Scientists investigating how diabetes affects amyotrophic lateral sclerosis risk are struggling to reconcile conflicting results. In the June 1 JAMA Neurology, researchers reported that diabetes protected against the motor neuron disease in Danish citizens. However, a paper in the May 2 Journal of Epidemiology concluded that, especially in younger men, diabetes increased the odds of developing ALS in people in Taiwan. Regardless, a study in the April 21 Muscle & Nerve claimed that diabetes made little difference to the prognosis of people already diagnosed with ALS. Scientists agreed they have insufficient data to make definite conclusions about diabetes and ALS risk. “This is just the beginning, not the end, of this issue,” said Chung-Yi Li of National Cheng Kung University in Tainan City, Taiwan, senior author on the J. Epidemiology paper. Researchers hope that if they can understand how diabetes affects ALS, it might point them to molecular pathways that influence the motor-neuron disease, which could be amenable to drug therapy.

Several studies have indicated that people with ALS survive longer if they are overweight, and that lean athletes might be at higher risk for the condition (see May 2011 newsScarmeas et al., 2002). Excess weight is a risk factor for diabetes, which has already been linked to neurodegenerative conditions such as Alzheimer’s (see AlzRisk). These facts have led neurologists to wonder about a possible association between diabetes and ALS.

Diabetes and Risk—Contradictory Findings
In the JAMA Neurology paper, Marianthi-Anna Kioumourtzoglou of the Harvard T.H. Chan School of Public Health and colleagues analyzed diabetes and ALS rates in Denmark. From the Danish National Patient Register of hospital visits, they identified 3,650 people who were treated for ALS from 1992 to 2009 and compared them to 365,000 ALS-free controls from a database of all Danes. They used the hospital registry to identify diabetics in both groups. Overall, people with diabetes had a 39 percent lower chance of having ALS than others.

Kioumourtzoglou’s results match with a recent study of the Swedish population by Daniela Mariosa and Fang Fang at the Karolinska Institutet in Stockholm (Mariosa et al., 2015). This study of 5,108 ALS cases and 25,540 controls suggested that diabetes was associated with a 21 percent reduction in ALS risk in Sweden.

However, the J. Epidemiology paper paints a different picture. First author Yu Sun of the En Chu Kong Hospital in New Taipei City used Taiwan’s National Health Insurance Research Database to identify 615,492 people who had received treatment for diabetes in 2000, but had never been diagnosed with ALS. The authors compared them to a control cohort of 614,835 with no evidence of either condition. By the end of 2008, 255 people in the diabetes group had been diagnosed with ALS, compared to 201 from the control cohort. Overall, people with diabetes had a 35 percent greater risk of ALS. When the authors broke their data down by gender and age, they calculated that men younger than 65 were most prone—diabetes came with a 67 percent higher risk of ALS. Li speculated that for older people, the risk from age outweighs any risk from diabetes.

“The Taiwanese results are throwing me for a loop,” said Marc Weisskopf of the Harvard T.H. Chan School of Public Health, senior author on the Danish study. “How we reconcile these results is not clear.” Perhaps the different genetic backgrounds of Scandinavians and Taiwanese account for the discrepancy, suggested Luc Dupuis of the Université de Strasbourg in France (see full comment below). Weisskopf also speculated that the two regions might differ in average body mass index or in treatments for diabetes. “I think we need to look more carefully at the profile of health factors that come with diabetes, and whether they differ between Taiwan and Scandinavia,” he said.

Type Matters
Neither Kioumourtzoglou nor Li were able to fully resolve their diabetes cases into type 1 and type 2. Li told Alzforum that in the Taiwanese population, the fraction of diabetics with the type 1 form accounted for less than 2 percent of total diabetes cases, so his data set must have been mostly type 2.

The Danish registry does distinguish between types, but doctors often enter the wrong code, said Kioumourtzoglou. However, she was able to use age as a proxy for diabetes type. Type 1 tends to arise in younger people than type 2 does. People diagnosed with diabetes after age 40 were protected from ALS, with a 48 percent lower risk. Conversely, those who had diabetes before 40 had a 66 percent higher risk, though this was not statistically significant because there were so few people in that group. “What we saw would indicate that type 2 diabetes is protective, and that type 1 might be a risk factor,” said Kioumourtzoglou.

