Observing the causal effect of exposure to a disease without the need for a traditional randomized clinical trial.

- Posted in English by

For anyone studying research against neurodegenerative diseases, it is striking to note the large number of studies which each claim to have identified a key element, different in each study, and which would be a causative factor of the disease. In addition, many studies are contradictory with each other.

In epidemiology, Mendelian randomization is a method of using measured gene variation, known to express the causal effect of exposure to a disease in observational studies, without the need for a traditional randomized clinical trial. Even better, it allows us to escape traditional biases in epidemiological studies, such as reverse causation and confusion. This method was first proposed in 1986 by Gray and Wheatley.

Since genotypes are randomly assigned when passed from parents to offspring during meiosis, the distribution of the population genotype should not be linked to the confounding factors that generally affect observational epidemiological studies. In this regard, Mendelian randomization can be viewed as a randomized controlled trial.

Because polymorphism is the instrument, Mendelian randomization depends on previous genetic association studies that have provided good candidate genes for the response to risk exposure.

Each of these selected genetic variants must satisfy three conditions, relevance, independence, exclusion restriction.

Among the various genetic and environmental factors that have been identified to be associated with ALS, the association between blood lipid metabolites and ALS has recently received considerable attention. The associations between lipids and ALS are strong and comparable in strength to many risk factors for ALS previously identified.

ALS patients suffer from increased energy expenditure at rest and from weight loss. Previous observational studies have shown that ALS patients frequently experience dyslipidemia. Dyslipidemia is characterized by abnormal levels of high density lipoproteins (HDL), low density lipoproteins (LDL), total cholesterol (TC) and triglycerides (TG).

The positive association between dyslipidemia and ALS suggests that elevated levels of non-HDL lipids may play a protective role in the progression of ALS. Consistent with observational studies in humans, research with ALS mouse models has also shown that the overall survival of ALS mice is reduced under calorie restriction. However, the relationship between dyslipidemia and ALS is also controversial, conflicting results have been reported for basal serum lipid levels, the cause of dyslipidemia, and the relationship between serum lipid levels and the progression of ALS disease.

For example, many observational studies following ALS have found no association between dyslipidemia and ALS. In addition, some studies have shown that patients with ALS often suffer from hypolipidemia - which is mainly characterized by low levels of LDL - in men and women with ALS. The association between hypolipidemia and ALS is further confirmed in a mouse ALS model. The conflicting results on the relationship between lipid levels and ALS may be due in part to the relatively small sample sizes used in previous studies and in part to uncontrolled confounders that are inevitable in observational studies .

Determining the causal impact of lipids on ALS is difficult using traditional randomized controlled trial studies because these studies necessarily require long-term follow-up, are expensive and often unethical. Therefore, it is desirable to determine the causal relationship between lipids and ALS through observational studies. Mendelian randomization is a powerful statistical tool for examining the causal relationship and estimating causal effects in observational studies.

Scientists have studied the causal effects of four blood lipid traits on the risk of ALS:

  • high density lipoprotein,
  • low density lipoprotein (LDL),
  • total cholesterol,
  • triglycerides.

The authors first selected SNPs (genetic variants) which can serve as valid instrumental variables for each of the four lipid traits (HDL, LDL, TC and TG).

Taking advantage of the instrument variables from several large-scale association studies on the genome in European and Asian populations, the authors performed one of the most important and comprehensive Mendelian randomization analyzes to date on the causal relationship between lipids and ALS. Among the four lipids, they found that only LDL is causally associated with ALS and that a higher level of LDL increases the risk of ALS in European and East Asian populations.

The large sample size used in this study allows the authors to fully establish a positive causal effect of the modifiable factor LDL on ALS in European and East Asian populations. The inferred causal relationship between LDL and ALS is robust in the choice of statistical methods and is carefully validated by various sensitivity analyzes.

The positive causal effect of LDL on ALS suggests that future development of strategies to reduce LDL levels would likely reduce the burden of ALS. LDL is a modifiable risk factor, the levels of which can be reduced by various intervention strategies. For example, dietary changes such as increased fiber intake, increased phytosterol consumption, and increased consumption of nuts can all lead to a reduction in LDL levels.

Restrictions on dietary cholesterol, restrictions on high-carbohydrate diets, and restrictions on the consumption of trans fats can also lower LDL levels. In addition to lifestyle and dietary changes, LDL reduction can be achieved by drug therapy.

This is not a breakthrough regarding the mechanism of onset of the disease, but it introduces a tool for managing this disease.

The future development of LDL reduction strategies and the development of public policies to promote such strategies are likely to reduce the burden of ALS in society.


This book retraces the main achievements of ALS research over the last 30 years, presents the drugs under clinical trial, as well as ongoing research on future treatments likely to be able stop the disease in a few years and to provide a complete cure in a decade or two.

Please, to help us continue to provide valuable information: