Abstract
Introduction
Current anthropometric cut-off points for defining overweight, obesity, and abdominal obesity are based on predominantly Caucasian populations, raising questions about their applicability in ethnically and geographically diverse populations such as the Peruvian population. The lack of population-specific references limits the appropriate assessment of anthropometric distributions and cardiometabolic risk in Peruvian adults.
Objective
To develop body mass index (BMI), waist circumference (WC), and waist-to-height ratio (WHtR) percentile curves specific for the Peruvian adult population, stratified by age and sex.
Methods
Analytical cross-sectional study based on data from the Demographic and Family Health Survey (DHS) 2015 to 2023. From an initial sample of 385,247 adults aged 20–97 years, 164,513 participants with complete anthropometric measurements and plausible values were included in the final analysis. Percentiles (P5, P10, P25, P50, P75, P85, P90, P95, P97) were calculated using the LMS (Lambda-Mu-Sigma) method for BMI, WC, and WHtR by five-year age groups and sex.
Results
The final sample included 164,513 participants (48.52% women, 51.48% men) with a mean (±standard deviation, SD) age of 44.38 ± 16.72 years. Mean BMI was 27.77 ± 4.80 kg/m². The prevalence of overweight was 42.16% and obesity 28.09% according to BMI criteria. Abdominal obesity varied dramatically according to the criterion used: 44.75% (ATP III) versus 68.31% (IDF) for WC, highlighting the problem of universal cut-off points. Sex-specific percentiles showed distinctive age patterns with increases until middle age followed by gradual decline. Altitude differences were clinically significant: 2.5–3.0 kg/m² for BMI, 4–6 cm for WC, and 0.04–0.06 units for WHtR between sea level versus high altitude (≥3000 m).
Conclusions
Age- and sex-specific anthropometric percentiles reveal distinctive age patterns that differ substantially from international cut-off points. Classical criteria for abdominal obesity (ATP III, IDF) and WHtR (≥0.5) systematically overestimate prevalence in the Peruvian population, with discrepancies reaching differences of up to 23.5 percentage points between criteria. Altitude differences are clinically significant, suggesting specific physiological adaptations in Andean populations.
Subject terms: Obesity, Epidemiology
Introduction
Anthropometric assessment constitutes a fundamental tool in clinical practice and nutritional epidemiology for identifying individuals at elevated risk of non-communicable chronic diseases [1]. However, the anthropometric cut-off points currently employed worldwide present inherent limitations derived from their specific methodological origins. Body mass index (BMI) cut-off points of 25 and 30 kg/m² were established based on mortality epidemiological studies in Caucasian populations and actuarial data from US insurance companies [2]. Waist circumference (WC) criteria from the Adult Treatment Panel III (ATP III: ≥102 cm men, ≥88 cm women) [3] were originally derived from specific BMI thresholds, particularly from the 30 kg/m² cut-off point for obesity; similarly, the International Diabetes Federation (IDF: ≥90 cm men, ≥80 cm women) [4] criteria represented empirical adjustments for Asian populations without exhaustive validation in other ethnic groups [5, 6]. The waist-to-height ratio (WHtR) cut-off point of 0.5 originates from heterogeneous systematic reviews that proposed this threshold as a universal predictor of cardiometabolic risk, without considering population variability [7].
The application of these international cut-off points in the Peruvian population has generated alarmingly high prevalence estimates of abdominal obesity, which question their validity and applicability. Recent studies document abdominal obesity prevalence reaching 60–80% according to ATP III and IDF criteria, and similar values for WHtR ≥0.5, figures that suggest possible systematic overestimation of risk in the Peruvian population [8, 9]. This extreme variability in prevalence estimates, depending on the criterion used, raises fundamental questions about the suitability of these cut-off points to adequately characterize cardiometabolic risk in the Peruvian context. The indiscriminate application of criteria developed in ethnically and geographically distinct populations may lead to erroneous classifications that affect both individual clinical practice and public health policy planning [10].
The Peruvian context presents unique characteristics that justify the development of specific anthropometric references. Peru is characterized by extraordinary ethnic diversity, including mestizos (60.2%), Quechuas (22.7%), Aymaras (2.7%), Amazonian groups (4.9%), and other ethnic groups, each with distinctive anthropometric characteristics resulting from millennia of adaptation to diverse environments [11]. Altitudinal variation from sea level to over 4000 m significantly influences body composition, metabolism, and body fat distribution patterns [12, 13]. Previous studies have documented that Andean populations present distinctive physiological adaptations, including different body fat distribution and unique metabolic patterns, which are reflected in specific anthropometric profiles that do not conform to conventional international standards [14, 15].
In pediatrics, the use of age- and sex-specific anthropometric percentiles constitutes the gold standard for assessing growth and development, explicitly recognizing that “normal” anthropometric values vary according to individual and population characteristics [16]. Recently, Banack et al. proposed extending this approach to older adults, developing age- and sex-specific BMI percentiles for the Canadian population, demonstrating that the use of percentiles significantly improves diagnostic accuracy compared to fixed cut-off points [16]. While validation against cardiometabolic outcomes would strengthen clinical utility, population-specific percentiles provide valuable information in their descriptive capacity by allowing classification of individuals relative to their reference population, facilitating identification of extreme values warranting clinical attention, and enabling evaluation of whether universal cut-off points appropriately classify specific populations. This approach establishes a necessary foundation for subsequent validation studies [16].
