European Journal of Clinical Nutrition (2014) 68, 1216–1219 © 2014 Macmillan Publishers Limited All rights reserved 0954-3007/14 www.nature.com/ejcn
REVIEW
Body mass index and mild cognitive impairment-to-dementia progression in 24 months: a prospective study T Sobów1, W Fendler2 and R Magierski3 BACKGROUND/OBJECTIVES: Mild cognitive impairment (MCI), often considered as an early stage of dementia, is heterogeneous, and not all subjects with MCI progress into clinically diagnosed dementia. Low body weight (and body mass index, BMI) as well as losing weight while in MCI stadium have been proposed as possible risk factors of MCI-to-dementia conversion. SUBJECTS/METHODS: A prospective, 2-year observation of 102 MCI subjects has been conducted. Data on MCI subtype, somatic and neuropsychiatric co-morbidity and demographic characteristics (including age, gender and education), were collected. In addition, baseline and yearly BMI were calculated. RESULTS: Data of 83 out of the originally included 102 subjects were available after 2 years; 27 of those (32.5%) progressed to dementia. In univariate analysis, multiple-deficit MCI subtype (as compared with pure amnestic), higher age, the presence of diabetes and apathy, and lower baseline BMI (and losing weight on 2-year follow-up) were associated with conversion to dementia. Variables retained in the multivariate backward stepwise logistic regression model for conversion after 24 months of observation included lower baseline BMI (odds ratio, OR (95% cofidence interval, CI): 0.6 (0.4–0.9)), weight loss on 2-year follow-up (OR (95% CI): 1.3 (1.1–1.5)), male gender (OR (95% CI): 0.1 (0.01–0.9)) and presence of apathy (OR (95% CI): 70.7 (5.6–699)). Apathetic subjects had lower BMI and higher weight loss after controlling for potential confounders (age, gender, years of education and baseline ADAScog (Alzheimer's Disease Assessment Scale-cognitive subscale) score). CONCLUSION: MCI subjects presenting with apathy, low initial BMI and losing weight on follow-up have a significantly greater risk of developing dementia. Nutritional and behavioural assessment should be considered as additional tools in evaluating the risk of dementia among MCI subjects. European Journal of Clinical Nutrition (2014) 68, 1216–1219; doi:10.1038/ejcn.2014.167; published online 13 August 2014
INTRODUCTION The relationship between body mass index (BMI), used as a proxy measure of nutritional status, and the risk of cognitive decline and dementia appears to be complex. Both high and low BMI have been associated with either the risk of dementia in the future1–3 or incipient dementia.4,5 Moreover, a decrease in BMI has been proposed as a marker of cognitive decline or incipient dementia.6–10 Recently, it has been also shown that lower baseline BMI is associated with a faster decline in mild cognitive impairment (MCI), a clinical construct conceptualized commonly as a preclinical stage of dementia.11,12 There are several possible explanations for the observed associations ranging directly from biological correlates of nutritional status and Alzheimer-like pathology6,13 to more psychosocial hypotheses for weight loss, including behavioural changes that influence both caloric intake and food quality such as forgetting to eat, inability to plan and prepare adequate meals and a tendency to exclude various foods, mostly vegetables and fruits, from everyday diet. One possible, yet understudied, explanation of the BMI to the rate of cognitive decline association could also be behavioural and include the presence of neuropsychiatric symptoms in both MCI and incipient dementia. Symptoms such as aberrant motor behaviour, agitation, irritability (possibly resulting in higher level of activities), as well as depression and apathy (possibly related to insufficient or inadequate eating), have been shown to be
correlated with weight loss in Alzheimer’s disease (AD) patients.14,15 Although some behavioural symptoms are common among MCI subjects,16 and even a clinical construct of mild behavioural impairment has been proposed recently,17 the data showing association between their presence and weight loss and also cognitive decline are lacking. In the present study, we try to address these gaps in knowledge by showing data from a well-characterized cohort of MCI subjects observed longitudinally at the academic, tertiary centre within a 24-month time frame. MATERIALS AND METHODS From a total number of 102 subjects (identified through screening of all potentially eligible patients diagnosed at our centre) with MCI, fulfilling core clinical criteria proposed by the National Institute of Aging—Alzheimer’s Association workgroup18 and a score of 0.5, according to the Washington University Clinical Dementia Rating Scale (CDR)—assignment algorithm,19 data for 83 who agreed to participate (25 men, mean age of 75 years) were available for multivariate analysis. The Washington University CDR is a 5-point global assessment instrument developed to clinically denote the presence of dementia of AD-type and stage its severity.20 The tool is used to characterize six domains of cognitive and functional
1 Department of Medical Psychology, Medical University of Lodz, Lodz, Poland; 2Department of Pediatrics, Oncology, Hematology and Diabetology, Medical University of Lodz, Lodz, Poland and 3Department of Old Age Psychiatry and Psychotic Disorders, Medical University of Lodz, Lodz, Poland. Correspondence: Professor T Sobów, Central Teaching Hospital, Medical University of Lodz, Czechoslowacka 8/10, 92-216 Lodz, Poland. E-mail: [email protected] Received 10 July 2014; accepted 12 July 2014; published online 13 August 2014
BMI as predictor of MCI-to-dementia progression T Sobów et al
1217 Table 1. Clinical and demographic baseline characteristics of the cohort included in the study and comparisons between converting and nonconverting groups Variable Age Gender (fraction of males) Year of education (median) MCI subtype (amnestic/multiple domain) BMI baseline BMI change (Δ-BMI) % at 24 months ADAS-cog NPI total number of symptoms (median) Apathy Depression Diabetes Hypertension Ischaemic heart disease Hypercholesterolaemia Atrial fibrillation
Entire study cohort (N = 83)
Converters (N = 27)
Non-converters (N = 56)
P-values (converters vs nonconverters)
75.0 ± 1.9 25/58 9 39/44 21.4 ± 1.4 − 2.8 (−4.0/2.9) 12.1 ± 1.8 4 20/63 12/71 14/69 16/67 23/60 17/66 8/75
73.3 ± 0.4 5/27 8 6/21 19.6 ± 1.3 − 6.5 (−9.5/1.9) 13.1 ± 1.7 5 18/9 7/20 12/15 9/18 4/23 7/20 3/24
78.3 ± 2.5 20/56 9 33/23 22.5 ± 1.3 − 1.3 (−4.7/3.5) 11.5 ±2.1 4 2/54 5/51 2/54 7/49 19/37 10/46 5/51
o 0.001 0.13 NS 0.002 o 0.001 o 0.001 o 0.001 NS 0.0001 0.05 o0.0001 0.04 0.1 0.6 0.7
Abbreviations: ADAS-cog, Alzheimer's Disease Assessment Scale-cognitive subscale; BMI, body mass index; MCI, mild cognitive impairment; NPI, Neuropsychiatric Inventory.
