Hala S. Sweed, MD,
Muhammad F. Abd-Al-Atty, MD,
Sherine M. El-Banouby M.B.B. CH,
Ahmed K Mortagy, MD,
Geriatrics & Gerontology Department
Faculty of Medicine - Ain Shams University
Cairo - EGYPT

Correspondence:
Dr Hala Samir Sweed
Assistant Professor of Geriatric Medicine
Geriatrics & Gerontology Department
Faculty of Medicine - Ain Shams University
Cairo - EGYPT
Tel: 002 0123349691 / 00973 39809954
Fax: 00202 22418307 / 00973 17243876
E-mail: halasweed@yahoo.co.uk

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

INTRODUCTION

Early recognition of cognitive impairment is important for diagnosis of potentially reversible medical conditions, and initiation of treatment interventions. Patients and caregivers will have time to prepare for lifestyle changes and plan for the future e.g. arranging finances and discussing end -of-life care while the patient is still competent (Boutsani et al, 2003).
Cognitive screening instruments developed for detecting dementia contain items that are partly reading and writing dependent. Therefore, they must be adapted for education and cultural difference before applying to illiterate populations (Xu et al, 2003). Less educated people were found to perform worse on mental status tests than those with more education (Mortimer & Graves, 1993).
Mini Mental Status Examination (MMSE) (Folstein et al, 1975) is the most widely used short cognitive test in clinical practice, research, and epidemiological studies (Huppert et al, 2005). It is considered the most commonly administered psychometric screening assessment of cognitive functioning. The MMSE is used to screen patients for cognitive impairment, track changes in cognitive functioning over time, and often to assess the effects of therapeutic agents on cognitive function (Strauss et al, 2006).
MMSE specificity and sensitivity are limited when the test is applied to subjects with little or no formal education, thus limiting its appropriateness as a screening instrument for dementia in populations with high illiteracy rates. Adjusting the MMSE cutoff scores according to schooling reduces the number of false-negatives in samples with more schooling and the number of false-positives in the less educated, especially the illiterate (Mungas et al, 2003).
The traditional MMSE cut-off score of 23 or less had a sensitivity of 69% and a specificity of 99%. Use of age- and education-specific cutoff scores improves the sensitivity to 82% with no loss of specificity. The clinical utility of the MMSE and acceptance by physicians may be improved through awareness of the influence of education on the MMSE (Tangalos et al, 1996).
The aim of the current study is to explore the impact of educational level on the score of the Mini Mental State Examination (MMSE) in a sample of elderly Egyptians.

METHODOLOGY

Study population:
The study was carried out - between July 2008 and December 2008 - in 8 elderly clubs randomly chosen from a list of Geriatric clubs in Cairo and Giza Governorate, was reviewed and updated by the Ministry of Social Affairs, and provided by the social worker of the Department.
The study was reviewed and approved by the Research Review Board of the Geriatrics and Gerontology Department, Faculty of medicine, Ain Shams University.
Three hundred participants were recruited for the study. The subjects who were apparently healthy, cognitively normal, and ambulant at the time of assessment, aged 60 years and over, both males and females, were included in the study.
A subject was defined as cognitively normal if there were no complaints about memory impairment or any other cognitive domain and no evidence of impairment in the activities of daily living stemming from cognitive disturbances (Inzelberg et al., 2007).
Explanation of the study aim and procedures was given to all subjects with informed consent taken from each and those who refused to participate were excluded from the study. Also, subjects who scored more than 5 in the Geriatric Depression Scale, fifteen items (GDS-15) were excluded from the study.

Tools of Assessment:
All participants were subjected to Comprehensive Geriatric Assessment (CGA) including:
* Full medical and personal history including Educational level, marital status and medical history
* Functional assessment was done using Activities of Daily Living questionnaire (ADL) (Katz et al, 1963), whereas, presence of depression was assessed using Geriatric depression scale 15 items (GDS-15) (Sheikh & Yesavage, 1986).
*Mini-mental status examination (MMSE) (Folstein et al, 1975) -Arabic version (El Okl et al, 2001) provided by the Department, was used for assessment of cognitive function. The MMSE assesses different domains of cognitive functions with a total score of 30.

