See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8355523

Validity of height and weight self-report in Mexican adults: Results from

the national health and aging study

Article  in  The Journal of Nutrition Health and Aging · February 2004

Source: PubMed

CITATIONS READS

93 1,026

3 authors, including:

José Alberto Avila-Funes Luis Miguel Gutiérrez-Robledo

Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán (INCMNSZ) Instituto Nacional De Geriatría
192 PUBLICATIONS   4,053 CITATIONS    283 PUBLICATIONS   12,259 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Burden of disease in old age in Mexico View project

Institutionnalisation et personnes âgées View project

All content following this page was uploaded by José Alberto Avila-Funes on 14 September 2015.

The user has requested enhancement of the downloaded file.

The Journal of Nutrition, Health & Aging©
Volume 8, Number 5, 2004

THE JOURNAL OF NUTRITION, HEALTH & AGING©

VALIDITY OF HEIGHT AND WEIGHT SELF-REPORT IN MEXICAN ADULTS:

RESULTS FROM THE NATIONAL HEALTH AND AGING STUDY

J.A. ÁVILA-FUNES1, L.M. GUTIÉRREZ-ROBLEDO1, S. PONCE DE LEON-ROSALES2

1. Clinic of Geriatrics, 2.Clinical Epidemiology Unit. Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico City, Mexico. Corresponding author: Luis Miguel
Gutiérrez-Robledo. Clinic of Geriatrics. Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán. Vasco de Quiroga 15, Tlalpan 14000, Distrito Federal, México. e-mail:
luismigr@data.net.mx

Abstract: Background: An adequate nutritional status is essential for maintaining the independence in the
elderly. The height and weight self-report is considered a useful alternative for the estimation of body mass index
(BMI). The validity of the self-report is an issue that has not been dealt with in developing countries. Aim: To
assess the validity of the height and weight self-report in adults Mexican citizens. Design: Transversal study.
Study population: 1707 persons (836 male, 871 female) were asked for their height and weight and were
measured. Results: Mean (± standard deviation) age was 59.09 ±9.86 SD; mean years of education was 5.51
±4.67 years. We found a high correlation between self-reported and measured weight (R2= 0.837); the difference
between both values rose along with the age (from 0.4 kg to 1.74 kg). A systematic difference between self-
reported and measured height was found. Self-reported height was over-estimated, and the bias increased along
with the age of the subjects (from 1.57 cm to 2.57 cm); further, over-estimation was larger in female individuals
(+2.22 cm in female vs. +1.21 cm in male subjects). We calculated a linear model that predicts real height from
self-reported height with moderate, although statistically significant results (R2= 0.39 y 0.50, for female and
male, respectively, p< 0.0001). Knee height was also used to estimate “adulthood height” and thus, BMI. This
method showed age-related dissimilarities, and the linear regression model yielded an unacceptably low
correlation (R2= <0.10). The best method to estimate real BMI was to consider self-reported parameters.
Discussion. Height and weight estimation using self-reported parameters is an acceptable method. Its precision is
not so high in subjects older than 75 years. Height over-estimation is an expected finding congruent with age-
related corporal changes. The estimation of height using knee height is not a useful method. Conclusion: Height
and weight self-report is a valid method that may be used to accurately estimate height and weight in Mexican
people.

Key words: Aging, body mass index, height, weight, obesity, self-report.

