol is an essential component of the cell membrane and is involved with many aspects of cellular function (Sabine, 1977), including the enhancement of immune function, such as inducing T-lymphocyte cytotoxicity (Heiniger et al., 1978) due to preventing cancer cell growth. It is possible that the inverse association between serum cholesterol and cancer is due to the existence of risk factors associated with low blood cholesterol levels, which accelerate or inhibit the development of cancer. These factors include serum vitamins A and E (retinol, β-carotene, and α-tocopherol), which can act as intracellular antioxidants and reduce carcinogenesis. The lower serum cholesterol is associated with lower concentrations of these vitamins because cholesterol carries fat-soluble vitamins in the general circulation. Several prospective studies indicated that the low serum concentrations of retinol (Kark et al., 1982), β-carotene (Menkens et al., 1986 ; Stahelin et al., 1991), and α-tocopherol (Wald et al., 1984 ; Stahelin et al., 1991) were associated with a higher risk for cancer. Moreover, cholesterol on plasma membrane lipids may result in a significant decrease in erythrocyte and platelet membrane cholesterol content.
Moreover, these findings indicated that the cholesterol component of the body may, by sex, provide either a protective shield against death, as in the female population, or place the individual at risk for death, such as an increased risk for cancer, as demonstrated in the male population. Women naturally store more fatty tissue compared with men. This physiological difference may be more protective against excess mortality events, with longer life expectancy among women than among men. These differences were distinguishable in statistical analysis, from the descriptive to analytical steps.
Our finding of age adjustment for all causes and specific causes of mortality rates had a decile term of serum cholesterol for each sex. There was a significant inverse association between serum total cholesterol with total causes of cancer, and noncancer liver dysfunction death for men. No such association was observed for women, but there was a higher rate of noncancer liver dysfunction mortality in the lowest cholesterol level (<160 mg/dl). Our findings are in agreement with the study conducted by Martin et al. (2015). According to a previous study total cholesterol levels were not statistically significantly associated with breast cancer risk in women (Martin et al., 2015).
Preclinical disease effects, perhaps, were mainly related to noncancer liver dysfunction because the liver is the major site of cholesterol synthesis; and second, because hypercholesterolemia can occur long before death in the liver dysfunction patient.
The findings of this analysis were able to associate cholesterol magnitude with death status in the study population. Analytical results of this study have placed strong emphasis on medical and pharmacological management for the low-cholesterol population, specifically cancer and noncancer liver dysfunction in the study population. The novel aspect of this study was the evaluation risk for cancer based upon change in serum total cholesterol levels from lowest decile to highest one precisely. Thus, our data may also provide important information to patients and clinicians regarding the risk of lowering cholesterol less than 160 mg/dl.
Strength and limitations
It was believed to be fairly accurate as the original data of 19 different large-scale dynamic risk factors and life expectancy observational cohort studies were conducted or coordinated by the same center and because the majority of measurements were made using standardization and quality control. An attempt to identify a low-cholesterol risk factor for excess mortality in this analysis may not be generalizable to other populations other than Italians, due to unknown or unmeasured factors.
Although this study may have reported an underestimation of risk factors in the study population, reporting biases were thought to be minimal. Despite study biases, these findings demonstrated a strong association between these risk factors as a factor of significant morbidity and mortality in low cholesterol level and excess death rate. The results of this research were able to associate cholesterol magnitude with death status in the study population. Analytical results of this study have placed strong emphasis on medical and pharmacological management for the low-cholesterol population, specifically noncancer liver dysfunction in the study population.
However, as in any observational cohort study, the findings may be confounded by unmeasured variables. This may have some limitation impact on study. However, this study like others has some limitations. First, information on post-treatment recurrence was insufficient. Second, the change in serum total cholesterol level at some point. A prospective study is required to determine the prognostic and treatment value of cholesterol. The inter-relation of different cardiovascular symptoms, deaths, and different cholesterol levels are complex and likely to diverse directions toward protective or risk factors. Furthermore, clarification of the biological and clinical pathway is needed to examine the sex issues of particular interests for further research.
Our findings showed an inverse association between serum cholesterol level and cancer-related death and noncancer liver dysfunction mortality. These associations remained statistically significant after controlling for age, cigarette smoking status, BMI, and systolic and diastolic blood pressures. The signification of this study highlighted important clinical implications and would be supportive of public health initiatives directed to prevention of cancer death, due to the major disability and morbidity and cancer diagnosis. The inter-relation of different cardiovascular symptoms, deaths, and different cholesterol levels are complex and likely to diverse directions toward protective or risk factors. Furthermore, clarification of the biological and clinical pathway is needed to examine the sex issues of particular interests for further research.
