The main finding of this cohort study is that LDL-C is a significant and independent predictor of all-cause mortality in community-dwelling adults. After adjustment for possible confounding factors, the results showed that participants with the very low LDL-C levels (< 70 mg/dL) were at a significantly higher risk for all-cause mortality than those with high LDL-C levels (≥ 144 mg/dL). To the best of our knowledge, few studies have demonstrated the relationship between LDL-C level and all-cause mortality in Japanese community-dwelling persons.
The results of this study, especially that low LDL-C levels are significantly associated with an increased risk of all-cause mortality, are consistent with the results of several existing studies. The Kangbuk Samsung Health Study on 347,971 individuals (mean age: 39.6 years old; male: 57.4%; mean follow-up: 5.64 ± 3.27 years) highlighted that the lowest LDL-C group (< 70 mg/dL) was at a higher risk of all-cause mortality (HR: 1.81; 95% CI: 1.44–2.28) compared with the reference group (120–139 mg/dL) . Further, Johannesen et al.  reported 2028 deaths among the total number of participants and 11,376 (10.5%) deaths during the study among the 108,243 individuals aged 20–100 years (male: 45.0%; median follow-up: 9.4 years). The study also showed a U-shaped relationship between LDL-C levels and the risk of all-cause mortality: that is, low levels (< 70 mg/dL; HR: 1.25, 95% CI: 1.15–1.36) and high levels (> 189 mg/dL; HR: 1.15, 95% CI: 1.05–1.27) were associated with an increased risk of all-cause mortality compared with the reference group (132–154 mg/L). The China Health and Retirement Longitudinal Study (follow-up: 4 years) recorded a total of 305 deaths out of 4981 male participants. Compared with the LDL-C baseline group (117–137 mg/dL), a lower LDL-C level (≤ 84 mg/dL) was associated with an increased risk of four-year all-cause mortality in middle-aged and older adult Chinese male participants . According to a recent systematic review of 19 cohort studies with more than 68,094 older adults, all-cause mortality was highest in the lowest LDL-C quartile group . In addition, a population-based register study on 118,160 individuals aged 50 years or older without baseline statin use showed an association between high LDL-C levels and lower mortality among older adults . The present study also reports that low LDL-C levels at baseline, as well as being male, older, having lower BMI, and having a history of diabetes, were linked with an increase in all-cause mortality. The association between LDL-C levels and all-cause mortality was particularly significant for male participants and those who have CKD. The opposite trend was not observed in people with a BMI of 25 kg/m2 or higher, those with a history of CVD, or those with diabetes. Higher LDL-C levels are often observed in these patient groups [19, 20], but this does not necessarily lead to increased mortality . In addition, for those with low LDL-C levels due to lipid-lowering medication, all-cause mortality was not significantly increased; for these individuals, low LDL-C levels were intentional, rather than an indicator of poor prognosis. However, in the study with 118,160 participants, it was reported that those with the highest LDL-C levels lived longer than those on lipid-lowering medication .
The study examined the association between LDL-C levels and mortality outcomes in a real-world setting among community-dwelling individuals, including those on lipid-lowering therapy and those with a baseline history of CVD, hypertension, or diabetes. However, the sub-analysis shows similar notable findings for those who were not undergoing lipid-lowering therapy or had no other diseases. The findings suggest that considerably lower LDL-C levels do not necessarily protect against all-cause mortality among community-dwelling persons who are not on lipid-lowering medication, thus supporting the lipid paradox . In addition, the difference in results between male and female participants can be attributed to fewer deaths among women than men, which results in insufficient association power . We believed that baseline serum LDL-C levels would be positively associated with CVD  and mortality in middle-aged people. On the other hand, a negative association between LDL-C levels and all-cause mortality has been observed in the elderly [23, 24]. In this population, mortality from non-CVD increased with decreasing LDL-C, which could be due to malnutrition or infectious diseases [25,26,27,28]. This would explain the reversed association between LDL-C levels and all-cause mortality that was only observed in the older population. We discuss this further in the following paragraph. In our study, the reversed association was consistently found in the age group of 55 years and older, but no significant association was found among the four LDL-C groups in participants younger than 55 years.
The mechanisms leading to increased all-cause mortality in individuals with very low LDL-C levels are not completely understood. Several explanations can be offered for these findings. Low LDL-C levels increase susceptibility to serious diseases . Conversely, it has been hypothesized that frailty and illnesses lower cholesterol levels . The latter concerns that cholesterol may be lowered by serious diseases shortly before death could be dispelled because mortality during the 3 years of observation were excluded in the current study. LDL-C may protect against viruses and cancers caused by viruses, and is therefore a component of innate immunity . Ravnskov et al.  reviewed nine cohort studies involving more than 140,000 individuals. The review showed that the occurrence of cancer was inversely associated with cholesterol levels measured 10–30 years earlier, and the association remained when cancer cases that appeared during the first 4 years were excluded. Moreover, the lowering of lipids in rodents have led to cancer in previous experiments . In addition, exotoxins produced by Gram-positive bacteria are absorbed by LDL-C . Thus, higher LDL-C levels have been associated with reduced infection-related mortality and other non-CVD mortality, which explains the inverse relationship with all-cause mortality . In this study, comorbidities such as hypertriglyceridemia, low HDL cholesterolemia, and hyperuricemia were more frequently observed in individuals with the lowest LDL-C levels. Finally, the relationship between LDL-C and mortality from coronary artery disease has also been reported in a paradoxical hypothesis regarding the relationship with mortality, and patients with higher LDL-C at admission had not higher all-cause mortality compared to patients with normal or low LDL-C [32, 33]. If high LDL-C were the cause, the effect should have been the opposite.
Study strengths and limitations
The strengths of this study are the fact that it is a long-term follow-up collection, the sample size, the adjustment for possible confounding factors, and the inclusion of sensitivity analyses. However, the authors acknowledge some limitations. First, the sample consisted mainly of relatively healthy middle-aged and elderly people (mean age: 64 ± 13 years) who lived in rural areas of Japan and participated in the health checkup. Therefore, it cannot be considered representative of the general population. Second, the survey covered people whose deaths were registered in the basic resident register. Those who moved out of the region during the survey period are not included. Third, the possible effects of medication (e.g., antihypertensive, lipid-lowering, and antidiabetic medication), underlying diseases, and lifestyle modifications at the baseline and during the follow-up period on the present findings cannot be overlooked. Fourth, the threshold for the high LDL-C group may have been too low to evaluate the U-shaped relationship between LDL-C levels and all-cause mortality. If high LDL-C was defined as 190 mg/dL or higher in this study, the high LDL-C group would have included 52 individuals, with a 7.7% mortality rate during the observation period, which was not significantly different from the reference value (data not shown). Fifth, this study did not measure certain specific lipoproteins (e.g., small dense LDL), which could be a possible explanation for this phenomenon. Finally, the relatively low number of participants and deaths may weaken the causal relationship between LDL-C levels and all-cause mortality.