Mariosa’s study hinted at something similar. In the Swedish data set, diabetes only protected people older than 70. Furthermore, diabetics treated with insulin—who are mostly type 1—who were younger than 50 had fivefold greater risk of ALS than normal. Overall, the Scandinavian studies suggest that people who were older and likely to have type 2 diabetes were protected from ALS, while perhaps those who were younger and had type 1 had increased risk.   

Diabetes and Prognosis—No Effect
What about after ALS starts—does diabetes change one’s prognosis? If that were true, it would suggest that some treatment to alter metabolism or glucose processing in people with ALS might be beneficial. In the April 21 Muscle & Nerve online, researchers led by senior author Timothy Miller of the Washington University School of Medicine in St. Louis asked if type 2 diabetes affected ALS progression. They collected data on 1,322 people with ALS who had participated in one of six previous clinical trials. Of these, 5.4 percent had diabetes at the start of their trial. However, this made no difference to how long they survived. “Once you have ALS, our data strongly suggest that there is no impact of diabetes,” said first author Sabrina Paganoni of Massachusetts General Hospital in Boston.

Weisskopf was disappointed by these findings. He was hoping that if diabetes or diabetes treatments changed the progression rate for ALS, it would offer some clue to treating ALS, but the clinical trial data suggest otherwise. Paganoni’s study was well done, he said, but not necessarily the final word. He speculated that some aspects of diabetes, or its treatment, might slow ALS progression, while other aspects might speed up the disease, such that the benefits and hazards canceled each other out in those trials.

The results were surprising to Mariosa and Fang, as well. “Since diabetes leads to complications and tends to have a negative effect on general survival, it is interesting to note that ALS patients with diabetes seemed to live just as long as ALS patients without this condition,” they wrote to Alzforum.

Though Paganoni and colleagues focused on diabetes and ALS prognosis, their evidence supported the idea that diabetes cuts down ALS risk. They found much lower rates of type 2 diabetes in people who came to ALS clinics, or participated in ALS trials, than in the general population.

“The story on diabetes and ALS is gradually coming along,” wrote Mariosa and Fang. “The important task now is to disentangle diabetes by subtype, duration, and severity.”—Amber Dance


  1. The study by Paganoni is a retrospective analysis of six clinical trials reanalyzed for the influence of diabetes on ALS survival. The authors found no such effect, meaning that the diabetic status did not correlate with either improved or worsened survival. Contrastingly, and consistent with multiple other studies, the authors found that patients with higher initial body mass index (BMI) had increased survival. Paganoni and collaborators also observed lower-than-expected frequency of diabetes in ALS patients in their cohort, and confirmed this observation in two other ALS cohorts. The authors are cautious in their interpretation since multiple biases might be responsible for this lower-than-expected frequency; yet, this latter observation is entirely consistent with previous studies by the Fang group showing decreased risk of ALS in diabetic patients.

    Intriguingly, the second study shows the opposite relationship, with increased risk of ALS in younger men with diabetes. The reasons for these discrepancies are not known, but might be related to different genetic backgrounds. Further studies are required to clarify this difference.

    In general, both studies are consistent with a role of altered energy metabolism in ALS. Diabetes is, however, a very distant proxy of metabolic health, and it is not surprising that BMI correlates better with survival than diabetes. These studies remain correlative, and a causal relationship between weight loss and ALS remains to be established. The first step in this direction was recently provided by Wills and collaborators, who showed that hypercaloric nutrition protects ALS patients who had needed gastrostomy for survival (Wills et al., 2014). These data suggest we urgently need to perform interventional studies in less-advanced ALS patients to determine whether providing energy-rich diets slows down disease progression.


    . Hypercaloric enteral nutrition in patients with amyotrophic lateral sclerosis: a randomised, double-blind, placebo-controlled phase 2 trial. Lancet. 2014 Jun 14;383(9934):2065-72. Epub 2014 Feb 28 PubMed.

  2. There are no obvious methodological flaws in these three studies. However, there are similarities and dissimilarities.

    In the age-specific analyses, these three studies all showed a higher relative risk of ALS in relation to diabetes in younger ages than in older ages.  