performance: memory, orientation, judgement and problemsolving, community affairs, home and hobbies and personal care. The necessary information to make each rating is collected through semi-structured interviews with the patient and informant, which is needed to rate the subject’s condition. The global score and sum of boxes (SOB) scores were used in clinical and research settings to stage dementia severity. CDR-0 means no cognitive impairment, and then the remaining four points are tracking a patient's level of impairment/dementia: CDR-0.5 = very mild dementia (mild cognitive impairment), CDR-1 = mild dementia, CDR-2 = moderate dementia and CDR-3 = severe dementia. With regard to cognitive measures, a battery of tests has been applied, including ADAS-cog (Alzheimer's Disease Assessment Scale—cognitive subscale)21 and Mini–Mental State Examination (as screening),22 The Trail Making Test (for executive functions),23 the Boston Naming Test,24 letter and category fluency (for language),25 copying of the complex Rey’s figure (for spatial skills)26 and digit span forward and back from Wechsler Adult Intelligence Scale (for attention).27 In addition, at baseline, all subjects had either computed tomography or magnetic resonance imaging scan and a set of standard laboratory tests to exclude cognitive impairment due to a known underlying aetiology. All subjects were evaluated at baseline, year 1 and year 2 with the same set of standard procedures, including, beyond cognitive tests (according to those cases were later classified as having either amnestic or multiple-domain MCI types), neuropsychiatric symptoms (using neuropsychiatric inventory), which were scored binary (as either present of absent only),28 co-morbid disorders (including diabetes, hypertension, ischaemic heart disease, atrial fibrillation and hypercholesterolaemia), as well as body weight and height, used for BMI calculations. The body weight (in kg to the nearest 0.1 kg) was measured on an electronic scale while participants were wearing underwear only and no shoes. The body height (in cm to the nearest 0.1 cm) was measured with a portable wall-mounting height scale with a measuring slide and a heel plate. The end point was an MCI-to-dementia ‘conversion’, defined in this study as a change in CDR score from 0.5 to 1 (or more). Statistical analysis For continuous variable comparisons, the Mann–Whitney U-test or Student’s t-test was used depending on the distribution of data. © 2014 Macmillan Publishers Limited
The Yates-corrected χ2 or Fisher’s two-sided exact tests were used for comparisons of nominal variables depending on the size of the compared groups. Backward stepwise logistic regression was used for multivariate analysis. After conducting univariate comparisons, factors for which P was o0.15 were entered into multivariate models. A threshold of P o 0.05 in the final multivariate model was considered as statistically significant. To ensure no significant colinearity of factors included in the final model, the variation inflation factors were calculated. All statistical analyses were performed using the STATISTICA 10.0 package (StatSoft, Tulsa, OK, USA). RESULTS From the initially identified 102 potentially appropriate subjects, data for 83 were available for analyses. The remaining (N = 19) either did not consent for data use (N = 6) or were lost to follow-up (N = 13). During 2 years of observation, 37 out of the 83 subjects progressed to dementia, as defined by a CDR score change from 0.5 to 1 (‘converters’). A baseline characteristic of the entire population and comparison between converters and nonconverters is provided in Table 1. As expected, converters differed from those who remained stable in a number of baseline characteristics, including higher age, lower score on ADAS-cog screening scale and higher fraction of multi-domain MCI subtype. Diabetes and hypertension were more prevalent among converters, whereas ischaemic heart disease, rather surprisingly, was prevalent in the stable group. Neuropsychiatric symptoms were quite common among both groups; only apathy and depression stood out as being different between the two groups. Converters proved to have significantly lower baseline BMI and bigger weight loss during 2 years of observation (Table 1). In multivariate analysis, baseline BMI, Δ-BMI, male gender and the presence of apathy were retained in the final model (Table 2). Of note, confidence intervals, particularly for the presence of apathy and gender distribution, are quite large, reducing their utility strongly as conclusive effects. Baseline BMI did not correlate with Δ-BMI (at 24 months), which allowed retaining both factors in the final model (R = 0.02; P = 0.8). However, both BMI and Δ-BMI (at 24 months) were correlated with the presence of apathy. Patients with apathy had lower baseline BMI (20.1 ± 2.3 versus European Journal of Clinical Nutrition (2014) 1216 – 1219
BMI as predictor of MCI-to-dementia progression T Sobów et al
1218 Table 2.
Variables retained in the multivariate backward stepwise logistic regression model for conversion after 24 months Variable Male gender Baseline BMI Δ-BMI at 24 months Apathy
OR (95% CI) 0.1 0.6 1.3 70.7
(0.01–0.9) (0.4–0.9) (1.1–1.5) (5.6–699)
P-value
VIF
0.04 o0.01 0.008 o0.001
1.0 1.2 1.1 1.3
Abbreviations: BMI, body mass index; Δ-BMI, body mass index change; CI, confidence level; VIF, variance inflation factors.
