Unfavourable gender effect of high body mass index on brain metabolism and connectivity

www.nature.com/scientificreports OPEN Received: 15 March 2018 Accepted: 18 July 2018 Published: xx xx xxxx Unfavourable gender effect of high body mass index on brain metabolism and connectivity Maura Malpetti1, Arianna Sala1,2, Emilia Giovanna Vanoli3, Luigi Gianolli3, Livio Luzi4 & Daniela Perani1,2,3 The influence of Body Mass Index (BMI) on neurodegeneration in dementia has yet to be elucidated. We aimed at exploring the effects of BMI levels on cerebral resting-state metabolism and brain connectivity, as crucial measures of synaptic function and activity, in a large group of patients with Alzheimer’s Dementia (AD) (n = 206), considering gender. We tested the correlation between BMI levels and brain metabolism, as assessed by 18F-FDG-PET, and the modulation of the resting-state functional networks by BMI. At comparable dementia severity, females with high BMI can withstand a lower degree of brain metabolism dysfunction, as shown by a significant BMI-brain metabolism correlation in the temporal-parietal regions, which are typically vulnerable to AD pathology (R = 0.269, p = 0.009). Of note, high BMI was also associated with reduced connectivity in frontal and limbic brain networks, again only in AD females (p < 0.05 FDR-corrected, k = 100 voxels). This suggests a major vulnerability of neural systems known to be selectively involved in brain compensatory mechanisms in AD females. These findings indicate a strong gender effect of high BMI and obesity in AD, namely reducing the available reserve mechanisms in female patients. This brings to considerations for medical practice and health policy. The interaction between Body Mass Index (BMI) and cognition in healthy status, and the role of BMI as a risk factor for dementia are complex and highly debated issues1–3. There is growing evidence that obesity (BMI > 30 kg/m2) is linked to reduced cognitive functions in adults, particularly in executive, attention, and memory domains3–5. Epidemiological evidence on the risk of developing dementia in obesity is controversial. The complex interplay between BMI and the risk of developing cognitive decline might be modulated by age. Some studies reported a significant association between high BMI and cognitive deficits in young subjects6, and also a higher risk of developing dementia in midlife7,8. Despite the results in young adults and middle-aged subjects consistently suggest a negative effect of high BMI on cognitive functions, the evidence in late life is less clear. Epidemiological findings in elderly suggest a protective effect of high BMI against dementia (see for reviews1–3). This evidence merges in the “obesity paradox” hypothesis9, which again suggests a positive correlation between BMI levels and dementia risk in mid-life (age up to 60–65 years), but a protective effect of high BMI in late life (age 65–70 years or older) instead. To explain this discrepancy in late life, it has been suggested that low BMI may be a consequence of the loss of weight preceding the onset of dementia in older people10,11. An accelerated reduction in BMI during late life has been also reported to precede the clinical detection of dementia, while a slower decline in BMI has been associated with a reduced dementia risk1. Therefore, it is possible that low or declining BMI is a factor associated with the development of dementia rather than a risk factor. On the other hand, a lifelong high BMI has been equally suggested as a risk factor for cognitive decline and dementia by several studies, which reported an increased risk of developing cognitive decline in overweight or obese individuals across the lifespans3,6,12. In spite of the extensive epidemiological data, neuroimaging studies are scarce and the neural mechanisms underlying the effect of BMI in dementia remain currently unknown. 18F-fluorodeoxyglucose Positron Emission Tomography (18F-FDG-PET) is a well-recognized tool to study brain dysfunction in dementia13,14. 18F-FDG-PET measures of resting-state brain metabolism, which is a direct in vivo index of synaptic function13, were applied in 1 Division of Neuroscience, San Raffaele Scientific Institute, Milan, Italy. 2Vita-Salute San Raffaele University, Milan, Italy. 3Nuclear Medicine Unit, San Raffaele Hospital, Milan, Italy. 4Metabolism Research Center and Endocrinology Unit, IRCCS Policlinico San Donato Milanese, Milan, Italy. Maura Malpetti and Arianna Sala contributed equally. Correspondence and requests for materials should be addressed to D.P. (email: ) Scientific REPOrtS | (2018) 8:12584 | DOI:10.1038/s41598-018-30883-y 1 www.nature.com/scientificreports/ studies on modifiable risk and protective factors for dementia, although not on BMI. For example, there is large evidence for low education and occupation levels, low physical and leisure activities as risk factors for all dementia conditions15,16. On the other hand, life experiences such as high education17–19, occupation17,18 and bilingualism20 are known to be involved in the modulation of cognitive reserve inducing brain plasticity and structural changes in both healthy subjects and patients with neurodegenerative conditions16. These proxies of cognitive reserve seem to be crucial for processes related to the brain maintenance, leading to protection of brain integrity in aging and disease, and resulting in a higher brain reserve16. Several neuroimaging studies17–20 focused on brain reserve16, which is now recognized as one of the most powerful protective mechanism for cognitive decline and dementia15. These studies have suggested that high levels of reserve allowing cope with brain neurodegeneration, delaying the onset of clinical manifestation and leading to a higher tolerance to underlying pathology17–20. We hypothesized possible modulatory effects of BMI on brain neurodegeneration and functional connectivity, and specifically a detrimental effect of high BMI and obesity on brain reserve, leading to major vulnerability to neuropathology. We thus studied a large group of patients with probable Alzheimer’s Dementia (AD), who underwent 18F-FDG-PET, in order to assess the effects of BMI levels on brain metabolism and connectivity, crucial measures of synaptic function and activity, while also considering the role of gender. Results Group analyses. We first estimated AD-related hypometabolism by comparing AD patients with a large group of healthy controls. In the patients, the comparison showed the typical AD-like pattern (p < 0.05 FWE corrected for multiple comparison, k:100 voxels), with an extended and significant hypometabolism in the parietal-temporal cortex, and in the posterior cingulate cortex and precuneus (Fig. 1a). We then tested for differences in brain hypometabolism in AD patients with low BMI vs. high BMI, separately for female and male subgroups (p < 0.05 FWE corrected for multiple comparison, cluster extent (k) threshold at 100 voxels). The results showed a less severe and extended pattern of hypometaboli (...truncated)