The authors thank Relative Risk for Life Expectancy (RIFLE) observational cohort studies pooled data in Italy, which the author Nader Parsa used as his cardiovascular course project at New York State University School of Medicine at Buffalo, by permission of Professor M. Trevisan as his teacher.
Conflicts of interest
There are no conflicts of interest.
Agner E, Hansen PF (1983). Fasting serum cholesterol and triglycerides in a ten-year prospective study in old age. Acta Med Scand 214:33–34.
Benn M, Watts GF, Tybjærg-Hansen A, Nordestgaard BG (2016). Mutations causative of familial hypercholesterolaemia: screening of 98 098 individuals from the Copenhagen General Population Study estimated a prevalence of 1 in 217. Eur Heart J 37:1384–1394.
Besseling J, Hovingh GK, Huijgen R, Kastelein JJ, Hutten BA (2016). Statins in familial hypercholesterolemia: consequences for coronary artery disease and all-cause mortality. J Am Coll Cardiol 68:252–260.
Beydoun MA, Beydoun HA, Dore GA, Fanelli-Kuczmarski MT, Evans MK, Zonderman AB (2015). Total serum cholesterol, atherogenic indices and their longitudinal association with depressive symptoms among US adults. Transl Psychiatry 5:e518.
Bombelli M, Facchetti R, Fodri D, Brambilla G, Sega R, Grassi G, Mancia G (2013). Impact of body mass index and waist circumference on the cardiovascular risk and all-cause death in a general population: data from the PAMELA study. Nutr Metab Cardiovasc Dis 23:650–656.
Chen W, Zheng R, Zhang S, Zhao P, Li G, Wu L, He J (2013). Report of incidence and mortality in China cancer
registries, 2009. Chin J Cancer
Cui R, Iso H, Yamagishi K, Saito I, Kokubo Y, Inoue M, Tsugane S, JPHC Study Group (2012). High serum total cholesterol levels is a risk factor
of ischemic stroke for general Japanese population: the JPHC study. Atherosclerosis 221:565–569.
Freeman MR, Solomon KR (2004). Cholesterol and prostate cancer
. J Cell Biochem 91:54–69.
Heiniger HJ, Brunner KT, Cerottini JC (1978). Cholesterol is a critical cellular component to T-lymphocyte cytotoxicity. Proc Natl Acad Sci USA 75:5683–5687.
Heron M, Anderson RN (2016). Changes in the leading cause of death: recent patterns in heart disease and cancer
mortality. Hyattsville, MD: National Center for Health Statistics. NCHS Data Brief 254:1–8.
Hiatt RA, Fireman BH (1986). Serum cholesterol and the incidence of cancer
in a large cohort. J Chr Dis 39:861–870.
International Collaborative Group (1982). Circulating cholesterol level and risk of death from cancer
in men aged 40–69 years. JAMA 248:2853–2859.
Iqbal A, Zaid M, Munir R, Usman H, Kalbacher H, Scandiuzzi L, Zaidi N (2016). Atypical plasma lipid profiles in leukemia. Clin Chim Acta 452:129–133.
Jung KW, Won YJ, Kong HJ, Oh CM, Seo HG, Lee JS (2013). Cancer
statistics in Korea: incidence, mortality, survival and prevalence in 2010. Cancer
Res Treat 45:1–4.
Kark JD, Smith AH, Hames CG (1982). Serum retinol and the inverse relationship between serum cholesterol and cancer
. Br Med J 284:152–154.
Knekt P, Reunanen A, Aromaa A, Heliovaara M, Hakulinen T, Hakama M (1988). Serum cholesterol and risk of cancer
in a cohort of 39,000 men and women. J Clin Epidemiol 41:519–530.
LaRosa JC, Pedersen TR, Somaratne R, Wasserman SM (2013). Safety and effect of very low levels of low-density lipoprotein cholesterol on cardiovascular events. Am J Cardiol 111:1221–1229.
Liu J, Zeng FF, Liu ZM, Zhang CX, Ling WH, Chen YM (2013). Effects of blood triglycerides on cardiovascular and all-cause mortality: a systematic review and meta-analysis of 61 prospective studies. Lipids Health Dis 12:1.
Malvezzi M, Bertuccio P, Levi F, La Vecchia C, Negri E (2013). European cancer
mortality predictions for the year 2013. Ann Oncol 24:792–800.
Mamtani R, Lewis JD, Scott FI, Ahmad T, Goldberg DS, Datta J, et al (2016). Disentangling the association between statins, cholesterol, and colorectal cancer
: a nested case-control study. PLoS Med 13:e1002007.