The lack of specific anthropometric references for the Peruvian population represents a significant gap in assessing nutritional status and cardiometabolic risk. The objective of this study is to develop anthropometric percentile curves (BMI, WC, and WHtR) specific for Peruvian adults, stratified by age and sex, using representative population data from the 2015–2023 period. Specifically, we propose to: [1] estimate anthropometric percentiles by five-year age groups and sex; [2] evaluate how current international cut-off points classify the Peruvian population by determining their equivalent percentile positions; [3] assess regional differences in anthropometric percentiles; and [4] provide population-specific percentile reference curves that can serve as a foundation for future development of validated criteria for anthropometric assessment in Peruvian adults.
Methods
Study design and data source
An analytical cross-sectional study was conducted based on secondary data from the Demographic and Family Health Survey (DHS) of Peru, corresponding to the 2015–2023 period. The DHS is a nationally representative survey executed annually by the National Institute of Statistics and Informatics (INEI), which is part of the global DHS program [17]. This survey uses a probabilistic, two-stage, stratified, and balanced sampling design at the departmental level and by area of residence (urban/rural), with a sampling frame based on cartographic and demographic information from the 2017 National Census.
Population and sample
The target population comprises Peruvian adults aged 20–97 years residing in private households during the study period. All adult participants with complete data on weight, height, and WC, who had provided documented informed consent, were included in the analysis. Pregnant women or those suspected of pregnancy, participants with extreme or biologically implausible anthropometric values (BMI < 10 kg/m² or >70 kg/m²; WC < 40 cm or >200 cm; height <120 cm or >220 cm; weight <30 kg or >300 kg), and those with missing data on main anthropometric variables were excluded.
Variables and measurements
The main anthropometric variables include BMI, calculated as weight (kg) divided by height squared (m²) and categorized according to WHO criteria as normal weight (18.5–24.9 kg/m²), overweight (25.0–29.9 kg/m²), and obesity (≥30.0 kg/m²); WC, measured in centimeters at the midpoint between the lower border of the last palpable rib and the superior iliac crest; and WHtR, calculated as WC (cm) divided by height (cm). Additionally, abdominal obesity was defined according to two internationally recognized criteria: ATP III (≥102 cm in men, ≥88 cm in women) [3] and IDF (≥90 cm in men, ≥80 cm in women) [4].
To characterize the study population, various sociodemographic covariates were included. Sex was categorized as women or men, and age was recorded in completed years. Natural region was classified into four categories: Metropolitan Lima, Rest of Coast, Highlands (Sierra), and Jungle (Selva). Educational level was categorized into four levels: no education, primary, secondary, and higher education (superior). Marital status was divided into single, married, cohabiting, widowed, and divorced.
Additionally, socioeconomic and geographic variables were included. The wealth index, calculated by ENDES using principal component analysis of household assets and characteristics, was classified into five quintiles: very poor (poorest), poor, middle, rich, and very rich (richest). The area of residence was categorized as urban or rural. Altitude was measured in meters above sea level (masl) and categorized into four levels: 0–499 masl, 500–1499 masl, 1500–2999 masl, and 3000 masl or higher.
Lifestyle variables were also collected. Alcohol consumption was categorized into three groups: never consumed or did not use in the last 12 months, non-excessive alcohol consumption, and excessive alcohol consumption [18]. Smoking status was classified into four categories: has never smoked, former smoker, currently smoker (occasional), and daily smoker.
Finally, blood pressure measurements included systolic and diastolic arterial pressure, measured in millimeters of mercury (mmHg) using standardized protocols with participants in a seated position after at least 5 min of rest.
Anthropometric measurement procedures
Anthropometric measurements were performed by trained personnel from the National Institute of Statistics and Informatics following standardized international protocols. Weight was measured using calibrated electronic scales (precision: 0.1 kg) with participants in light clothing and without shoes. Height was measured using portable stadiometers (precision: 1 mm) following standardized head positioning. WC was measured using an inextensible measuring tape (precision: 0.1 cm) at the midpoint between the lower border of the last palpable rib and the upper iliac crest, with participants standing after normal exhalation. All measurements were performed in duplicate, using the average for analysis. When discrepancies exceeded predefined thresholds (0.5 kg for weight, 1 cm for height or WC), a third measurement was obtained. Quality control procedures included direct supervision and repeated measurements in a subsample of participants to ensure data reliability.
Statistical analysis
All analyses considered the complex sampling design of DHS, using sampling weights, strata, and primary sampling units. The analysis was performed in R (version 4.3.0; R Foundation for Statistical Computing, Vienna, Austria), mainly using the ‘survey’ package [19] for complex design management, ‘gamlss’ [20] for LMS model fitting, ‘gtsummary’ [21] for descriptive analyses, and ‘ggplot2’ [22] for visualizations.
Descriptive statistics weighted by the sampling design were calculated to characterize the study population and examine the distribution of sociodemographic, anthropometric, and clinical variables stratified by sex. Continuous variables are presented as means ± standard deviations (SD), and categorical variables are presented as frequencies and weighted percentages.
The LMS (Lambda-Mu-Sigma) method was used to generate smoothed percentile curves, where L represents the skewness parameter (Box–Cox), M the median, and S the coefficient of variation. Percentiles P5, P10, P25, P50, P75, P85, P90, P95, P97 were calculated for each anthropometric indicator, by five-year age groups and sex. The analysis included prevalence comparisons using international criteria versus population percentiles, concordance analysis between classifications, and an evaluation of regional differences through altitude-stratified subgroup analysis. Validation of LMS method assumptions was performed through Q-Q plots and normality tests, applying appropriate smoothing to avoid artificial oscillations in the curves.
Ethical aspects
This study is based on secondary analysis of de-identified data from the Demographic and Family Health Survey (DHS), which is publicly available through the National Institute of Statistics and Informatics (INEI) of Peru. The original DHS survey received ethical approval from the institutional review board of the implementing agency, and all participants provided informed consent before data collection.
According to international guidelines for research ethics, secondary analyses of completely de-identified, publicly available datasets do not constitute human subjects research requiring additional institutional review board approval when specific criteria are met. The U.S. Code of Federal Regulations (45 CFR 46.102) establishes that analysis of de-identified, publicly available data does not constitute human subjects research as defined by federal regulations [23]. Similarly, the Canadian Tri-Council Policy Statement (TCPS2, Articles 2.2 and 2.4) indicates that research using exclusively publicly available information where individuals have no reasonable expectation of privacy, and secondary use of anonymous information that does not generate identifiable data, does not require ethics review [24].
Our study meets these criteria: (1) the data are fully anonymized without possibility of re-identification; (2) the data are publicly accessible through official government repositories; (3) no data linkage was performed that could enable identification of individuals; (4) the original data collection received ethical approval and informed consent; and (5) the research poses no additional risk to participants. All data management and analysis procedures followed ethical principles for research, including data confidentiality and appropriate use of information.
Results
General characteristics of the study population
The sociodemographic and anthropometric characteristics of the study population are presented in Table 1. The sample comprised 164,513 participants, with 79,821 women (48.52%) and 84,692 men (51.48%), and had a mean age of 44.38 ± 16.72 years (women: 45.36 ± 16.91 years; men: 43.46 ± 16.49 years). The majority of participants resided in urban areas (81.16%) compared to rural areas (18.84%). Regarding educational level, 42.15% had completed secondary education, 36.38% higher education, 21.24% primary education, and only 0.23% had no formal education. The wealth index distribution showed balanced participation among different quintiles, without marked differences between the poorest (18.31%) and the richest (20.45%).
Table 1.
General characteristics of the study population.
| Characteristic | Total, n = 164,513 | Sex | |
|---|---|---|---|
| Women, n = 79,821 (48.52%) | Men, n = 84,692 (51.48%) | ||
| Age, years | 44.38 (16.72) | 45.36 (16.91) | 43.46 (16.49) |
| Residence | |||
| Urban | 133,521 (81.16%) | 65,118 (48.77%) | 68,403 (51.23%) |
| Rural | 30,992 (18.84%) | 14,703 (47.44%) | 16,289 (52.56%) |
| Education | |||
| No Level | 361 (0.23%) | 237 (65.70%) | 124 (34.30%) |
| Primary | 33,667 (21.24%) | 18,647 (55.39%) | 15,020 (44.61%) |
| Secondary | 66,818 (42.15%) | 29,775 (44.56%) | 37,043 (55.44%) |
| Superior | 57,676 (36.38%) | 26,202 (45.43%) | 31,474 (54.57%) |
| Wealth Index | |||
| The poorest | 30,125 (18.31%) | 14,383 (47.75%) | 15,741 (52.25%) |
| Poor | 32,849 (19.97%) | 15,645 (47.63%) | 17,204 (52.37%) |
| Medium | 34,110 (20.73%) | 16,617 (48.72%) | 17,493 (51.28%) |
| Rich | 33,778 (20.53%) | 16,563 (49.03%) | 17,216 (50.97%) |
| Richest | 33,651 (20.45%) | 16,613 (49.37%) | 17,038 (50.63%) |
| Alcohol consumption | |||
| Never or did not use in the last 12 months | 49,203 (29.91%) | 30,987 (62.98%) | 18,215 (37.02%) |
| Non-excessive alcohol consumption | 110,262 (67.02%) | 46,915 (42.55%) | 63,347 (57.45%) |
| Excessive alcohol consumption | 5049 (3.07%) | 1919 (38.01%) | 3130 (61.99%) |
| Body Mass Index, kg/m² | 27.77 (4.80) | 28.31 (5.14) | 27.26 (4.41) |
| Waist Circumference, cm | 93.90 (11.57) | 93.52 (11.77) | 94.27 (11.37) |
| Waist to Height Ratio | 0.60 (0.08) | 0.62 (0.08) | 0.57 (0.07) |
| Systolic arterial pressure, mmHg | 122.97 (19.19) | 118.27 (19.61) | 127.39 (17.69) |
| Diastolic arterial pressure, mmHg | 74.84 (12.12) | 72.61 (11.31) | 76.95 (12.48) |
| Smoking status | |||
| Has never smoked | 134,257 (81.61%) | 74,704 (55.64%) | 59,553 (44.36%) |
| Former smoker | 12,865 (7.82%) | 2350 (18.27%) | 10,515 (81.73%) |
| Currently smoker | 14,723 (8.95%) | 2227 (15.13%) | 12,495 (84.87%) |
| Daily smoker | 2669 (1.62%) | 540 (20.23%) | 2129 (79.77%) |
| Abdominal obesity - ATP III | |||
| Normal | 90,892 (55.25%) | 25,847 (28.44%) | 65,045 (71.56%) |
| Obesity | 73,622 (44.75%) | 53,974 (73.31%) | 19,648 (26.69%) |
| Abdominal obesity - IDF | |||
| No | 52,127 (31.69%) | 9696 (18.60%) | 42,432 (81.40%) |
| Yes | 112,386 (68.31%) | 70,126 (62.40%) | 42,260 (37.60%) |
| Nutritional status | |||
| Normal weight | 48,931 (29.75%) | 21,536 (44.01%) | 27,395 (55.99%) |
| Overweight | 69,355 (42.16%) | 31,999 (46.14%) | 37,356 (53.86%) |
| Obesity | 46,204 (28.09%) | 26,272 (56.86%) | 19,932 (43.14%) |
Mean (SD); n (%).