The MMSE comprises:
30 questions with 10 devoted to orientation (five regarding time and five regarding place);
three items requiring registration of new information (repeating three words);
five questions addressing attention and calculation (mental control questions requiring patient to make five serial subtractions of 7 from 100 or spell word backwards);
three recall items (remembering the three registration items):
eight items assessing language skills (two naming items, repeating phrase, following a three-step command, reading and following a written command and writing a sentence);
and one construction question (copying a figure consisting of two overlapping pentagons).
A score less than 24/30, indicates cognitive impairment.

Statistical methods
The data was collected, coded and entered into a personal computer (P.C.) The data was analyzed with the program (SPSS) statistical package for social science under Windows version 13.0. Qualitative data was presented in form of frequency tables (number and percentage). Quantitative data was presented in form of mean ± standard deviation and range.
Education level was treated as an ordinal variable by grouping the subjects by years of formal schooling. Subjects were divided into 4 groups according to educational levels: Group A (less than 3 years of education), Group B (from 3 to 9 years of education), Group C (from 10 to 12 years of education), Group D (more than 12 years of education)
Pearson correlation coefficient was performed to test correlation between 2 quantitative variables, while One way Analysis of Variance (ANOVA) was used to test for comparison between multiple groups with Quantitative continuous variables. Independent sample-t test was also used to compare two groups with quantitative continuous variables.
Multinomial logistic regression analysis was done to determine the independent association of different factors. P value was always set as significant at 0.05.

RESULTS

Of the subjects who were approached, 20 elderly who had depressive symptoms (GDS more than 5) were excluded as depression itself sometimes is associated with lower cognitive functioning even when dementia is excluded (Ganguli et al, 2006), 13 refused to participate, whereas 300 cognitively normal subjects constituted the study group. There were 165 (55%) male and 135 (45%) female.
Age of the studied sample ranged from 60 to 90 years with mean age 70.7± 6.5(SD) (males 71.03±6.6, females 70.4± 6.4). 108 of them (36%) were unmarried (single or widow) while 192 (64%) were married.
Educational years of the studied sample ranged from 0 to 16 years with mean educational years 9.3±6.5 (SD) (Table 1). Males (11.5±6.1) had more educational years than females (6.6±6.1) with statistically significant difference (t=6.98, p=0.000)

Table 1 Distribution of education level in the studied sample

Of the studied sample, 57.7% (n=173) had at least one chronic disease, 42.3% (n=127) were hypertensive on therapy, 23.7% (n=71) were diabetic, 14.3% (n=43) had ischemic heart disease (IHD) and 10.7% (n=32) had chronic obstructive pulmonary disease (COPD) previously diagnosed.
Mean MMSE score of the studied sample was found to be 26.02± 3.9 (Table 2). Males were found to have higher mean MMSE (27± 3.5) compared to females (24.7± 4.0) with the difference being highly significant statistically (t=5.2, p=0.000), whereas, no statistically significant correlation was found between age of the studied cases and the total MMSE score (r=-0.087, p=0.1).
Eighty one participants were within the lowest percentile (25th) and 61.1% of them were females (n=56).

Table 2 Descriptive statistics of the total MMSE score among studied sample

There was a lower mean MMSE among unmarried (single or widowed) participants (25.3±3.7) compared to married participants (26.3±4.0) and the difference was found to be significant statistically (t=2.1, p=0.03).
Variables with the lowest scores among the bstudied group were the calculation and then the recall variables (Table 3). All studied subjects got 100% scores on registration point.
When comparing the MMSE score of the studied group and their education, a highly significant positive correlation was found between number of educational years and the total mean MMSE (r=0.742, p=0.000).