22.5 ± 2.1 kg/m; P o0.001) and higher Δ-BMI (5.0 ± 5.7 versus 1.6 ± 5.0 kg/m; P = 0.01), and the differences remained statistically significant after controlling for potential confounding factors, such as age, gender, years of education and baseline ADAScog score. To ensure that despite these correlations all factors should be retained in the final model, variance inflation factors have been calculated and none of them exceeded the potentially harmful value of 2 (Table 2). Given all the above, we came up with the final predictive model: Probability of outcome
¼ 1=1þe - ð10:26þApathy 0=1 ´ 4:26þgender 0=1 ´ ð - 2:71ÞþBMI baseline ´ ð - 0:57ÞþΔBMI ´ 0:23Þ which proved to efficiently classify at least 85% cases. It must be, however, noted that the model should be next applied to independent and bigger cohorts to ensure its validity and predictive efficiency. DISCUSSION Weight loss is a well-known phenomenon in the course of dementia, particularly in AD, with various possible explanations proposed. These include lower energy intake (in part due to olfactory system failure and lower leptin levels), higher resting energy expenditure as well as increased physical activity (due to behavioural disturbances), and may lead to micronutrient deficiencies, consequent oxidative tissue damage and acceleration of neurodegenerative processes.29 In longitudinal observations of healthy elderly, it has been noticed that weight loss may occur as early as 4–0 years before the onset of clinically overt dementia.30,31 These observations lead to a dispute whether weight loss is a risk factor or a very early sign of ongoing neurodegenerative pathology.3,5,32 Limited clinical evidence also indicates that low initial BMI as well as aggravated weight loss might increase the rate of cognitive decline in preclinical dementia stages.11,12 The results of the present study seem to confirm these findings. Interestingly, we have also found that baseline BMI is not correlated with an ongoing weight loss, which may indicate that there are different pathogenetic processes involved. We hypothesize that low BMI might be related more to subtle changes in appetite regulation that may precede overt dementia,33 whereas further weight loss is at least in part associated with behavioural changes that tend to aggravate in the course of cognitive decline and dementia.34 Moreover, we observed a strong association between apathy, one of the most common behavioural symptom of MCI, and early dementia,35 with both low BMI and later weight loss apparent in patients progressing more rapidly from MCI to an overt dementia. This is in line with earlier observations of our group and others that apathy might be associated with faster cognitive decline.36–38 On the other hand, the link between apathy and weight loss might be in shared neural substrates, as it has been shown that subjects with apathy and MCI or dementia present more olfactory deficits as compared with those without this behavioural complex.39 European Journal of Clinical Nutrition (2014) 1216 – 1219
The present study is not free of limitations. Although well characterized clinically and psychometrically, our group is relatively small and statistical analyses are rather strongly affected by the number of subjects analysed (for example, confidence intervals in logistic regression are wide, particularly for apathy, and de facto exclude firm conclusions). We also did not account for a type of incident dementia, although is it probably useful to mention that in most cases patients were classified as having AD. Finally, the present analysis did not take into consideration the genetic variability of APOE (or other genes such as CLU (clusterin) gene) that may possibly change the picture owing to known, complex interactions with MCI and the natural course of dementia as well as the contribution of dietary factors.40 Last but not the least, BMI is a very rough indicator of nutritional status of the body, and analyses of body composition (using, for instance, bioimpedance methods, dual energy X-ray absorptiometry or magnetic resonance techniques) could be applied in future studies to better understand the relationship between body weight and cognitive decline. This might particularly involve visceral fat that appears to confer an increased risk for cognitive impairment or dementia41 or body hydration status. In conclusion, taking the results of our study into account, we advocate for inclusion of body weight (and BMI) monitoring to the standard care of subjects with MCI and dementia. Further studies are needed to clarify whether low BMI and weight loss are just markers of faster cognitive decline or a target for a possible nutritional intervention. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS The authors are (partially) supported by Healthy Ageing Research Centre (REGPOT-2012-2013-1, 7FP). The work was supported by own resources of Medical University of Lodz (503/8-040-01/503-01).
REFERENCES 1 Whitmer RA, Gunderson EP, Quesenberry Jr CP, Zhou J, Yaffe K. Body mass index in midlife and risk of Alzheimer disease and vascular dementia. Curr Alzheimer Res 2007; 4: 103–109. 2 Fitzpatrick AL, Kuller LH, Lopez OL, Diehr P, O'Meara ES, Longstreth WT Jr et al. Midlife and late-life obesity and the risk of dementia: cardiovascular health study. Arch Neurol 2009; 66: 336–342. 3 Anstey KJ, Cherbuin N, Budge M, Young J. Body mass index in midlife and late-life as a risk factor for dementia: a meta-analysis of prospective studies. Obes Rev 2011; 12: e426–e437. 4 Johnson DK, Wilkins CH, Morris JC. Accelerated weight loss may precede diagnosis in Alzheimer disease. Arch Neurol 2006; 63: 1312–1317. 5 Gustafson DR, Bäckman K, Joas E, Waern M, Östling S, Guo X et al. 37 years of body mass index and dementia: observations from the prospective population study of women in Gothenburg, Sweden. J Alzheimers Dis 2012; 28: 163–171. 6 Buchman AS, Wilson RS, Bienias JL, Shah RC, Evans DA, Bennett DA. Change in body mass index and risk of incident Alzheimer disease. Neurology 2005; 65: 892–897. 7 Atti AR, Palmer K, Volpato S, Winblad B, De Ronchi D, Fratiglioni L. Late-life body mass index and dementia incidence: nine-year follow-up data from the Kungsholmen Project. J Am Geriatr Soc 2008; 56: 111–116. 8 Hughes TF, Borenstein AR, Schofield E, Wu Y, Larson EB. Association between latelife body mass index and dementia: The Kame Project. Neurology 2009; 72: 1741–1746. 9 Power BD, Alfonso H, Flicker L, Hankey GJ, Yeap BB, Almeida OP. Changes in body mass in later life and incident dementia. Int Psychogeriatr 2013; 25: 467–478. 10 Tolppanen AM, Ngandu T, Kåreholt I, Laatikainen T, Rusanen M, Soininen H et al. Midlife and late-life body mass index and late-life dementia: results from a prospective population-based cohort. J Alzheimers Dis 2014; 38: 201–209. 11 Cronk BB, Johnson DK, Burns JM. Alzheimer's Disease Neuroimaging Initiative. Body mass index and cognitive decline in mild cognitive impairment. Alzheimer Dis Assoc Disord 2010; 24: 126–130.