Sex-specific anthropometric characteristics showed important differences (Table 1). Women presented higher mean BMI (28.31 ± 5.14 kg/m²) compared to men (27.26 ± 4.41 kg/m²), higher WHtR (0.62 ± 0.08 vs 0.57 ± 0.07), but similar WC (93.52 ± 11.77 cm vs 94.27 ± 11.37 cm). Men presented higher systolic blood pressure (127.39 ± 17.69 vs 118.27 ± 19.61 mmHg) and diastolic blood pressure (76.95 ± 12.48 vs 72.61 ± 11.31 mmHg). According to traditional BMI classification, 29.75% of the population presented normal weight, 42.16% overweight, and 28.09% obesity, resulting in a combined prevalence of overweight and obesity of 70.25%. The prevalence of abdominal obesity varied significantly according to the criterion used: 44.75% according to ATP III criteria (≥102 cm in men, ≥88 cm in women) and 68.31% according to IDF criteria (≥90 cm in men, ≥80 cm in women), with women showing substantially higher prevalence than men under both criteria (73.31% vs 26.69% for ATP III; 62.40% vs 37.60% for IDF).
Anthropometric percentile curves by age and sex
Figure 1 presents smoothed percentile curves for the three anthropometric indicators by age and sex. In panel A (BMI), it is observed that all percentiles follow a characteristic pattern of increase from 20 years until reaching maximum values in middle age (40–50 years), followed by a gradual decline in advanced ages. In women, the 50th percentile (median) increases from ~25 kg/m² at 20 years to 29 kg/m² at 45 years, decreasing to 25 kg/m² at 80 years. In men, the pattern is similar, but with consistently lower values, with the median varying between 24 kg/m² in young adults, 27 kg/m² in middle age, and 24 kg/m² in older adults.
Fig. 1. Age- and sex-specific percentile curves for anthropometric indicators in Peruvian adults aged 20–80+ years.
A BMI, B WC, C WHtR ratio. Curves represent 5th, 10th, 25th, 50th, 75th, 85th, 90th, 95th, and 97th percentiles derived from DHS 2015 to 2023 data (n = 385,247).
Panel B (WC) shows similar patterns, with progressive increases until the sixth decade of life. In women, the 50th percentile increases from 85 cm at 20 years to 96 cm at 55–60 years, remaining relatively stable until declining to 88 cm at 80 years. In men, values range from 83 cm in young adults to 95 cm in the sixth decade, declining to 91 cm at advanced ages. Panel C (WHtR) exhibits the same age pattern, with the median in women increasing from 0.55 at 20 years to 0.64 at 55–60 years, while in men it varies from 0.50 to 0.59 in the same age range.
BMI percentiles
BMI percentiles by age and sex are presented in Table 2. In men, the 50th percentile increased from 24.1 kg/m² in the 20–24 years group to a maximum of 27.1 kg/m² in the 35–44 years groups, remaining stable until 49 years and declining progressively to 24.3 kg/m² in the 80+ years group. The 85th percentile, used as a reference point for overweight, varied from 28.8 kg/m² in the youngest group to 31.6 kg/m² in the 50–54 years group, declining to 28.4 kg/m² in those over 80 years.
Table 2.
BMI percentiles by age and sex for Peruvian adults aged 20–80+ years.