Table 3 Descriptive statistics of the mean parameters included in the calculation of MMSE

There was a higher mean of total MMSE among cases with 10-12 years and above 12 years compared to other groups and the difference was highly significant statistically (F=106, p=0.000) (Table 4).
LSD test showed significant difference between; Group A vs B, C, D -
Group B vs C, D, whereas, no significant difference was shown between group C and D.
There was a highly significant positive correlation between the educational years and the orientation score, calculation score, and language score whereas no significant correlation was found with the recall score (Table 5).

Table 4 Correlation between EDUCATION of the studied patients and the mean MMSE

Table 5 Correlation coefficient between the educational years and the parameters included in the calculation of MMSE

There was a statistically significant higher mean educational year among cases with positive visuo-spatial score compared to cases negative for visuo-spatial (Table 6).
When assessing each gender separately, again a statistically significant difference was found between different educational levels as regards their MMSE score (Table 7).
Factors found significantly related to MMSE in the current study included education, marital status and female gender. Logistic regression analysis showed powerful education correlates to test performance independent of other factors (Table 8).

Table 6 Correlation between results of visuo-spatial item and the mean educational years

** P<0.01 highly significant

Table 7 Correlation between EDUCATION of studied males and females and the mean MMSE
ANOVA test

** P<0.01 highly significant

Table 8 Logistic regression model for the independent risk factors for LOW MMSE score (below 50th percentile) median score

DISCUSSION

Education was found to be the most powerful independent factor affecting MMSE score compared to age and gender. The results of the current study showed that there was a highly significant positive correlation between number of educational years and the total mean MMSE. Those with low educational years have twelve times the risk of having lower MMSE score compared to those with higher educational years.
Education introduces a psychometric bias leading to a misclassification of individuals from different educational backgrounds (Tombaugh & McIntyre, 1992). Subjects of limited literacy do worse than educated subjects, having 6 years of compulsory formal primary education, on a variety of neuropsychological tests later in life clarifying the value of education in early life on cognitive performance in the elderly (Elwan et al, 1996).
Crum et al. (1993) reported the MMSE scores to be related to schooling level. The mean was found to be 29 for individuals with more than 9 years of schooling, 26 for those with 5 to 8 years of schooling, and 22 for those with 0 to 4 years of schooling.
In another study employing the CERAD battery (Consortium to Establish a Registry for Alzheimer's Disease), which includes the MMSE, regression analyses indicated powerful education and less marked age and gender correlates of test performance (Unverzagt et al. 1996).
Marcopulos et al., (1997) attempted to establish preliminary norms for nine commonly administered neuropsychological tests including the MMSE. The use of previously existent test norms with lower-educated, rural-dwelling, older adults was found to result in over-estimation of cognitive impairment. Persons with fewer than 8 years of education often scored below the cutoff originally suggested for indicating cognitive impairment, including on the MMSE, irrespective of racial identity.