© 2014 Macmillan Publishers Limited
BMI as predictor of MCI-to-dementia progression T Sobów et al
1219 12 Besser LM, Gill DP, Monsell SE, Brenowitz W, Meranus DH, Kukull W et al. Body mass index, weight change, and clinical progression in mild cognitive impairment and Alzheimer disease. Alzheimer Dis Assoc Disord 2014; 28: 36–43. 13 Grundman M, Corey-Bloom J, Jernigan T, Archibald S, Thal LJ. Low body weight in Alzheimer's disease is associated with mesial temporal cortex atrophy. Neurology 1996; 46: 1585–1591. 14 White H, Pieper C, Schmader K, Fillenbaum G. Weight change in Alzheimer’s disease. J Am Geriatr Soc 1996; 44: 265–272. 15 Volicer L, Frijters DH, van der Steen JT. Apathy and weight loss in nursing home residents: longitudinal study. J Am Med Dir Assoc 2013; 14: 417–420. 16 Monastero R, Mangialasche F, Camarda C, Ercolani S, Camarda R. A systematic review of neuropsychiatric symptoms in mild cognitive impairment. J Alzheimers Dis 2009; 18: 11–30. 17 Dillon C, Serrano CM, Castro D, Leguizamon PP, Heisecke SL, Taragano FE. Behavioral symptoms related to cognitive impairment. Neuropsychiatr Dis Treat 2013; 9: 1443–1455. 18 Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman HH, Fox NC et al. The diagnosis of mild cognitive impairment due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement 2011; 7: 270–279. 19 Morris JC. The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology 1993; 43: 2412–2414. 20 Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL. A new clinical scale for the staging of dementia. Br J Psychiatry 1982; 140: 566–572. 21 Rosen WG, Mohs RC, Davis KL. A new rating scale for Alzheimer's disease. Am J Psychiatry 1984; 141: 1356–1364. 22 Folstein MF, Folstein SE, McHugh PR. Mini-mental-state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–198. 23 Reitan RM. The relation of the trail making test to organic brain damage. J Consult Psychol 1955; 19: 393–394. 24 Kaplan EF, Goodglass H, Weintraub S. The Boston Naming Test: Experimental edition (1978). Boston, USA: Kaplan & Goodglass, 1978, 1983. 25 Lezak MD, Howieson DB, Loring DW. Neuropsychological Assessment. 4th edn. Oxford University Press, USA: New York, 2004. 26 Corwin J, Bylsma FW. Translations of excerpts from the Andre Rey's Psychological examination of traumatic encephalopathy and P.A. Osterreith's The complex figure copy test. Clin Neuropsychol 1993; 7: 3–15. 27 Wechsler D. Wechsler Adult Intelligence Scale—Revised. The Psychological Corporation, USA: New York, 1981.
© 2014 Macmillan Publishers Limited
28 Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi DA, Gornbein J. The Neuropsychiatric Inventory: comprehensive assessment of psychopathology in dementia. Neurology 1994; 44: 2308–2314. 29 Sergi G, De Rui M, Coin A, Inelmen EM, Manzato E. Weight loss and Alzheimer's disease: temporal and aetiologic connections. Proc Nutr Soc 2013; 72: 160–165. 30 Barrett-Connor E, Edelstein SL, Corey-Bloom J, Wiederholt WC. Weight loss precedes dementia in community-dwelling older adults. J Am Geriatr Soc 1996; 44: 1147–1152. 31 Knopman DS, Edland SD, Cha RH, Petersen RC, Rocca WA. Incident dementia in women is preceded by weight loss by at least a decade. Neurology 2007; 69: 739–746. 32 Nourhashémi F, Deschamps V, Larrieu S, Letenneur L, Dartigues JF, BarbergerGateau P. PAQUID study. Personnes Agées Quid. Body mass index and incidence of dementia: the PAQUID study. Neurology 2003; 60: 117–119. 33 Fadel JR, Jolivalt CG, Reagan LP. Food for thought: the role of appetitive peptides in age-related cognitive decline. Ageing Res Rev 2013; 12: 764–776. 34 Guerin O, Soto ME, Brocker P, Robert PH, Benoit M, Vellas B. REAL.FR Group. Nutritional status assessment during Alzheimer's disease: results after one year (the REAL French Study Group). J Nutr Health Aging 2005; 9: 81–84. 35 Benoit M, Berrut G, Douissant J, Bakchine S, Bonin-Guillaume S, Fremont P et al. Apathy and depression in mild Alzheimer's disease: a cross-sectional study using diagnostic criteria. J Alzheimers Dis 2012; 31: 325–334. 36 Sobow T, Magierski R, Flirski M, Wojtera M, Kloszewska I. Apathy but not other affective symptoms is predicting conversion to dementia in mild cognitive impairment. Alzheimers Dementia 2010; 6: S355. 37 Richard E, Schmand B, Eikelenboom P, Yang SC, Lightart SA, Moll van Charante EP et al. Symptoms of apathy are associated with progression from mild cognitive impairment to Alzheimer's disease in non-depressed subjects. Dement Geriatr Cogn Disord 2012; 33c: 204–209. 38 Somme J, Fernandez-Martinez M, Molano A, Zarranz JJ. Neuropsychiatric symptoms in amnestic mild cognitive impairment: increased risk and faster progression to dementia. Curr Alzheimer Res 2013; 10: 86–94. 39 Seligman SC, Kamath V, Giovanetti T, Arnold SE, Moberg PJ. Olfaction and apathy in Alzheimer's disease, mild cognitive impairment, and healthy older adults. Aging Mental Health 2013; 17: 564–570. 40 Barberger-Gateau P, Lambert JC, Feart C, Peres K, Ritchie K, Dartigues JF et al. From genetics to dietetics: the contribution of epidemiology to understanding Alzheimer's disease. J Alzheimers Dis 2013; 33: S457–S463. 41 West NA, Haan MN. Body adiposity in late life and risk of dementia or cognitive impairment in a longitudinal community-based study. J Gerontol A Biol Sci Med Sci 2009; 64: 10 3–109.