| Age group | n | P5 | P10 | P25 | P50 | P75 | P85 | P90 | P95 | P97 |
|---|---|---|---|---|---|---|---|---|---|---|
| Men | ||||||||||
| 20–24 | 6594 | 19.6 | 20.5 | 22.1 | 24.1 | 27 | 28.8 | 30.2 | 32.5 | 34.2 |
| 25–29 | 8971 | 20.6 | 21.6 | 23.3 | 25.7 | 28.6 | 30.4 | 31.5 | 33.8 | 35.3 |
| 30–34 | 10,908 | 21.3 | 22.2 | 24.1 | 26.6 | 29.3 | 31.1 | 32.4 | 34.5 | 35.9 |
| 35–39 | 10,243 | 21.5 | 22.6 | 24.5 | 27.1 | 29.8 | 31.5 | 32.8 | 34.8 | 36.5 |
| 40–44 | 7744 | 21.6 | 22.7 | 24.6 | 27.1 | 29.9 | 31.5 | 32.8 | 34.8 | 36.4 |
| 45–49 | 5679 | 21.6 | 22.6 | 24.5 | 27.1 | 29.9 | 31.5 | 32.8 | 35 | 36.7 |
| 50–54 | 4664 | 21.3 | 22.3 | 24.3 | 27 | 29.9 | 31.6 | 32.7 | 34.9 | 36.2 |
| 55–59 | 4228 | 21 | 22 | 24 | 26.8 | 29.6 | 31.3 | 32.6 | 34.7 | 36.1 |
| 60–64 | 3524 | 20.8 | 21.8 | 23.8 | 26.4 | 29.5 | 31.2 | 32.4 | 34.3 | 35.4 |
| 65–69 | 2971 | 20.4 | 21.4 | 23.3 | 25.8 | 28.9 | 30.4 | 31.7 | 33.5 | 34.9 |
| 70–74 | 2163 | 20.3 | 21.2 | 23.1 | 25.5 | 28.3 | 29.9 | 31.2 | 33.1 | 34.1 |
| 75–79 | 1521 | 19.9 | 20.9 | 22.6 | 24.9 | 27.5 | 29.2 | 30.2 | 31.9 | 33.5 |
| 80+ | 1580 | 19.3 | 20.3 | 22.1 | 24.3 | 26.8 | 28.4 | 29.6 | 31.3 | 32.5 |
| Women | ||||||||||
| 20–24 | 8877 | 19.6 | 20.7 | 22.6 | 25.2 | 28.4 | 30.6 | 32.2 | 34.3 | 36.2 |
| 25–29 | 12,252 | 20.6 | 21.8 | 23.9 | 26.8 | 30 | 31.9 | 33.5 | 36 | 37.7 |
| 30–34 | 12,811 | 21.5 | 22.7 | 24.9 | 27.7 | 30.8 | 32.6 | 34.1 | 36.5 | 38.4 |
| 35–39 | 10,250 | 21.9 | 23.2 | 25.5 | 28.3 | 31.4 | 33.4 | 34.8 | 37.2 | 39 |
| 40–44 | 7248 | 22.1 | 23.5 | 25.9 | 28.8 | 31.9 | 33.7 | 35.1 | 37.6 | 39.3 |
| 45–49 | 4844 | 21.9 | 23.3 | 26 | 28.9 | 32.1 | 34.1 | 35.7 | 38 | 40 |
| 50–54 | 5303 | 21.5 | 22.9 | 25.6 | 28.5 | 31.9 | 34 | 35.5 | 37.9 | 39.6 |
| 55–59 | 4869 | 21 | 22.4 | 25.1 | 28.3 | 31.9 | 33.9 | 35.4 | 37.6 | 39.5 |
| 60–64 | 4240 | 20.7 | 22.2 | 24.9 | 28 | 31.6 | 33.7 | 35.1 | 37.4 | 39.2 |
| 65–69 | 3375 | 20 | 21.3 | 24 | 27.4 | 30.9 | 33 | 34.5 | 37 | 38.8 |
| 70–74 | 2469 | 19.5 | 20.8 | 23.2 | 26.4 | 30.1 | 32.2 | 33.6 | 35.8 | 37.7 |
| 75–79 | 1676 | 19.1 | 20.4 | 22.8 | 25.7 | 29.2 | 31.2 | 32.5 | 35 | 36.7 |
| 80+ | 1779 | 18.7 | 19.8 | 21.9 | 24.8 | 28 | 29.7 | 31.4 | 33.6 | 35.1 |
P5–P97 represent the 5th through 97th percentiles.
Sample sizes (n) represent the number of participants in each age-sex group.
In women, the 50th percentile showed a similar pattern, but with consistently higher values, increasing from 25.2 kg/m² at 20–24 years to 28.9 kg/m² in the 45–49 years group, and decreasing to 24.8 kg/m² in the 80+ years group. The 85th percentile varied between 30.6 kg/m² in the youngest group and 34.1 kg/m² in the 45–49 years group, declining to 29.7 kg/m² in those over 80 years.
The upper percentiles (P90, P95, P97) followed similar patterns, with progressive increases until middle age and subsequent declines. Sex differences were more pronounced in the upper percentiles, with women showing values 2–3 kg/m² higher than men in the same age groups.
WC percentiles
WC percentiles by age and sex are shown in Table 3. In men, the 50th percentile increased from 83.3 cm in the 20–24 years group to 95.5 cm in the 60–64 years group, followed by a gradual decline to 91.3 cm in the 80+ years group. The 85th percentile showed a similar pattern, varying from 95.5 cm in the youngest group to 107.4 cm in the 60–64 years group.
Table 3.
