New observational data add one more reason to strive for a healthy middle age. Researchers led by Helena Hörder, University of Gothenburg, Sweden, report that women who scored high on a fitness test in midlife were nearly 90 percent less likely than their moderately fit or unfit peers to develop dementia decades later. Also, the fittest women held dementia at bay 10 years longer. The study, now available online, appears in the April 10 Neurology.

  • From 1968, dementia incidence was tracked among women who were then middle aged.
  • Over 44 years, the fittest women were nearly 90 percent less likely to develop dementia.
  • If they did, they got it 10 years later than did less-fit women.

“This is a very important study because of the extremely long follow-up,” said Prashanthi Vemuri, Mayo Clinic in Rochester, Minnesota. Miia Kivipelto, National Institute for Health and Welfare, Helsinki, also valued the report. “Together with previous studies, it supports the idea that what is good for the heart is also good for the brain,” she said. Both cautioned, however, that the sample size was small.

Several observational studies have suggested physical activity lowers the risk of dementia, and a handful of clinical trials testing the effects of lifestyle interventions, including exercise, support this proposal (e.g., Norton et al., 2014; Blondell et al., 2014; Nov 2015 conference news). Compiling data from many studies, a commission convened by The Lancet attributed physical inactivity late in life to a 3 percent increase in dementia risk (Livingston et al., 2017). 

To gain a long-term view of the link between physical fitness and brain health, Hörder mined data from the Prospective Population Study of Women. This tracked the health of 1,462 middle-aged Swedish women beginning in 1968. Hörder analyzed data from a subgroup of 191 women who were 38–60 years old at that time. She focused specifically on fitness, which is determined by both physical activity and genetic factors. At baseline, the women rode stationary bikes fitted with ergometers that measured their work output. First, they pedaled easy for six minutes, producing 32 and then 64 watts. A fast walk is equivalent to about 100 watts. After a five-minute break, the women began pedaling again, but this time, based on their performance on the first test, the resistance was adjusted to measure their peak workload. Participants pedaled as hard as they could at this load until they felt exhausted and had to stop, typically after about six minutes. If they were still going strong at six minutes, study researchers added an additional 10 watts every minute to the workload, until they had to stop pedaling (Bengtsson et al., 1978). To track dementia, Hörder used data from neuropsychological exams and interviews conducted over a follow-up period of 44 years. Each person was tracked, on average, for 29 years. During this time, 44 women developed dementia.

On average, participants put out 103 watts at peak load. Output ranged from less than 48 watts for those unable to complete the test, often because their heart rate or blood pressure rose too much, to more than 136 watts for the fittest. The 103-watt mean output was lower than recent reference measurements for middle-aged women: 125 watts for 40- to 49-year-olds and 117 watts for those 50–59 (Loe et al., 2013). “This is probably because women in the 1960s were not that physically active,” noted Hörder, adding that differences in the test procedures may also contribute to the discrepancy.

The researchers grouped the participants into low-, medium-, and high-fitness categories, which generated 80 watts, 88–112 watts, or more than 120 watts, respectively. Anyone unable to complete the test was put in the low-fitness bin. Examinations from the 44-year study as well as data from the Swedish Hospital Discharge Register until December 2012, indicated that 19 of 59 women in the low-fitness group developed some form of dementia, 23 of 92 women with medium fitness did, and only two of 40 women from the high-fitness group did, and it was not Alzheimer’s disease. After Hörder adjusted for age, height, triglyceride levels, hypertension, smoking, wine consumption, physical inactivity, and income, the fittest women had a hazard ratio of 0.12 for all-cause dementia relative to the medium fitness group. For the least-fit women, the hazard ratio was 1.4, which was significantly different from the fittest group, but not from the medium group. The authors noted that the fitness effect was stronger than that previously reported for physical activity (Hamer and Chida, 2009; Sofi et al., 2011). The cumulative incidence of dementia was 32 percent for those in the low-fitness group, 25 percent for the medium, and 5 percent for high fitness.

Women who failed to complete the test ran the highest risk of developing dementia, with 45 percent becoming demented during the follow-up period. “These results suggest that underlying poor cardiovascular health may partially explain the relationship between fitness and brain health,” wrote Nicole Spartano, Boston University, and Tiia Ngandu, National Institute for Health and Welfare, Helsinki, in a Neurology editorial. Notably, the two women in the high-fitness group who developed dementia did so very late in life, when they were 90 years old. The average age of onset was 79 for the medium-fitness group, and 81 for the low group.

Hörder and colleagues cautioned that the study has several limitations. For example, the less-healthy women likely received more medical care than fitter women, which could indirectly influence dementia risk while increasing the likelihood of it being diagnosed earlier.