European Journal of Clinical Nutrition (2014) 1216 – 1219
REVIEW
Body mass index and mild cognitive impairment-to-dementia progression in 24 months: a prospective study T Sobów1, W Fendler2 and R Magierski3 BACKGROUND/OBJECTIVES: Mild cognitive impairment (MCI), often considered as an early stage of dementia, is heterogeneous, and not all subjects with MCI progress into clinically diagnosed dementia. Low body weight (and body mass index, BMI) as well as losing weight while in MCI stadium have been proposed as possible risk factors of MCI-to-dementia conversion. SUBJECTS/METHODS: A prospective, 2-year observation of 102 MCI subjects has been conducted. Data on MCI subtype, somatic and neuropsychiatric co-morbidity and demographic characteristics (including age, gender and education), were collected. In addition, baseline and yearly BMI were calculated. RESULTS: Data of 83 out of the originally included 102 subjects were available after 2 years; 27 of those (32.5%) progressed to dementia. In univariate analysis, multiple-deficit MCI subtype (as compared with pure amnestic), higher age, the presence of diabetes and apathy, and lower baseline BMI (and losing weight on 2-year follow-up) were associated with conversion to dementia. Variables retained in the multivariate backward stepwise logistic regression model for conversion after 24 months of observation included lower baseline BMI (odds ratio, OR (95% cofidence interval, CI): 0.6 (0.4–0.9)), weight loss on 2-year follow-up (OR (95% CI): 1.3 (1.1–1.5)), male gender (OR (95% CI): 0.1 (0.01–0.9)) and presence of apathy (OR (95% CI): 70.7 (5.6–699)). Apathetic subjects had lower BMI and higher weight loss after controlling for potential confounders (age, gender, years of education and baseline ADAScog (Alzheimer's Disease Assessment Scale-cognitive subscale) score). CONCLUSION: MCI subjects presenting with apathy, low initial BMI and losing weight on follow-up have a significantly greater risk of developing dementia. Nutritional and behavioural assessment should be considered as additional tools in evaluating the risk of dementia among MCI subjects. European Journal of Clinical Nutrition (2014) 68, 1216–1219; doi:10.1038/ejcn.2014.167; published online 13 August 2014
INTRODUCTION The relationship between body mass index (BMI), used as a proxy measure of nutritional status, and the risk of cognitive decline and dementia appears to be complex. Both high and low BMI have been associated with either the risk of dementia in the future1–3 or incipient dementia.4,5 Moreover, a decrease in BMI has been proposed as a marker of cognitive decline or incipient dementia.6–10 Recently, it has been also shown that lower baseline BMI is associated with a faster decline in mild cognitive impairment (MCI), a clinical construct conceptualized commonly as a preclinical stage of dementia.11,12 There are several possible explanations for the observed associations ranging directly from biological correlates of nutritional status and Alzheimer-like pathology6,13 to more psychosocial hypotheses for weight loss, including behavioural changes that influence both caloric intake and food quality such as forgetting to eat, inability to plan and prepare adequate meals and a tendency to exclude various foods, mostly vegetables and fruits, from everyday diet. One possible, yet understudied, explanation of the BMI to the rate of cognitive decline association could also be behavioural and include the presence of neuropsychiatric symptoms in both MCI and incipient dementia. Symptoms such as aberrant motor behaviour, agitation, irritability (possibly resulting in higher level of activities), as well as depression and apathy (possibly related to insufficient or inadequate eating), have been shown to be
correlated with weight loss in Alzheimer’s disease (AD) patients.14,15 Although some behavioural symptoms are common among MCI subjects,16 and even a clinical construct of mild behavioural impairment has been proposed recently,17 the data showing association between their presence and weight loss and also cognitive decline are lacking. In the present study, we try to address these gaps in knowledge by showing data from a well-characterized cohort of MCI subjects observed longitudinally at the academic, tertiary centre within a 24-month time frame. MATERIALS AND METHODS From a total number of 102 subjects (identified through screening of all potentially eligible patients diagnosed at our centre) with MCI, fulfilling core clinical criteria proposed by the National Institute of Aging—Alzheimer’s Association workgroup18 and a score of 0.5, according to the Washington University Clinical Dementia Rating Scale (CDR)—assignment algorithm,19 data for 83 who agreed to participate (25 men, mean age of 75 years) were available for multivariate analysis. The Washington University CDR is a 5-point global assessment instrument developed to clinically denote the presence of dementia of AD-type and stage its severity.20 The tool is used to characterize six domains of cognitive and functional
1 Department of Medical Psychology, Medical University of Lodz, Lodz, Poland; 2Department of Pediatrics, Oncology, Hematology and Diabetology, Medical University of Lodz, Lodz, Poland and 3Department of Old Age Psychiatry and Psychotic Disorders, Medical University of Lodz, Lodz, Poland. Correspondence: Professor T Sobów, Central Teaching Hospital, Medical University of Lodz, Czechoslowacka 8/10, 92-216 Lodz, Poland. E-mail: [email protected] Received 10 July 2014; accepted 12 July 2014; published online 13 August 2014
BMI as predictor of MCI-to-dementia progression T Sobów et al
1217 Table 1. Clinical and demographic baseline characteristics of the cohort included in the study and comparisons between converting and nonconverting groups Variable Age Gender (fraction of males) Year of education (median) MCI subtype (amnestic/multiple domain) BMI baseline BMI change (Δ-BMI) % at 24 months ADAS-cog NPI total number of symptoms (median) Apathy Depression Diabetes Hypertension Ischaemic heart disease Hypercholesterolaemia Atrial fibrillation
Entire study cohort (N = 83)
Converters (N = 27)
Non-converters (N = 56)
P-values (converters vs nonconverters)
75.0 ± 1.9 25/58 9 39/44 21.4 ± 1.4 − 2.8 (−4.0/2.9) 12.1 ± 1.8 4 20/63 12/71 14/69 16/67 23/60 17/66 8/75
73.3 ± 0.4 5/27 8 6/21 19.6 ± 1.3 − 6.5 (−9.5/1.9) 13.1 ± 1.7 5 18/9 7/20 12/15 9/18 4/23 7/20 3/24
78.3 ± 2.5 20/56 9 33/23 22.5 ± 1.3 − 1.3 (−4.7/3.5) 11.5 ±2.1 4 2/54 5/51 2/54 7/49 19/37 10/46 5/51
o 0.001 0.13 NS 0.002 o 0.001 o 0.001 o 0.001 NS 0.0001 0.05 o0.0001 0.04 0.1 0.6 0.7
Abbreviations: ADAS-cog, Alzheimer's Disease Assessment Scale-cognitive subscale; BMI, body mass index; MCI, mild cognitive impairment; NPI, Neuropsychiatric Inventory.