Marrer E, Wagner A, Montaye M, Luc G, Amouyel P, Dallongeville J, et al (2013). Lipoprotein(a) plasma levels and the risk of cancer
: the PRIME study. Eur J Cancer
Martin LJ, Melnichouk O, Huszti E, Connelly PW, Greenberg CV, Minkin S, Boyd NF (2015). Serum lipids, lipoproteins, and risk of breast cancer
: a nested case-control study using multiple time points. J Natl Cancer
Menkens MS, Constock GW, Vuillemier JP, Helsing KJ, Rider AA, Brookmeyer R (1986). Serum beta-carotene, vitamins A and E, selenium, and the risk of lung cancer
. N Engl J Med 315:1250–1254.
Nishikura T, Koba S, Yokota Y, Hirano T, Tsunoda F, Shoji M, et al (2014). Elevated small dense low-density lipoprotein cholesterol as a predictor for future cardiovascular events in patients with stable coronary artery disease. J Atheroscler Thromb 21:755–767.
Prineas RJ (1978). Blood pressure sounds: their measurements and meaning A training manual. Minneapolis: Gamma Medical Products.
Ravnskov U, Diamond DM, Hama R, Hamazaki T, Hammarskjöld B, Hynes N, et al (2016). Lack of an association or an inverse association between low-density lipoprotein cholesterol and mortality in the elderly: a systematic review. BMJ Open 6:e010401.
Reed DR, Yano K, Kagan A (1986). Lipids and lipoproteins as predictors of coronary heart disease, stroke and cancer
in the Honolulu Heart Program. Am J Med 80:871–878.
Rose G (1965). Standardization of observers in blood pressure measurements. Lancet I:673–674.
Rose G, Blackburn H (1968). Cardiovascular survey methods. Geneva: WHO.
Sabine J (1977). Cholesterol. New York: Marcel Dekker Inc.
Sahebzamani FM, D’Aoust RF, Friedrich D, Aiyer AN, Reis SE, Kip KE (2013). Relationship among low cholesterol
levels, depressive symptoms, aggression, hostility, and cynicism. J Clin Lipidol 7:208–216.
Singh GM, Danaei G, Farzadfar F, Stevens GA, Woodward M, Wormser D, et al (2013). The age-specific quantitative effects of metabolic risk factors on cardiovascular diseases and diabetes: a pooled analysis. PLoS One 8:e65174.
Smith GD, Shipley MJ, Marmot MG, Rose G (1992). Plasma cholesterol concentration and mortality. The Whitehall study. J Am Med Assoc 267:70–76.
Stahelin HB, Gey KF, Eichholzer M, Ludin E (1991). Beta-carotene and cancer
prevention: the basel study. Am J Clin Nutr 53 (Suppl 1):265s–269s.
Stefanko A, Thiede C, Ehninger G, Simons K, Grzybek M (2017). Lipidomic approach for stratification of acute myeloid leukemia patients. PLoS One 12:e0168781. doi: 10.1371/journal.pone.0168781.
Strohmaier S, Edlinger M, Manjer J, Stocks T, Bjørge T, Borena W, et al (2013). Total serum cholesterol and cancer
incidence in the metabolic syndrome and cancer
project (Me-Can). PLoS One 8:e54242.
Tamosiunas A, Luksiene D, Baceviciene M, Bernotiene G, Radisauskas R, Malinauskiene V, et al (2014). Health factors and risk of all-cause, cardiovascular, and coronary heart disease mortality: findings from the MONICA and HAPIEE studies in Lithuania. PLoS One 9:e114283.
Tanaka T, Okamura T (2012). Blood cholesterol level and risk of stroke in community-based or worksite cohort studies: a review of Japanese cohort studies in the past 20 years. Keio J Med 61:79–88.
The RIFLE Research Group (1993). Presentation of the RIFLE project, risk factors and life expectancy. Eur J Epidemiol 9:459–476.
Trialists CT (2010). Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomized trials. Lancet 376:1670–1681.
Statistics Working Group (2015). United States cancer
statistics: 1999–2012 incidence and mortality web-based report. Atlanta: Department of Health and Human Services, Centers for Disease Control and Prevention, and National Cancer
Wald NJ, Boreham J, Hayward JL, Bulbrook RD (1984). Plasma retinol, beta-carotene and vitamin E levels in relation to the future risk of breast cancer
. Br J Cancer
Whelton SP, Roy P, Astor BC, Zhang L, Hoogeveen RC, Ballantyne CM, Coresh J (2013). Elevated high-sensitivity C-reactive protein as a risk marker of the attenuated relationship between serum cholesterol and cardiovascular events at older age. The ARIC Study. Am J Epidemiol 178:1076–1084.
WHO (2010). International statistical classification of diseases and related health problems – 10th revision, 2010 ed. Geneva, Switzerland: WHO.
Wingard DL, Criqui MH, Holdbrook MJ, Barrett-Connor E (1984). Plasma cholesterol and cancer
morbidity and mortality in an adult community. J Chronic Dis 37:401–406.