Introduction Several studies that have evaluated nutritional status have

In Mexico, as in the rest of the world, the proportion of elder
people has steadily risen. In the year 2000, the subset accounted
for 7% of the population, approximately 6.9 million individuals
(1). The aging population represents a rising economic and
social burden; unless a radical shift in the distribution of health
services is made, the demographic change will reflect itself in
higher morbidity and mortality, as well as in a significant
detriment of the quality of life and independence of this
component of the population (2,3).
Autonomy of the older person is strongly linked to
nutritional status. Age-related changes in body composition,
such as decrease in muscle and increase in fatty tissue, are
associated with a reduced basal metabolic rate and total energy
requirement; accordingly, caloric intake decreases (4,5).
Chronic, debilitating illnesses, in the setting of non biologic
problems (e.g. poverty, social isolation) add to anomalous
caloric intake. Common results are obesity or malnutrition.
The assessment of the nutritional status of older people is
relevant. It allows the health professional to detect risk factors,
undertake pertinent preventive measures, and decrease
complications.
relied in the body mass index (BMI, weight in kg/[height in
meters] 2 ). In most cases, it is an adequate indicator of
malnutrition; however age-related corporal changes may
invalidate its meaning (6). Height reduction is one of the most
conspicuous corporal changes in older people. It may cause
BMI to be imprecise and yield over-estimations not related to a
real increase in body fatty tissue (7-10). On the other hand, age-
related weight changes have been associated to an increased
mortality (11).
Anthropometrical indicators are an economic and practical
method to evaluate corporal dimensions and thus estimate
nutritional status. However, even though the measurements are
relatively simple to obtain, they commonly are limited by the
impaired mobility of elder people. Furthermore, they are not
used in epidemiological studies because they require
cumbersome equipment standardization and personnel training.
Perhaps the most straightforward method to estimate
nutritional status is to simply ask the subject his/her height and
weight: the height and weight self-report. Although it relies on
the memory of the audited person, several studies performed in
developed countries have demonstrated its validity since 1981
(12-16). Some of these have been conducted in older people,

355
The Journal of Nutrition, Health & Aging©
Volume 8, Number 5, 2004
VALIDITY OF HEIGHT AND WEIGHT SELF-REPORT IN MEXICAN ADULTS
(17) including patients with cognitive impairment (18).
Nevertheless, the validity of self-reported health information is
still sometimes doubted; moreover, it is an issue not evaluated
in Mexico and other developing countries.
The aim of the present work was to determine the validity of
the height and weight self-report as an anthropometrical
method useful in the setting of epidemiological studies, where
conventional clinical methods are not applicable.
Methods
The Mexican Health and Aging Study (MHAS) is a
prospective project that includes Mexican citizens born before
1951 and their spouse or partner disregarding their age. The
aim and design of the MHAS have been published previously
(19,20) It was derived from the fourth round of the National
Employment Survey and it is regarded as a representative
sample of the Mexican citizenship born before 1951. It
considers individuals from urban and rural areas. Data was
obtained from direct, face to face, individual audit. The MHAS
is only represents the non-institutionalized component of the
population aged 50 and over in 2000. In the case of Mexico,
however, it is not a gross omission because, according to the
1990 National Population Survey, only 0.4% of people older
than 60 years lived in an institution. The MHAS includes data
from 15230 interviews (9806 index cases and 5424
spouses/parteners). It comprises information regarding self-
reports of symptoms, height and weight, health related