performance: memory, orientation, judgement and problemsolving, community affairs, home and hobbies and personal care. The necessary information to make each rating is collected through semi-structured interviews with the patient and informant, which is needed to rate the subject’s condition. The global score and sum of boxes (SOB) scores were used in clinical and research settings to stage dementia severity. CDR-0 means no cognitive impairment, and then the remaining four points are tracking a patient's level of impairment/dementia: CDR-0.5 = very mild dementia (mild cognitive impairment), CDR-1 = mild dementia, CDR-2 = moderate dementia and CDR-3 = severe dementia. With regard to cognitive measures, a battery of tests has been applied, including ADAS-cog (Alzheimer's Disease Assessment Scale—cognitive subscale)21 and Mini–Mental State Examination (as screening),22 The Trail Making Test (for executive functions),23 the Boston Naming Test,24 letter and category fluency (for language),25 copying of the complex Rey’s figure (for spatial skills)26 and digit span forward and back from Wechsler Adult Intelligence Scale (for attention).27 In addition, at baseline, all subjects had either computed tomography or magnetic resonance imaging scan and a set of standard laboratory tests to exclude cognitive impairment due to a known underlying aetiology. All subjects were evaluated at baseline, year 1 and year 2 with the same set of standard procedures, including, beyond cognitive tests (according to those cases were later classified as having either amnestic or multiple-domain MCI types), neuropsychiatric symptoms (using neuropsychiatric inventory), which were scored binary (as either present of absent only),28 co-morbid disorders (including diabetes, hypertension, ischaemic heart disease, atrial fibrillation and hypercholesterolaemia), as well as body weight and height, used for BMI calculations. The body weight (in kg to the nearest 0.1 kg) was measured on an electronic scale while participants were wearing underwear only and no shoes. The body height (in cm to the nearest 0.1 cm) was measured with a portable wall-mounting height scale with a measuring slide and a heel plate. The end point was an MCI-to-dementia ‘conversion’, defined in this study as a change in CDR score from 0.5 to 1 (or more). Statistical analysis For continuous variable comparisons, the Mann–Whitney U-test or Student’s t-test was used depending on the distribution of data. © 2014 Macmillan Publishers Limited
The Yates-corrected χ2 or Fisher’s two-sided exact tests were used for comparisons of nominal variables depending on the size of the compared groups. Backward stepwise logistic regression was used for multivariate analysis. After conducting univariate comparisons, factors for which P was o0.15 were entered into multivariate models. A threshold of P o 0.05 in the final multivariate model was considered as statistically significant. To ensure no significant colinearity of factors included in the final model, the variation inflation factors were calculated. All statistical analyses were performed using the STATISTICA 10.0 package (StatSoft, Tulsa, OK, USA). RESULTS From the initially identified 102 potentially appropriate subjects, data for 83 were available for analyses. The remaining (N = 19) either did not consent for data use (N = 6) or were lost to follow-up (N = 13). During 2 years of observation, 37 out of the 83 subjects progressed to dementia, as defined by a CDR score change from 0.5 to 1 (‘converters’). A baseline characteristic of the entire population and comparison between converters and nonconverters is provided in Table 1. As expected, converters differed from those who remained stable in a number of baseline characteristics, including higher age, lower score on ADAS-cog screening scale and higher fraction of multi-domain MCI subtype. Diabetes and hypertension were more prevalent among converters, whereas ischaemic heart disease, rather surprisingly, was prevalent in the stable group. Neuropsychiatric symptoms were quite common among both groups; only apathy and depression stood out as being different between the two groups. Converters proved to have significantly lower baseline BMI and bigger weight loss during 2 years of observation (Table 1). In multivariate analysis, baseline BMI, Δ-BMI, male gender and the presence of apathy were retained in the final model (Table 2). Of note, confidence intervals, particularly for the presence of apathy and gender distribution, are quite large, reducing their utility strongly as conclusive effects. Baseline BMI did not correlate with Δ-BMI (at 24 months), which allowed retaining both factors in the final model (R = 0.02; P = 0.8). However, both BMI and Δ-BMI (at 24 months) were correlated with the presence of apathy. Patients with apathy had lower baseline BMI (20.1 ± 2.3 versus European Journal of Clinical Nutrition (2014) 1216 – 1219
BMI as predictor of MCI-to-dementia progression T Sobów et al
1218 Table 2.
Variables retained in the multivariate backward stepwise logistic regression model for conversion after 24 months Variable Male gender Baseline BMI Δ-BMI at 24 months Apathy
OR (95% CI) 0.1 0.6 1.3 70.7
(0.01–0.9) (0.4–0.9) (1.1–1.5) (5.6–699)
P-value
VIF
0.04 o0.01 0.008 o0.001
1.0 1.2 1.1 1.3
Abbreviations: BMI, body mass index; Δ-BMI, body mass index change; CI, confidence level; VIF, variance inflation factors.