    The Swedish study showed a protective effect of diabetes only in those aged 70 years and older; similar findings were noted in those older than 60 in the U.S. study. While the Taiwanese study showed a significant elevated hazard ratio for men younger than 65 with diabetes, it did not show such adverse effect of diabetes in people aged 65 years and older.

    The co-morbidities and other socioeconomic factors considered in the analysis were not the same in these three studies. For example, the Swedish study considered education, occupation-based socioeconomic status, age, sex, and area of residence in its analysis. In addition to socioeconomic status, the U.S. study additionally adjusted for chronic obstructive pulmonary disorder, obesity, and marital status. Compared to the U.S. and Swedish studies, the Taiwanese study took more factors into consideration, but I don’t believe this explains why the Taiwanese study arrived at different conclusions.

    The other difference in these three studies is the age of study population. The mean age in the Taiwanese study was 60, but it was 66 in the U.S. study and 68 in the Swedish.

    All told, there was a tendency in these three studies for diabetes to be a risk factor for ALS in younger ages, but it may pose a protective effect on risk of ALS overall.

    Therefore, if we took only a single value showing the overall effect of diabetes on ALS, the Taiwanese study may be overrepresented by younger people, and both the Swedish and U.S. studies tended to show an association of diabetes with ALS in an older people. I believe such a “population mix” is the most likely reason for the dissimilarities in findings between ours and the other two studies.

    Based on what I observed. I would make the following three suggestions:

    1. Age could significantly modify the relationship between diabetes and ALS. Therefore, future studies should present age-specific relative risk of ALS in relation to diabetes.

    2. We need more studies that include diabetics with wider age ranges (e.g., younger than 50 years or older than 70 years) to see if age is truly a significant effect-modifier.

    3. For our own Taiwanese study, or other future studies, it would be better to make a more detailed age-stratified analyses (like the Swedish study) to look into the age-specific relationship between diabetes and ALS.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Being Pleasantly Plump: Way to Live Longest with ALS?

Paper Citations

  1. . Premorbid weight, body mass, and varsity athletics in ALS. Neurology. 2002 Sep 10;59(5):773-5. PubMed.
  2. . Association between diabetes and amyotrophic lateral sclerosis in Sweden. Eur J Neurol. 2015 Jan 19; PubMed.

External Citations

  1. AlzRisk

Further Reading


  1. . Effects of diabetes mellitus on amyotrophic lateral sclerosis: a systematic review. BMC Res Notes. 2014 Mar 24;7:171. PubMed.
  2. . Insulin resistance in amyotrophic lateral sclerosis. J Neurol Sci. 1984 Mar;63(3):317-24. PubMed.
  3. . Reduction in insulin receptors in amyotrophic lateral sclerosis correlates with reduced insulin sensitivity. Neurology. 1987 Aug;37(8):1375-9. PubMed.
  4. . Impaired glucose tolerance in patients with amyotrophic lateral sclerosis. Amyotroph Lateral Scler. 2010;11(1-2):166-71. PubMed.
  5. . Vascular risk factors: imaging and neuropathologic correlates. J Alzheimers Dis. 2010;20(3):699-709. PubMed.
  6. . Antecedent Disease is Less Prevalent in Amyotrophic Lateral Sclerosis. Neurodegener Dis. 2015;15(2):109-13. Epub 2015 Feb 20 PubMed.
  7. . Trials of Antidiabetic Drugs in Amyotrophic Lateral Sclerosis: Proceed with Caution?. Neurodegener Dis. 2013 Oct 2; PubMed.

Primary Papers

  1. . Diabetes Mellitus, Obesity, and Diagnosis of Amyotrophic Lateral Sclerosis: A Population-Based Study. JAMA Neurol. 2015 Aug;72(8):905-11. PubMed.
  2. . Risk of Amyotrophic Lateral Sclerosis in Patients With Diabetes: A Nationwide Population-Based Cohort Study. J Epidemiol. 2015;25(6):445-51. Epub 2015 May 2 PubMed.
  3. . Pre-morbid type 2 diabetes mellitus is not a prognostic factor in amyotrophic lateral sclerosis. Muscle Nerve. 2015 Sep;52(3):339-43. Epub 2015 Jun 1 PubMed.