behaviors (smoking habit, alcohol use), access to health
services, depression, pain, cognitive performance, and
anthropometrical measurements. Additionally, it considers
childhood demographic data, education, migratory history and
marital status. In approximately 20% of the interviewed the
following measurements were performed: weight, height, waist,
hip and calf circumference, knee height, and balance over one
limb.
Studied Population
We randomly chose 2944 persons from the MHAS database
(19.3%). Complete height and weight measurements were
performed in 2527 of them (85.8%); 417 were excluded
because they were unable to move, they could not stand up
alone, or had severe spinal deformities. Further, 820 had to be
excluded from the final analysis because they had not answered
the self-report. Thus, 1707 individuals comprised the final
studied population (57.9% of the 2944 initially considered).
Experimental procedure
The investigator obtained the self-reported height and weight
during a direct interview. Patients were asked: How many
kilograms do you weight? and What is your height without
shoes?
Next, patients were weighted and measured. Interviewers
were trained to make standardized anthropometrical
356
measurements. Patients wore clothes with specific
characteristics. Weight was quantified with a traditional scale.
It stood on a flat floor, next to a wall so the patient could lean
in case he lost his balance. The patient was asked to stand still
on the scale, without shoes, with his hands hanging free in a
vertical position. Weight was recorded in kilograms. Next,
patients were asked to stand with their back facing a wall, in an
erect position, with their ankles together, on a flat hard floor.
Height was recorded in centimeters.
Knee height was measured and the estimated adulthood
height of each individual was calculated according to Chulmea
et al. (21,22). Knee height was measured in the right leg, except
in those individuals who had an injury in the mentioned limb.
In those cases, left knee height was recorded. Patient was asked
to sit in a chair and the knee was bended 90 degrees. The
distance in centimeters from the floor to the patella was
recorded in centimeters.
Three body mass indexes were calculated for each patient.
One with data obtained from the clinical measurements, another
considering the estimated adulthood height, and a third with the
height and weight obtained in the self-report.
Statistical analyses
For descriptive purposes, the arithmetic mean and the
standard deviation (SD) of each somatometric index was
calculated. Additionally, the following statistical procedures
were used: (23-25)
1. Spearman’s correlation coefficient (rs), to quantify the
magnitude of associations.
2. Wilcoxon’s rank test, to determine the statistical
significance of the difference between two paired variables.
3. Linear regression analysis, considering the real
measurement as the dependent variable (y) and the self-
report (or the estimated height) as the independent variable
(x), with the calculation of the significance of the difference
from the expected results according to the null hypothesis
of identity between methods: slope = 1 and intercept = 0.
4. Intraclass correlation coefficient (ICC) derived from a 1-
way random effects analysis of variance as described by
Bartko (26). It was interpreted according to Landis and
Koch suggestions about kappa index, its equivalent for
categorical variables (27).
5. Deremination of concordance limits at 95% according to
the method by Bland y Altman.28. Briefly, this consist in
the identification of the interval of distribution of the
differences between each pair of measurements, made with
each of the two methods.Depending on the amplitud of this
interval, in usual units of each method, it may be resolved if
a method shows acceptable concordance or not.
To consider results compatible with good accuracy or
reliability, the following criteria were set a priori for each
procedure: Absence of significance in the Wilcoxon statistical
test, and the slope and intercept with respect to the straight line
The Journal of Nutrition, Health & Aging©
Volume 8, Number 5, 2004