22.5 ± 2.1 kg/m; P o0.001) and higher Δ-BMI (5.0 ± 5.7 versus 1.6 ± 5.0 kg/m; P = 0.01), and the differences remained statistically significant after controlling for potential confounding factors, such as age, gender, years of education and baseline ADAScog score. To ensure that despite these correlations all factors should be retained in the final model, variance inflation factors have been calculated and none of them exceeded the potentially harmful value of 2 (Table 2). Given all the above, we came up with the final predictive model: Probability of outcome
¼ 1=1þe - ð10:26þApathy 0=1 ´ 4:26þgender 0=1 ´ ð - 2:71ÞþBMI baseline ´ ð - 0:57ÞþΔBMI ´ 0:23Þ which proved to efficiently classify at least 85% cases. It must be, however, noted that the model should be next applied to independent and bigger cohorts to ensure its validity and predictive efficiency. DISCUSSION Weight loss is a well-known phenomenon in the course of dementia, particularly in AD, with various possible explanations proposed. These include lower energy intake (in part due to olfactory system failure and lower leptin levels), higher resting energy expenditure as well as increased physical activity (due to behavioural disturbances), and may lead to micronutrient deficiencies, consequent oxidative tissue damage and acceleration of neurodegenerative processes.29 In longitudinal observations of healthy elderly, it has been noticed that weight loss may occur as early as 4–0 years before the onset of clinically overt dementia.30,31 These observations lead to a dispute whether weight loss is a risk factor or a very early sign of ongoing neurodegenerative pathology.3,5,32 Limited clinical evidence also indicates that low initial BMI as well as aggravated weight loss might increase the rate of cognitive decline in preclinical dementia stages.11,12 The results of the present study seem to confirm these findings. Interestingly, we have also found that baseline BMI is not correlated with an ongoing weight loss, which may indicate that there are different pathogenetic processes involved. We hypothesize that low BMI might be related more to subtle changes in appetite regulation that may precede overt dementia,33 whereas further weight loss is at least in part associated with behavioural changes that tend to aggravate in the course of cognitive decline and dementia.34 Moreover, we observed a strong association between apathy, one of the most common behavioural symptom of MCI, and early dementia,35 with both low BMI and later weight loss apparent in patients progressing more rapidly from MCI to an overt dementia. This is in line with earlier observations of our group and others that apathy might be associated with faster cognitive decline.36–38 On the other hand, the link between apathy and weight loss might be in shared neural substrates, as it has been shown that subjects with apathy and MCI or dementia present more olfactory deficits as compared with those without this behavioural complex.39 European Journal of Clinical Nutrition (2014) 1216 – 1219
The present study is not free of limitations. Although well characterized clinically and psychometrically, our group is relatively small and statistical analyses are rather strongly affected by the number of subjects analysed (for example, confidence intervals in logistic regression are wide, particularly for apathy, and de facto exclude firm conclusions). We also did not account for a type of incident dementia, although is it probably useful to mention that in most cases patients were classified as having AD. Finally, the present analysis did not take into consideration the genetic variability of APOE (or other genes such as CLU (clusterin) gene) that may possibly change the picture owing to known, complex interactions with MCI and the natural course of dementia as well as the contribution of dietary factors.40 Last but not the least, BMI is a very rough indicator of nutritional status of the body, and analyses of body composition (using, for instance, bioimpedance methods, dual energy X-ray absorptiometry or magnetic resonance techniques) could be applied in future studies to better understand the relationship between body weight and cognitive decline. This might particularly involve visceral fat that appears to confer an increased risk for cognitive impairment or dementia41 or body hydration status. In conclusion, taking the results of our study into account, we advocate for inclusion of body weight (and BMI) monitoring to the standard care of subjects with MCI and dementia. Further studies are needed to clarify whether low BMI and weight loss are just markers of faster cognitive decline or a target for a possible nutritional intervention. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS The authors are (partially) supported by Healthy Ageing Research Centre (REGPOT-2012-2013-1, 7FP). The work was supported by own resources of Medical University of Lodz (503/8-040-01/503-01).
REFERENCES 1 Whitmer RA, Gunderson EP, Quesenberry Jr CP, Zhou J, Yaffe K. Body mass index in midlife and risk of Alzheimer disease and vascular dementia. Curr Alzheimer Res 2007; 4: 103–109. 2 Fitzpatrick AL, Kuller LH, Lopez OL, Diehr P, O'Meara ES, Longstreth WT Jr et al. Midlife and late-life obesity and the risk of dementia: cardiovascular health study. Arch Neurol 2009; 66: 336–342. 3 Anstey KJ, Cherbuin N, Budge M, Young J. Body mass index in midlife and late-life as a risk factor for dementia: a meta-analysis of prospective studies. Obes Rev 2011; 12: e426–e437. 4 Johnson DK, Wilkins CH, Morris JC. Accelerated weight loss may precede diagnosis in Alzheimer disease. Arch Neurol 2006; 63: 1312–1317. 5 Gustafson DR, Bäckman K, Joas E, Waern M, Östling S, Guo X et al. 37 years of body mass index and dementia: observations from the prospective population study of women in Gothenburg, Sweden. J Alzheimers Dis 2012; 28: 163–171. 6 Buchman AS, Wilson RS, Bienias JL, Shah RC, Evans DA, Bennett DA. Change in body mass index and risk of incident Alzheimer disease. Neurology 2005; 65: 892–897. 7 Atti AR, Palmer K, Volpato S, Winblad B, De Ronchi D, Fratiglioni L. Late-life body mass index and dementia incidence: nine-year follow-up data from the Kungsholmen Project. J Am Geriatr Soc 2008; 56: 111–116. 8 Hughes TF, Borenstein AR, Schofield E, Wu Y, Larson EB. Association between latelife body mass index and dementia: The Kame Project. Neurology 2009; 72: 1741–1746. 9 Power BD, Alfonso H, Flicker L, Hankey GJ, Yeap BB, Almeida OP. Changes in body mass in later life and incident dementia. Int Psychogeriatr 2013; 25: 467–478. 10 Tolppanen AM, Ngandu T, Kåreholt I, Laatikainen T, Rusanen M, Soininen H et al. Midlife and late-life body mass index and late-life dementia: results from a prospective population-based cohort. J Alzheimers Dis 2014; 38: 201–209. 11 Cronk BB, Johnson DK, Burns JM. Alzheimer's Disease Neuroimaging Initiative. Body mass index and cognitive decline in mild cognitive impairment. Alzheimer Dis Assoc Disord 2010; 24: 126–130.