Overall, the results are consistent with shorter observational studies reporting associations between cardiovascular fitness and dementia (Defina et al., 2013; Nyberg et al., 2014; Kulmala et al., 2014). The long follow-up of the new study, however, builds a stronger bridge between middle and old age, said Kivipelto, who noted that it avoids the potential for reverse causation that might confound studies of older people. In those individuals, dementia might limit fitness, rather than fitness protecting from dementia.

Should all middle-aged women exercise strenuously to stave off dementia? Genetics, age, gender, and body size also play a role. “A very fit person may not necessarily be a very active person, and vice versa,” noted Spartano and Ngandu. The combined genetic and environmental effects may explain why the effect of fitness seems greater than that reported for physical activity alone, noted Kivipelto. Also, she and Vemuri said that more data are needed. “We don’t have conclusive data yet to say what level of exercise is going to be best,” said Vemuri.

Going forward, Hörder wants to search for links between fitness and MRI scans from this cohort. Vemuri suggested examining vascular health in more detail, noting that it plays both direct and indirect roles in dementia-associated neurodegeneration, without significantly affecting amyloid deposition (Vemuri et al., 2017).—Marina Chicurel


  1. While this study is interesting since it seeks to correlate midlife (rather than elderly) fitness with AD risk, it has two major limitations that caution against reading too much into these findings. In addition to the small sample size mentioned above by Drs. Vemuri and Kivipelto, the authors did not report any APOE genotype information for the sample, which consisted of 191 women divided into low (n=59), medium (n=92), and high (n=40) levels of midlife fitness. It is easily possible with a sample of this size that the high fitness group randomly had fewer carriers of the APOE E4 allele than would be expected from population genetics (or, the other groups had more carriers than expected).

    The authors mention in their discussion that without this genetic information they "cannot draw any conclusions about the impact of APOE E4 on the relation between fitness and dementia." But my feeling is that without this genetic information, a study of this size should not draw any conclusions about the relation between fitness and dementia, period. Given the ~300 percent increase in AD risk conferred by each copy of the APOE E4 allele that one might carry, correcting for this genetic factor is an essential component for any epidemiological study of AD risk, particularly when the sample is small.

    Despite these limitations, the study does provide some added motivation for me to increase the frequency of my visits to our company's physical fitness facility!

  2. I absolutely agree that favourable genes (e.g not APOE4) could explain both the high cardiovascular fitness (healthy heart) and the decreased dementia risk (healthy brain).

Make a Comment

To make a comment you must login or register.


News Citations

  1. Health Interventions Boost Cognition—But Do They Delay Dementia?

Paper Citations

  1. . Potential for primary prevention of Alzheimer's disease: an analysis of population-based data. Lancet Neurol. 2014 Aug;13(8):788-94. PubMed.
  2. . Does physical activity prevent cognitive decline and dementia?: A systematic review and meta-analysis of longitudinal studies. BMC Public Health. 2014 May 27;14:510. PubMed.
  3. . Dementia prevention, intervention, and care. Lancet. 2017 Jul 19; PubMed.
  4. . Maximal work performance test in middle-aged women: results from a population study. Scand J Clin Lab Invest. 1978 Apr;38(2):181-8. PubMed.
  5. . Aerobic capacity reference data in 3816 healthy men and women 20-90 years. PLoS One. 2013;8(5):e64319. Print 2013 PubMed.
  6. . Physical activity and risk of neurodegenerative disease: a systematic review of prospective evidence. Psychol Med. 2009 Jan;39(1):3-11. PubMed.
  7. . Physical activity and risk of cognitive decline: a meta-analysis of prospective studies. J Intern Med. 2011 Jan;269(1):107-17. Epub 2010 Sep 10 PubMed.
  8. . The association between midlife cardiorespiratory fitness levels and later-life dementia: a cohort study. Ann Intern Med. 2013 Feb 5;158(3):162-8. PubMed.
  9. . Cardiovascular and cognitive fitness at age 18 and risk of early-onset dementia. Brain. 2014 May;137(Pt 5):1514-23. Epub 2014 Mar 6 PubMed.
  10. . Association between mid- to late life physical fitness and dementia: evidence from the CAIDE study. J Intern Med. 2014 Jan 20; PubMed.
  11. . Age, vascular health, and Alzheimer disease biomarkers in an elderly sample. Ann Neurol. 2017 Nov;82(5):706-718. Epub 2017 Oct 26 PubMed.

Further Reading

Primary Papers

  1. . Midlife cardiovascular fitness and dementia: A 44-year longitudinal population study in women. Neurology. 2018 Apr 10;90(15):e1298-e1305. Epub 2018 Mar 14 PubMed.