THE JOURNAL OF NUTRITION, HEALTH & AGING©

of identity between methods. higher in patients older than 75 years. In this group of age it
In order to achieve the best prediction from self reported ranged between a 14.75 kg under-estimation to a 16.6 kg over-
data, we looked for a polynomial model beyond the simple estimation in males, and from a 8.6 kg under-estimation to a
straight line. Usually it was useless to go further than the 14.7 kg over-estimation in females. Intraclass correlation
quadratic model and even so we didn’t find a significant coefficient were high in all subgroups (ICC >0.80), except in
improvement. Thus, these results won’t be mentioned anymore male subjects older than 75 years (ICC = 0.77). Refer to Table
along this paper. 3 and 4.
Statistical test were performed using the Stata package The equation for the prediction of weight using self-reported
version 7.0 for Windows. weight exhibited a lower precision in male individuals older
than 75 years (R2= 0.596). Nevertheless, it was statistically
Results significant (p < 0.0001). In the female subgroup it showed a
high R2 value (0.852).
General characteristics The equation for the prediction of real weight using self-
The studied population included 1707 individuals. Mean (± reported weight in subjects older than 75 years was:
standard deviation) age was 59.09 ±9.86 years (range 24 to 95);
mean years education was 5.51 ±4.67 years (range 0 to 19). Male: y= 11.5 + 0.81 (self-reported wieght)
Table 1 summarizes the demographic characteristics of the Female: y= 9.9 + 0.78 (self-reported weight)
studied population. As expected, anthropometrical parameters
differed between male and female subjects (Table 2). Self-reported vs. real height
We observed an age-related, obvious tendency, to over-
Table 1 estimate height (from 1.57 to 2.75 cm). Once again, the bias
Demographic Characteristics. was larger in female subjects (+2.22 cm in women vs. +1.21 cm
in men) of all the groups of age. Further, the error had a higher
Age Male Female magnitude in persons of older age. In women older than 75
years, we found a 11.8 cm under-estimation and a 19.3 cm
No. (%) No. (%) over-estimation. In men, values fluctuated between –13.5 cm
< 59 years 418 (50) 544 (62) and 17.8 cm. ICC was high, except in women older than 75
60-74 years 340 (40) 278 (32) years (ICC= 0.56). Refer to Table 3 and 4.
≥ 75 years 78 (10) 49 (6) The performance of the prediction model designed to
Total 836 (100) 871 (100) estimate real height using self-reported height was not as good
(R2=0.39 for women and R2= 0.50 for men), yet it reached
Table 2 statistical significance (p < 0.0001).
Anthropometric Parameters According to Sex. The equation to estimate real height using self-reported
height in individuals older than 75 years is:
Male Women Both
Mean ± SD Mean ± SD Mean ± SD Male: y = 85.4 + 0.46 (self-reported height [cm])
Female: y = 67.9 + 0.52 (self-reported height [cm])
Real weight (kg) 74.1 ± 14.3a 66.7 ± 13.2b 70.3 ± 14.2
Self-reported weight 74.5 ± 14.2a 67.4 ± 13.1b 70.89 ± 14.1
Estimated adulthood height vs. real height
(kg)
Real height (cm) 165.2 ± 8.0a 153.1 ± 7.7b 159.0 ± 9.9 The method proposed to estimate height using knee height,
Self-reported height 166.4 ± 8.6a 155.3 ± 7.9b 160.7 ± 10.0 showed large age-related variability; analogously to the other
(cm) procedures, the imprecision was higher in female subjects.
Estimated height (cm) 165.8 ± 9.4a 151.1 ± 8.4b 158.2 ± 11.5 In male individuals, differences ranged from nearly cero
Real BMI 27.1 ± 5.0a 28.4 ± 5.5b 27.8 ± 5.3 (0.04 cm, p= 0.21) in the younger age group (< 59 years), to
Self-reported BMI 26.8 ± 4.6a 27.9 ± 5.2b 27.4 ± 5.0 approximately 1 cm (0.95 cm, p= 0.0002) in subjects 60 to 74
Estimated BMI 27.1 ± 5.9a 29.4 ± 6.6b 28.3 ± 6.4 years old, to an under-estimation of almost 3 cm (3.1 cm; p=
0.0013) in persons older than 75 years.
a,b p < 0.0001 (male vs. female)
In female subjects, differences ranged from an almost 3 cm
over-estimation (3.1 cm, p<0.00001) in the younger group, to a
Self-reported versus real weight correct estimation in subjects 60 to 74 years old (0.6 cm,
We found a systematic over-estimation of the self-reported p=0.98), to a roughly 5 cm under-estimation in women older
weight. The bias increased along with the age (from 0.4 kg to than 75 years (4.9 cm, p=0.0017). ICC showed a slight to
1.74 kg); further, it was larger in female subjects (+3.06 kg in moderate correlations (Table 3 and 4).
males vs. +0.92 kg in females). The magnitude of the error was The Bland and Altman prediction error was approximately

357
The Journal of Nutrition, Health & Aging©
Volume 8, Number 5, 2004

VALIDITY OF HEIGHT AND WEIGHT SELF-REPORT IN MEXICAN ADULTS

20 cm in all the studied groups. Linear regression equations was acceptable amongst women (R2 = 0.65-0.76); but lower in
yielded unacceptably low values (R 2< 0.10). Thus, height men (R2 = 0.33-0.56). Refer to Table 3 and 4.
estimation using knee height is not a valid procedure. The equation to estimate real BMI using self-reported

Table 3
Differences Between the Self-Reported/Real Parameters (CI 95%), Bland and Altman Concordance Interval, and Intraclass
Correlation Coefficient (ICC).