Scores at low levels of education should be treated with caution to prevent false positive interpretation (Inzelberg et al, 2007). In a Brazilian study, analysis of covariance taking age into account showed that MMSE scores were significantly lower among those with no formal education (Almiedo, 1998).
In a healthy older Hispanic sample, educational level effects were found to be stronger than age effects on the MMSE scores. The performance of participants with no schooling was similar to that of moderately demented patients, while the performance of participants with 1-4 years of education resembled the performance of mildly demented patients (Solis et al, 2000).
Also when the Korean Mini Mental State Examination (K-MMSE) was applied to the cognitively normal, results showed that the K-MMSE scores were related to the level of education and concluded that the normative data of those with lower educational levels should be different than those with higher educational levels (Changsu et al, 2008).
When correlation coefficient was studied between the educational years and the individual items included in the MMSE, there was a highly significant positive correlation between the educational years and the "calculation", "Language", and "orientation" scores among studied patients (P<0.01). But, there was no significant correlation between educational years and the recall score.
These results agree with Solis et al., (2000) who mentioned that items that were most sensitive to educational level were those that involve reading, writing, and calculation.
Also in an Italian study, MMSE results were coded in10 item bundles. Six of the 10 item bundles were found to have differential functioning related to education. Items that required literacy skills (writing a sentence, following a written command) mathematics-based serial sevens and interlocking pentagons items, were much more difficult for those with less schooling (Crane et al, 2006).
In the current study, there was no statistically significant correlation between age of the studied subjects and the total MMSE score. This might disagree with some of the normative scores published for both age and educational levels (Grigoletto et al, 1999; Solis et al, 2000). Crum et al. (1993) found an inverse relationship between MMSE scores and age, with a mean of 29 for those 18 to 24 years of age and 25 for individuals 80 years of age. The current study had a narrower range of age (from 60 to 90 years) to assess.
When Mini-Mental State Examination (MMSE) was adapted to the Slovenian language and the influences of age and education on its score were evaluated, the positive effect of education was stronger than the negative effect of age. Older non-demented subjects with a high level of education achieved higher scores than did similar subjects with less education (Rakuska et al, 2006).
As regard gender influence on the MMSE, there was a higher mean MMSE among males compared to females and the difference is highly significant statistically.
There was actually more of those with lowest education level (A) among females in the studied sample (n=52/135, 38.5%) compared to the males (n=27/165, 16.4%) and males had more mean educational years. Yet, within the same education level (B, C), females tend to have lower mean score, suggesting additional factors beyond education. One of them could be the different social exposure and lifestyle of males versus females in this cultural group.
Gender was also found to be associated with the Korean version of the MMSE (K-MMSE) scores (Changsu et al, 2008).
While in those with more than 12 years education (D), females tend to have a higher score, a possible explanation could be that in the studied generation, girls who were intellectually privileged were given the opportunity to spend more years in education, while boys had the opportunity to study routinely (Inzelberg et al, 2007).
Nowadays, females have the same opportunities of higher education, health literacy and employment as males, so gender difference in MMSE may disappear in a few decades.
Effect of companionship on cognitive functions was clear in the current study as we have noticed that being married was associated with higher MMSE scores than being single or widow.
Intimate, confiding relationships may be most valuable to a person's well-being and mental health in old age. About four out of five elderly persons report having confidantes. When available, spouses are most likely to be listed as confidantes, followed by friends, children, and siblings (Johnson & Troll, 1994).
The main limitation of our work is the possibility that very mild dementia might have been unrecognized and misclassified as cognitively normal. Older adults in the Egyptian community mostly are being cared for by their families, and certain functional limitations may be underestimated owing to low expectations from the elderly.
Because of the effect of education on performance, different cut-off scores should be used for different educational strata. Further studies are needed to develop norms accordingly for such an Arabic translated form. Adjusting cut-off scores according to educational level can raise the sensitivity and specificity of the MMSE.

CONLCUSION AND RECOMMENDATION

Both male and female elderly Yorubas in Nigeria had a low intake of calories, vitamin A and protein. Elderly women have a greater nutritionally vulnerability than men, since they consume significantly less calories, protein and iron. There is a particular need for a nutrition intervention program, specifically for elderly individuals in Nigeria.

1. Almiedo OP (1998): Mini mental state examination and the diagnosis of dementia in Brazil. Arq Neuropsiquiatr. 56(3B):605-12.
   
2. Boutsani M, Peterson B, Hansoni I et al., (2003): Early Identification of Alzheimer's disease. Ann Intern Med 138:925-6
   
3. Changsu H, Sangmee A, Inho J et al., (2008): An adaptation of the Korean mini-mental state examination (K-MMSE) in elderly Koreans: Demographic influence and population-based norm (The AGE study) Archives of Gerontology and Geriatrics 47: 302-310.
   
4. Crane PK, Gibbons LE, Jolly L et al., (2006): Differential item functioning related to education and age in the Italian version of the Mini-mental State Examination International Psycho geriatrics; 18:3, 505-515.
   
5. Crum RM, Anthony JC, Bassett SS, et al., (1993): Population-based norms for the Mini-Mental State Examination by age and education level JAMA; 269:2386-2391.
   