WC percentiles by age and sex for Peruvian adults aged 20–80+ years.
| Age group | n | P5 | P10 | P25 | P50 | P75 | P85 | P90 | P95 | P97 |
|---|---|---|---|---|---|---|---|---|---|---|
| Men | ||||||||||
| 20–24 | 6594 | 71.8 | 73.9 | 77.7 | 83.3 | 91 | 95.5 | 98.9 | 104.2 | 107.5 |
| 25–29 | 8971 | 74.6 | 77 | 81.7 | 88.6 | 95.9 | 99.9 | 103 | 107.6 | 111.3 |
| 30–34 | 10,908 | 76.7 | 79.5 | 84.7 | 91.4 | 98.1 | 102.2 | 105.2 | 110 | 113.4 |
| 35–39 | 10,243 | 78 | 80.8 | 86.6 | 93.1 | 100 | 104 | 107 | 111.5 | 115.2 |
| 40–44 | 7744 | 78.9 | 81.9 | 87.5 | 94.1 | 101 | 105 | 107.8 | 112.3 | 115.9 |
| 45–49 | 5679 | 79 | 82.1 | 88 | 94.7 | 101.3 | 105.5 | 108.5 | 113.2 | 116.8 |
| 50–54 | 4664 | 78.7 | 82 | 88.1 | 95.3 | 102.3 | 106.2 | 108.9 | 113.7 | 117.2 |
| 55–59 | 4228 | 78.3 | 81.7 | 88 | 95.3 | 102.4 | 106.7 | 109.3 | 113.5 | 116.8 |
| 60–64 | 3524 | 78.6 | 81.6 | 87.7 | 95.5 | 103.4 | 107.4 | 110.2 | 114.4 | 117.4 |
| 65–69 | 2971 | 77.3 | 80.5 | 86.8 | 94.3 | 102.2 | 106.5 | 109.2 | 114 | 116.5 |
| 70–74 | 2163 | 77.4 | 80.2 | 86.6 | 93.9 | 101.4 | 105.3 | 108.4 | 113 | 116.5 |
| 75–79 | 1521 | 76 | 79.6 | 85.6 | 92.3 | 99.9 | 104.6 | 107.5 | 111.1 | 114 |
| 80+ | 1580 | 75.5 | 78.4 | 84.2 | 91.3 | 98.6 | 102.6 | 105.6 | 110.2 | 113.7 |
| Women | ||||||||||
| 20–24 | 8877 | 70.5 | 73.3 | 78.4 | 84.8 | 92.1 | 96.5 | 99.9 | 104.7 | 108.4 |
| 25–29 | 12,252 | 73.7 | 77 | 82.2 | 89 | 96.2 | 100.4 | 103.4 | 108.4 | 111.5 |
| 30–34 | 12,811 | 76.5 | 79.8 | 85.2 | 91.5 | 98.7 | 102.6 | 105.5 | 110.2 | 114.1 |
| 35–39 | 10,250 | 77.9 | 81.3 | 87 | 93.4 | 100.5 | 104.6 | 107.4 | 112.2 | 115.7 |
| 40–44 | 7248 | 78.6 | 82 | 87.9 | 94.7 | 101.8 | 105.9 | 108.7 | 113 | 116.3 |
| 45–49 | 4844 | 78 | 81.5 | 88.3 | 95.3 | 102.5 | 106.9 | 109.7 | 114.6 | 118.1 |
| 50–54 | 5303 | 75.7 | 80.6 | 88.1 | 95.5 | 102.8 | 107 | 110.1 | 115 | 118.2 |
| 55–59 | 4869 | 74.2 | 79.6 | 87.4 | 95.5 | 103.5 | 108.2 | 111.2 | 115.8 | 119.1 |
| 60–64 | 4240 | 74.5 | 79.1 | 86.8 | 95.1 | 103 | 107.5 | 110.5 | 115.2 | 118.4 |
| 65–69 | 3375 | 71.3 | 76.7 | 85.2 | 94 | 102.3 | 106.5 | 109.6 | 114 | 117.4 |
| 70–74 | 2469 | 69.3 | 74 | 82.4 | 91.9 | 100.5 | 105 | 108.2 | 112.9 | 115.9 |
| 75–79 | 1676 | 68.2 | 72.5 | 80.7 | 90.3 | 98.6 | 104 | 106.8 | 111.8 | 115 |
| 80+ | 1779 | 68.2 | 71.7 | 78.1 | 87.3 | 96.1 | 100.5 | 103.9 | 108.5 | 111.5 |
P5–P97 represent the 5th through 97th percentiles.
Sample sizes (n) represent the number of participants in each age-sex group.
In women, the 50th percentile increased from 84.8 cm at 20–24 years to 95.5 cm in the 50–59 years groups, remaining relatively stable until 64 years and declining to 87.3 cm in the 80+ years group. The 85th percentile varied between 96.5 cm in the youngest group and 108.2 cm in the 55–59 years group.
Sex differences were less pronounced than in BMI, especially in young ages where values were similar. In middle and advanced ages, women showed slightly higher values in central percentiles, but lower values in upper extreme percentiles.
WHtR percentiles
WHtR percentiles are presented in Table 4. In men, the 50th percentile increased from 0.505 in the 20–24 years group to 0.591 in the 60–64 years group, with subsequent stability. The 85th percentile varied from 0.574 in the youngest group to 0.662 in the 60–64 years group. In the case of women, the 50th percentile showed a more pronounced increase, from 0.554 at 20–24 years to 0.637 in the 55–64 years groups, declining to 0.613 in the 80+ years group. The 85th percentile increased from 0.633 in the youngest group to 0.722 in the 60–64 years group.