Difference Bland and Altman

Mean (95% CI) Inferior Limit Superior Limit ICC

Self-reported vs. real weight (kg)

Female < 59 years 0.47 (-0.03 – 0.99) -11.78 +12.78 .882
60-74 years 0.51 (0.02-0.99) -7.75 +8.77 .950
≥ 75 years 3.06 (1.38-4.73) -8.59 +14.71 .888
< 59 years 0.27 (-0.34 – 0.90) -12.78 +13.34 .903
Male 60-74 years 0.49 (-0.04 – 1.03) -9.64 +10.64 .924
≥ 75 years 0.92 (-0.84 – 2.69) -14.75 +16.60 .771
Self reported vs. real height (cm)
< 59 years 1.88 (1.43 - 2.33) -8.79 +12.55 .723
Female 60-74 years 2.64 (2.01 – 3.27) -8.03 +13.32 .722
≥ 75 years 3.73 (1.50-5.96) -11.81 +19.28 .555
< 59 years 1.17 (0.56 – 1.79) -11.53 +3.89 .691
Male 60-74 years 1.05 (0.34 – 1.76) -12.33 +14.42 .661
≥ 75 years 2.14 (0.38 – 3.90) -13.47 +17.75 .663
Estimated vs. real height (cm)
< 59 years 3.08 (2.20 – 3.97) -23.34 +17.16 .138
Female 60-74 years 0.55 (-0.71 – 1.82) -21.23 +20.13 .214
≥ 75 years -4.93 (-8.39 – 1.47) -18.07 +27.95 .000
< 59 years 0.04 (-1.01 – 1.09) -21.12 +21.03 .309
Male 60-74 years -0.95 (-2.07 – 0.16) -19.30 +21.20 .287
≥ 75 years -3.08 (-5.60 –0.56) -18.21 cm +24.38 cm .120
Self-reported BMI vs. real BMI
< 59 years 0.51 (0.22 – 0.80) -7.35 +6.32 .799
Female 60-74 years 0.73 (0.41 – 1.04) -6.01 +4.55 .861
≥ 75 years -0.07 (-0.98 – 0.83) -6.26 +6.40 .807
< 59 years 0.40 (-0.009 – 0.82) -9.10 +8.28 .661
Male 60-74 years 0.02 (-0.31 – 0.36) -6.35 6.30 .734
≥ 75 years 0.19 (-0.64 – 1.04) -7.70 +7.31 .576

parameters in individuals 60 to 74 years old is:

Body mass index (BMI)
BMI is a linear product of the elements formerly discussed. Male: y = 1.3 + .61 (self-reported BMI)
Thus, the best method to estimate real BMI is to use self- Female: y = 10.3 + .61 (self-reported BMI)
reported height and weight. When it is performed, differences
between real and estimated BMI are minute, smaller than 1 The equation to estimate real BMI using self-reported
percent. Further, no age- or sex-related bias was evident. parameters in individuals 60 to 74 years old is:
The Bland and Altman prediction error fluctuated between 5
and 8 percent, except in the younger age male group were the Male: y = 9.9 + .60 (self-reported BMI)
under-estimation was 9%. ICC were moderately high (in men Female: y= 7.4 + .71 (self-reported BMI).
older than 75 years) to excellent (in women older than 60
years).
Prediction of the real BMI using self-reported parameters

358
The Journal of Nutrition, Health & Aging©
Volume 8, Number 5, 2004

THE JOURNAL OF NUTRITION, HEALTH & AGING©

Table 4
Lineal Regression estimate to Predict Somatometric Data Using Self-Reported Parameters