6. CSHA (1994): The Canadian study of health and aging: risk factors for Alzheimer's disease in Canada. Neurology; 44: 2073-2080.
   
7. El Okl MA, El banouby MH, Mortagy AK, et al, (2001): Prevalence of AD and other types of dementia in Egypt. Tenth Congress of the International Psychogeriatrics Association. International Psychogeriatrics 13(2):114-S
   
8. Elwan 0, Hassan A , Abdel Naseer M , et al., (1996): Brain aging in normal Egyptians: Neuropsychological, electrophysiological and cranial tomographic assessment. J. Neurol. Sci;136: 73-80.
   
9. Folstein MF, Folstein SE, & McHugh PR (1975): Mini-Mental State: A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research; 12: 189-198.
   
10. Fratiglioni L, Viitanen M, von Strauss E, et al., (1997): Very old women at highest risk of dementia and Alzheimer's disease: incidence data from the Kungsholmen Project, Stockholm. Neurology; 48: 132-138.
    
11. Ganguli M, Du Y, Dodge HH, et al., (2006): Depressive symptoms and cognitive decline in late life. A prospective epidemiological study. Arch Gen Psychiatry 63:153-160.
    
12. Grigoletto F, Zappala G, Anderson DW, et al., (1999): Norms for the Mini-Mental State Examination in a healthy population. Neurology 53:315 -320.
    
13. Heyman A, Peterson B, Fillenbaum G, et al., (1996): The consortium to establish a registry for Alzheimer's disease (CERAD). Part XIV. Demographic and clinical predictors of survival in patients with Alzheimer's disease. Neurology; 46:656-660.
    
14. Huppert FA, Cabelli ST, Matthews FE (2005): Brief cognitive assessment in a UK population sample - distributional properties and the relationship between the MMSE and an extended mental state examination BMC Geriatrics; 5-7.
    
15. Inzelberg R, Schechtman E, Abuful A, et al.,(2007): Education effects on cognitive function in a healthy aged Arab population. International Psychogeriatrics 19(3); 593-603.
    
16. Johnson CL, Troll LE (1994): Constraints and facilitators to friendships in late life. Gerontologist; 34:79-87.
    
17. Katz S, Ford AB, Moskowitz AW et al. (1963): Studies of illness in the aged. The index of ADL: a standardized measure of biological and psychosocial function. JAMA 185: 914-919.
    
18. Marcopulos BA, McLain CA, & Giuliano AJ (1997): Cognitive impairment or inadequate norms? A study of healthy, rural, older adults with limited education. The Clinical Neuropsychologist; 11: 111-13.
    
19. Mortimer J& Graves A (1993): Education and other socioeconomic determinants of dementia and Alzheimer's disease. Neurology; 135:1051-1059.
    
20. Mungas D, Reed BR, Kramer JH (2003): Psychometrically matched measures of global cognition, memory, and executive function for assessment of cognitive decline in older persons. Neuropsychology; 17:380-92.
    
21. Rakuska M, Granda G, Kogo A, et al.,(2006): Mini-Mental State Examination: standardization and validation for the elderly Slovenian population. European Journal of Neurology; 13: 141-145.
    
22. Sheikh JI, and Yesavage JA. (1986): Geriatric Depression Scale: Recent evidence and development of a shorter version. Clin Gerontol 5;165-172.
    
23. Solís FO, López-Arango G, and Ardila A (2000): Sensitivity and Specificity of the Mini-Mental State Examination in a Spanish-Speaking Population. Applied Neuropsychology; 7(1):25-31.
    
24. Strauss E, Sherman EM, Spreen O (2006): A Compendium of Neuropsychological Tests: Administration, Norms, and Commentary. (3rd ed.) Oxford University Press Inc, New York.
    
25. Xu G, Meyer JS, Huang Y et al., (2003): Adapting Mini-Mental State Examination for dementia screening among illiterate or minimally educated elderly Chinese. Int J Geriatr Psychiatry; 18: 609-616.