Table 4.
WHtR percentiles by age and sex for Peruvian adults aged 20–80+ years.
| Age group | n | P5 | P10 | P25 | P50 | P75 | P85 | P90 | P95 | P97 |
|---|---|---|---|---|---|---|---|---|---|---|
| Men | ||||||||||
| 20–24 | 6594 | 0.433 | 0.447 | 0.471 | 0.505 | 0.549 | 0.574 | 0.595 | 0.622 | 0.643 |
| 25–29 | 8971 | 0.456 | 0.469 | 0.498 | 0.537 | 0.58 | 0.603 | 0.619 | 0.647 | 0.666 |
| 30–34 | 10,908 | 0.47 | 0.487 | 0.516 | 0.555 | 0.593 | 0.617 | 0.634 | 0.659 | 0.681 |
| 35–39 | 10,243 | 0.478 | 0.495 | 0.529 | 0.566 | 0.606 | 0.628 | 0.645 | 0.671 | 0.691 |
| 40–44 | 7744 | 0.486 | 0.503 | 0.536 | 0.574 | 0.613 | 0.635 | 0.651 | 0.677 | 0.7 |
| 45–49 | 5679 | 0.49 | 0.507 | 0.541 | 0.579 | 0.619 | 0.642 | 0.659 | 0.687 | 0.705 |
| 50–54 | 4664 | 0.488 | 0.508 | 0.544 | 0.584 | 0.625 | 0.648 | 0.665 | 0.691 | 0.709 |
| 55–59 | 4228 | 0.489 | 0.508 | 0.546 | 0.587 | 0.63 | 0.654 | 0.671 | 0.695 | 0.715 |
| 60–64 | 3524 | 0.49 | 0.511 | 0.549 | 0.591 | 0.639 | 0.662 | 0.677 | 0.705 | 0.721 |
| 65–69 | 2971 | 0.487 | 0.51 | 0.545 | 0.588 | 0.635 | 0.659 | 0.676 | 0.703 | 0.724 |
| 70–74 | 2163 | 0.493 | 0.512 | 0.547 | 0.59 | 0.635 | 0.659 | 0.679 | 0.706 | 0.723 |
| 75–79 | 1521 | 0.488 | 0.509 | 0.546 | 0.588 | 0.633 | 0.658 | 0.674 | 0.699 | 0.72 |
| 80+ | 1580 | 0.492 | 0.509 | 0.543 | 0.588 | 0.633 | 0.659 | 0.673 | 0.702 | 0.72 |
| Women | ||||||||||
| 20–24 | 8877 | 0.459 | 0.477 | 0.512 | 0.554 | 0.603 | 0.633 | 0.653 | 0.689 | 0.709 |
| 25–29 | 12,252 | 0.481 | 0.503 | 0.538 | 0.584 | 0.631 | 0.658 | 0.677 | 0.71 | 0.729 |
| 30–34 | 12,811 | 0.5 | 0.521 | 0.557 | 0.6 | 0.647 | 0.673 | 0.693 | 0.724 | 0.745 |
| 35–39 | 10,250 | 0.51 | 0.531 | 0.57 | 0.614 | 0.661 | 0.689 | 0.706 | 0.739 | 0.758 |
| 40–44 | 7248 | 0.515 | 0.54 | 0.579 | 0.625 | 0.67 | 0.698 | 0.717 | 0.748 | 0.768 |
| 45–49 | 4844 | 0.515 | 0.541 | 0.583 | 0.631 | 0.68 | 0.708 | 0.728 | 0.757 | 0.78 |
| 50–54 | 5303 | 0.508 | 0.537 | 0.586 | 0.634 | 0.683 | 0.711 | 0.733 | 0.765 | 0.786 |
| 55–59 | 4869 | 0.501 | 0.534 | 0.582 | 0.637 | 0.69 | 0.719 | 0.741 | 0.77 | 0.795 |
| 60–64 | 4240 | 0.504 | 0.533 | 0.582 | 0.637 | 0.691 | 0.722 | 0.743 | 0.775 | 0.796 |
| 65–69 | 3375 | 0.489 | 0.521 | 0.577 | 0.635 | 0.691 | 0.722 | 0.744 | 0.774 | 0.795 |
| 70–74 | 2469 | 0.479 | 0.509 | 0.564 | 0.626 | 0.684 | 0.716 | 0.736 | 0.77 | 0.793 |
| 75–79 | 1676 | 0.473 | 0.504 | 0.557 | 0.624 | 0.68 | 0.715 | 0.738 | 0.772 | 0.786 |
| 80+ | 1779 | 0.48 | 0.506 | 0.554 | 0.613 | 0.669 | 0.702 | 0.726 | 0.752 | 0.767 |
P5–P97 represent the 5th through 97th percentiles.
Sample sizes (n) represent the number of participants in each age-sex group.
Anthropometric percentiles by altitude
BMI percentiles showed systematic differences according to altitude of residence. In men aged 30–34 years, the 50th percentile was 27.5 kg/m² at sea level, 26.3 kg/m² at moderate altitude, and 25.1 kg/m² at high altitude. In women of the same age group, the corresponding values were 27.9 kg/m², 27.3 kg/m², and 26.9 kg/m², respectively. This pattern of decreasing values with higher altitude remained consistent across all age groups and percentiles evaluated (P75, P85, P90, P95, P97).