intercept slope R2 Model P

Self-reported vs. real weight

< 59 years 7.89 0.879 0.779 .000
Female 60-74 years 0.74 0.980 0.903 .000
≥ 75 years 9.85 0.778 0.852 .000
< 59 years 7.74 0.895 0.815 .000
Male 60-74 years 4.20 0.936 0.855 .000
≥ 75 years 11.5 0.809 0.596 .000
Self reported vs. real height
< 59 years 37.9 0.744 0.561 .000
Female 60-74 years 41.8 0.711 0.596 .000
≥ 75 years 67.9 0.524 0.393 .000
< 59 years 42.2 0.740 0.490 .000
Male 60-74 years 68.6 0.581 0.455 .000
≥ 75 years 85.4 0.464 0.504 .000
Estimated vs. real height
< 59 years 128.0 0.172 0.035 .000
Female 60-74 years 122.2 0.194 0.051 .000
≥ 75 years 152.5 -0.035 0.001 .807
< 59 years 117.5 0.291 0.104 .000
Male 60-74 years 123.9 0.247 0.093 .000
≥ 75 years 139.8 0.131 0.027 .168
Self-reported BMI vs. real BMI
< 59 years 4.84 0.847 0.645 .000
Female 60-74 years 2.35 0.940 0.760 .000
≥ 75 years 7.35 0.707 0.657 .000
< 59 years 4.17 0.863 0.464 .000
Male 60-74 years 10.3 0.613 0.559 .000
≥ 75 years 9.91 0.604 0.328 .000

Discussion with vertebral osteoporosis (8,9). The work by Chumlea et al

Our results demonstrate that weight and height estimation
using self-reported parameters is valid, specially in subjects
under age 75.
Weight, as has been showed in several published reports,
tends to be under- or over-estimated (12,16,17,29-34). The
phenomenon seems to be related to cultural trends: people with
overweight under-estimate their weight and vice versa. In our
sample, the over-estimation was specially significant in people
older than 75 years, disregarding gender. These results are
congruent with the available information: there is a high
correlation between self-reported weight and direct
measurement with a scale (Table 5).
Likewise, we observed a tendency towards a over-estimation
of height. This occurrence might have a physiological
explanation, related to the normal aging process: older people
probably report the height they remember had in their youth,
before they experienced the reduction in tallness associated
reports a high correlation (0.82) between methods, and consider
it good enough to validate self-reported height (21). In our
sample we found a good correlation when we considered the
total sample or age groups. However, when we analyzed
according to gender, the correlation was moderate (0.701 for
men and 0.763 for women). This fact diminished the prediction
capability of the model.
On the other hand, the differences observed between real
height and adulthood estimated height (using knee height as
reference) are too large to consider it a useful meted. Tibial
height is used because it is a relatively constant proportion of
total body height that does not linger with aging (8). We found
the differences were larger in the group of patients older than
75 years, disregarding gender. If one employs the adulthood
estimated height to calculate the BMI, it will be falsely low in
women, and falsely high in men. The discrepancy between our
results and the ones reported by Chumlea et al are probably due
to methodological differences. They use a special artifact to

359
The Journal of Nutrition, Health & Aging©
Volume 8, Number 5, 2004

VALIDITY OF HEIGHT AND WEIGHT SELF-REPORT IN MEXICAN ADULTS

measure knee height which we did not use (35) This probably
confers our measurements higher variability and lower
accuracy.
According to our results, the principal factor that bias the
weight self-report is age. Conversely, age and sex are the
principal causes of the inaccuracy of height self-report. It is
important to mention that half of the sample was younger than
59 years. Thus, our results are not exclusively applicable to
senior adults. The influence of other factors such as years of
education and cognitive status remain to be sought for in
further work. Finally, it is important to consider the scenario
from where data was obtained; height and weight information
retrieved during face-to-face survey might be less accurate than
information filled into a self answered format.
Conclusions
The weight and height self-report is a valid instrument. It
will prove useful in large epidemiological studies in which real
height and weight cannot be directly measured.
Adulthood estimated height calculated according to knee
height is not a convenient method to estimate height and BMI.
Even though we can rely on this statistical valid instrument
that allows us to predict real weight and height using self-
reported parameters, our results suggest caution during the use
of the self-report, because it entails the risk of introducing bias,
and thus diagnostic errors, specially when applied to certain
subsets of the elder population.