© 2014 Macmillan Publishers Limited
BMI as predictor of MCI-to-dementia progression T Sobów et al
1219 12 Besser LM, Gill DP, Monsell SE, Brenowitz W, Meranus DH, Kukull W et al. Body mass index, weight change, and clinical progression in mild cognitive impairment and Alzheimer disease. Alzheimer Dis Assoc Disord 2014; 28: 36–43. 13 Grundman M, Corey-Bloom J, Jernigan T, Archibald S, Thal LJ. Low body weight in Alzheimer's disease is associated with mesial temporal cortex atrophy. Neurology 1996; 46: 1585–1591. 14 White H, Pieper C, Schmader K, Fillenbaum G. Weight change in Alzheimer’s disease. J Am Geriatr Soc 1996; 44: 265–272. 15 Volicer L, Frijters DH, van der Steen JT. Apathy and weight loss in nursing home residents: longitudinal study. J Am Med Dir Assoc 2013; 14: 417–420. 16 Monastero R, Mangialasche F, Camarda C, Ercolani S, Camarda R. A systematic review of neuropsychiatric symptoms in mild cognitive impairment. J Alzheimers Dis 2009; 18: 11–30. 17 Dillon C, Serrano CM, Castro D, Leguizamon PP, Heisecke SL, Taragano FE. Behavioral symptoms related to cognitive impairment. Neuropsychiatr Dis Treat 2013; 9: 1443–1455. 18 Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman HH, Fox NC et al. The diagnosis of mild cognitive impairment due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement 2011; 7: 270–279. 19 Morris JC. The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology 1993; 43: 2412–2414. 20 Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL. A new clinical scale for the staging of dementia. Br J Psychiatry 1982; 140: 566–572. 21 Rosen WG, Mohs RC, Davis KL. A new rating scale for Alzheimer's disease. Am J Psychiatry 1984; 141: 1356–1364. 22 Folstein MF, Folstein SE, McHugh PR. Mini-mental-state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–198. 23 Reitan RM. The relation of the trail making test to organic brain damage. J Consult Psychol 1955; 19: 393–394. 24 Kaplan EF, Goodglass H, Weintraub S. The Boston Naming Test: Experimental edition (1978). Boston, USA: Kaplan & Goodglass, 1978, 1983. 25 Lezak MD, Howieson DB, Loring DW. Neuropsychological Assessment. 4th edn. Oxford University Press, USA: New York, 2004. 26 Corwin J, Bylsma FW. Translations of excerpts from the Andre Rey's Psychological examination of traumatic encephalopathy and P.A. Osterreith's The complex figure copy test. Clin Neuropsychol 1993; 7: 3–15. 27 Wechsler D. Wechsler Adult Intelligence Scale—Revised. The Psychological Corporation, USA: New York, 1981.
© 2014 Macmillan Publishers Limited
28 Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi DA, Gornbein J. The Neuropsychiatric Inventory: comprehensive assessment of psychopathology in dementia. Neurology 1994; 44: 2308–2314. 29 Sergi G, De Rui M, Coin A, Inelmen EM, Manzato E. Weight loss and Alzheimer's disease: temporal and aetiologic connections. Proc Nutr Soc 2013; 72: 160–165. 30 Barrett-Connor E, Edelstein SL, Corey-Bloom J, Wiederholt WC. Weight loss precedes dementia in community-dwelling older adults. J Am Geriatr Soc 1996; 44: 1147–1152. 31 Knopman DS, Edland SD, Cha RH, Petersen RC, Rocca WA. Incident dementia in women is preceded by weight loss by at least a decade. Neurology 2007; 69: 739–746. 32 Nourhashémi F, Deschamps V, Larrieu S, Letenneur L, Dartigues JF, BarbergerGateau P. PAQUID study. Personnes Agées Quid. Body mass index and incidence of dementia: the PAQUID study. Neurology 2003; 60: 117–119. 33 Fadel JR, Jolivalt CG, Reagan LP. Food for thought: the role of appetitive peptides in age-related cognitive decline. Ageing Res Rev 2013; 12: 764–776. 34 Guerin O, Soto ME, Brocker P, Robert PH, Benoit M, Vellas B. REAL.FR Group. Nutritional status assessment during Alzheimer's disease: results after one year (the REAL French Study Group). J Nutr Health Aging 2005; 9: 81–84. 35 Benoit M, Berrut G, Douissant J, Bakchine S, Bonin-Guillaume S, Fremont P et al. Apathy and depression in mild Alzheimer's disease: a cross-sectional study using diagnostic criteria. J Alzheimers Dis 2012; 31: 325–334. 36 Sobow T, Magierski R, Flirski M, Wojtera M, Kloszewska I. Apathy but not other affective symptoms is predicting conversion to dementia in mild cognitive impairment. Alzheimers Dementia 2010; 6: S355. 37 Richard E, Schmand B, Eikelenboom P, Yang SC, Lightart SA, Moll van Charante EP et al. Symptoms of apathy are associated with progression from mild cognitive impairment to Alzheimer's disease in non-depressed subjects. Dement Geriatr Cogn Disord 2012; 33c: 204–209. 38 Somme J, Fernandez-Martinez M, Molano A, Zarranz JJ. Neuropsychiatric symptoms in amnestic mild cognitive impairment: increased risk and faster progression to dementia. Curr Alzheimer Res 2013; 10: 86–94. 39 Seligman SC, Kamath V, Giovanetti T, Arnold SE, Moberg PJ. Olfaction and apathy in Alzheimer's disease, mild cognitive impairment, and healthy older adults. Aging Mental Health 2013; 17: 564–570. 40 Barberger-Gateau P, Lambert JC, Feart C, Peres K, Ritchie K, Dartigues JF et al. From genetics to dietetics: the contribution of epidemiology to understanding Alzheimer's disease. J Alzheimers Dis 2013; 33: S457–S463. 41 West NA, Haan MN. Body adiposity in late life and risk of dementia or cognitive impairment in a longitudinal community-based study. J Gerontol A Biol Sci Med Sci 2009; 64: 10 3–109.
European Journal of Clinical Nutrition (2014) 1216 – 1219