Table 5
Summary of Nine Studies That Evaluate Height Self-Report

Authors and Sample Size Age Group Correlation Mean Difference Standard
Publication Date % (n) Coefficient (kg) Deviation
(Self-Reported vs.
Measured)

Stunkard & Albaum, Total 550 0.99 -1.20 3.10

1981 (34) Male 19 % 40 (mean) NA§ NA NA
Female 81 % NA NA NA
Stewart, 1982 (29) Total 3226 0.99 -1.09 2.49
Male 46 % 15-61 0.98 -0.73 2.42
Female 54 % 0.98 -1.41 2.57
Pirie et al, 198131 Total 3407 NA -1.23 NA
Male 47 % 20-59 0.96 -0.54 3.24
Female 53 % 0.97 -1.86 3.03
Palta et al, 198232 Total 1337 NA NA NA
Male 63 % 30-69 NA -1.50 3.20
Female 37 % NA -2.40 4.00
Rowland, 198912 Total 11284 0.98 -0.34 3.11
Male 48 % 20-74 0.97 0.41 3.02
Female 52 % 0.98 -1.03 3.02
Stewart et al, 198730Total 1519 0.98 -0.58‡ NA
Male 64 % 36-65 NA NA NA
Female 36 % NA NA NA
Stunkard & Albaum, 198134 Total 16-66 NA NA NA
Male 79 % 752 0.95-0.96 NA NA
Female 21 % 0.91-0.97 NA NA
Jalkanen et al, 198733 Total 30-64 NA -0.50NA
Male 49 % 11880 NA -0.40 3.00
Female 51 % NA -0.60 2.00
MHAS, 2001 Total 1707 0.91 0.53
Male 49 % 24-95 0.90 0.42 5.84
Female 51 % 0.91 0.63

§ Information not available

360
The Journal of Nutrition, Health & Aging©
Volume 8, Number 5, 2004
THE JOURNAL OF NUTRITION, HEALTH & AGING©
Acknowledgements: We thank Dr. JC Crispín for critically reviewing the manuscript.
References
1. Peláez M. “La salud pública y el envejecimiento en América Latina y el Caribe”,
presented at the IV Congreso Latinoamericano de Geriatría y Gerontología.
COMLAT, IAG, OPS. Santiago de Chile, Sep 2003.
2. Novelo de López H.I. Aspectos demográficos del adulto mayor. Nutrición clínica
2003;6:63-69.
3. Estadísticas de mortalidad en México en el año 2001. Instituto Nacional de
Estadística, Geografía e Informática. Dirección General de Información y Evaluación
del Desempeño. Secretaría de Salud, México. Salud Pública 2002;44:6.
4. Fogt E.J., Bell S.J., Blackburn G.L. Nutritional assessment of the elderly. In: Morley
EJ, Glick Z, Rubenstein LZ, eds. Geriatric nutrition, a comprehensive review. New
York: Raven Press, 1995.
5. Poehlman E.T. Regulation of energy expediture in aging humans. J. Am. Geriatric.
Soc., 1993;41:552-559.
6. Alemán-Mateo H, Pérez-Flores F.A. Los indicadores del estado de nutrición y el proceso
de envejecimiento. Nutrición clínica, 2003;6:46-52.
7. Bates C.J., Benton D., Biesalski H.K. et al. Nutrition and aging: A consensus
statement. J. Nutr. Health. Aging, 2002;6:103-116.
8. Freidlaender J.S., Costa P.T. Jr., Bosse R. et al. Longitudinal physique changes
among healthy white veterans at Boston. Hum. Biol., 1977;49:541-558.
9. Borkan G.A., Hults D.E., Glynn R.J. Role of longitudinal change and secular trend in
age differences in male body dimensions. Hum. Biol., 1983;55:629-641.
10. Soto D., Cariaga L., Gaete M.C., et al. Clasificación nutricional de senescentes de
centros institucionales de la ciudad de Santiago-Chile. Rev. Chil. Nutr., 1988;3:306-
311.
11. Somes G.W., Kritchevsky S.B., Shorr R.I., et al. Body mass index, weight change, and
death in older adults: The systolic hypertension in the elderly program. Am. J.
Epidemiol., 2002;2:132-138.
12. Rowland M.L. Self-reported weight and height. Am. J. Clin. Nutr., 1990;52:1125-1133.
13. Heaney P., Ryan R. Relation between measured and recalled body height. N. Engl. J.
Med., 1988;319:795-796.
14. Casey V. Longterm memory of body weight and past weight satisfaction: a longitudinal
follow-up study. Am. J. Clin. Nutr., 1991;53:1493-1498.
15. Milliar W.J. Distribution of body weight and height. J. Epidemiology Comm. Health.,
1986;40:319-323.
16. Schlichting P., Carlsen P.F., Quaade F. Comparation of self-reported height and weight
with controlled height and weight in women and men. Int. J. Obesity, 1981;5:67-76.
361
View publication stats
17. Boutier V., Payette H. Validity of weight and height given from memory in an elderly
population. Age & Nutrition, 1994;5:17-21.
18. Payette H., Kergoat M.J., Boutier V., et al. Validity of self-reported height and
weight estimates on cognitively-intact and impaired elderly individuals. Journal of
Nutrition, Health & Aging, 2000;4:223-228.
19. The MHAS. Available at: http://www.ssc.upenn.edu/mhas/ Accesibility verified
february 10, 2004.
20. Palloni A., Soldo B.J., Wong R. “Health Status and Functional Limitations in a
National Sample of Elderly Mexicans”, presented at the Population Association of
America Conference, Atlanta, May 2002.
21. Chumlea W.C., Roche A.F., Steinbaugh M.L. Estimating stature from knee height for
persons 60 to 90 years of age. J. Am. Geriatr. Soc., 1985;33:116-120.
22. Chumlea W.C. Accuracy and reliability of a new sliding caliper. Am. J. Physical
Anthropol., 1985;68:425-427.
23. Tzu G., Twowey S.L., Thiers R.E. Statistical evaluation of method-comparison data.
Clin. Chem., 1975;21:315-320.
24. Lubran M. Comparison of two methods of measurement of the same substance: An
evaluation of some statistical procedures. Ann. Clin. Lab. Sci., 1982;12:135-142.
25. Kramer M., Feinstein A.R. Clinical biostatistics. LIV. The biostatistics of
concordance. Clin. Pharmacol. Ther., 1981;29:111-123.
26. Bartko J.J: The intraclass correlation coefficient as a measure of reliability. Psicol.
Reports, 1966;19:3-11.
27. Dunn G. Design and analisis of reliability studies. New York. Oxford University
Press. 1989:37.
28. Bland J.M., Altman D.G. Statistical methods for assessing agreement between two
methods of clinical measurement. Lancet, 1986;1:307-310.
29. Stewart A.L. The reliability and validity of self-reported weight and height. J. Chron.
Dis., 1982;35:295-309.
30. Stewart A.W., Jackson R.T., Ford M.A. et al. Underestimation of relative weight by
use of self-reported height and weight. Am. J. Epidemiol., 1987;125:122-126.
31. Pirie P., Jacobs D., Jeffery R., et al. Distortion in self-reported height and weight
data. J. Am. Diet. Assoc., 1981;78:601-606.
32. Palta M., Prineas R.J., Berman R., et al. Comparation of self-reported and mesured
height and weight. Am. J. Epidemiol., 1982;115:223-230.
33. Jalkanen L., Tuomilehto J., Tanskanen A., et al. Accuracy of self-reported body
weight compared to measured body weight. Scan. J. Soc. Med., 1987;15:191-198.
34. Stunkard A.J., Albaum J.M. The accuracy of self-reported wights. Am. J. Clin. Nutr.,
1981;34:1593-1599.
35. Laboratorios ROSS, Chumlea WC. Directions for the ROSS Knee Height Caliper.
Fascicle distributed by ROSS laboratories, 1990.