The differences between extreme altitudes (sea level vs high altitude) in the 50th percentile varied between 2.0 and 3.0 kg/m² in men and 1.5 and 2.5 kg/m² in women, depending on the age group. The upper percentiles (P85, P90, P95) showed similar or slightly greater differences.
WC presented a similar pattern to BMI, with decreasing values at higher altitudes. In men aged 40–44 years, the 50th percentile was 94.9 cm at sea level, 91.9 cm at moderate altitude, and 88.2 cm at high altitude. For women of the same group, the values were 95.7 cm, 93.4 cm, and 91.6 cm, respectively. The differences between sea level and high altitude in the 50th percentile ranged between 4 and 7 cm in men and 3 and 6 cm in women according to age group. In the upper percentiles (P85–P97), the differences remained in similar ranges, with variations of 5–8 cm between extreme altitudes.
The WHtR followed the same decreasing pattern with higher altitude observed in the other indicators. In the 35–39 years group, the 50th percentile in men was 0.577 at sea level, 0.558 at moderate altitude, and 0.543 at high altitude. In women of the same group, the corresponding values were 0.618, 0.602, and 0.592. The differences between extreme altitudes in the 50th percentile varied between 0.030 and 0.050 in men and 0.025 and 0.040 in women according to age. The upper percentiles showed comparable differences, with variations of 0.035–0.055 units between sea level and high altitude.
A continuous altitudinal gradient was observed in the three anthropometric indicators, where moderate altitude values consistently positioned between those of sea level and high altitude. In all cases, the differences between adjacent altitude categories were approximately half of the total difference between extremes. Additionally, the magnitude of altitude differences remained relatively constant across different age groups, although it was slightly greater in middle ages (30–50 years) compared to younger and older groups. The pattern was consistent in both sexes, although absolute differences tended to be slightly greater in men for BMI and WC, while they were similar between sexes for WHtR. Complete altitude-stratified percentile values (P5–P97) for BMI, WC, and WHtR by age and sex are presented in Supplementary Tables S1, S2, and S3, respectively.
Discussion
Main findings
This study presents the first age- and sex-specific anthropometric percentile curves for the Peruvian adult population, demonstrating substantial differences from currently used international cut-off points. The findings reveal distinctive age patterns in the three anthropometric indicators evaluated, with increasing trends from young adulthood to middle age, followed by a gradual decline in advanced ages. This pattern, consistent with observations by Banack et al. [25] in the older Canadian population, suggests that the use of fixed cut-off points across the life cycle does not adequately reflect normal physiological variations associated with aging. Stratification by age and sex is fully justified, considering that women presented consistently higher values in BMI and WHtR, while men showed greater variability in WC, reflecting known differences in adiposity distribution between sexes, as found by other studies [26].
The three anthropometric indicators evaluated in this study have demonstrated robust associations with cardiovascular and metabolic outcomes in diverse populations, which underscores the importance of developing appropriate reference values. BMI, despite its limitations in distinguishing between fat and lean mass, has been consistently associated with all-cause mortality, cardiovascular disease, and type 2 diabetes in prospective studies [27, 28]. However, its predictive capacity varies considerably by age and ethnicity, with some populations showing cardiovascular risk at lower BMI thresholds than those established in Caucasian populations [29]. WC and WHtR, as indicators of central adiposity, have shown superior predictive capacity for cardiometabolic risk compared to BMI alone in multiple studies [30, 31]. WC has been independently associated with metabolic syndrome, type 2 diabetes, and cardiovascular events, even after adjusting for BMI [30, 32]. The WHtR has emerged as a particularly useful screening tool due to its simplicity and strong association with cardiometabolic risk factors independent of age and sex, with meta-analyses suggesting it may be a better predictor of cardiovascular events than WC or BMI alone [7, 33]. Importantly, recent evidence from Peruvian workers demonstrated that population-specific percentiles of WC and WHtR showed superior predictive capacity for myocardial infarction and stroke compared to conventional ATP III, IDF, or WHtR ≥0.5 cut-offs [34]. This finding reinforces the clinical utility of developing population-specific references that better capture the relationship between anthropometric measures and cardiovascular risk in specific ethnic and geographic contexts.
The age patterns observed in our study reveal specific characteristics that support the need for population references. BMI showed the most marked pattern of decline in advanced ages, particularly after age 70, coinciding with international evidence on body composition changes during aging [34, 35]. In contrast, WC maintained relatively stable values until very advanced ages, while WHtR showed progressive increase with age, suggesting that these indicators may be more appropriate for assessing cardiometabolic risk in older adults. These findings are consistent with previous studies that have documented the redistribution of body fat toward the abdominal region with aging, independent of changes in total body weight [36, 37]. The persistence of elevated WHtR values at advanced ages could explain why this indicator has shown greater association with cardiovascular risk in middle-aged and older populations [33].
The altitude analysis reveals clinically significant differences that cannot be ignored in the context of a geographically diverse country like Peru. The observed differences between populations at sea level versus high altitude (≥3000 m) were 2.5–3.0 kg/m² for BMI, 4–6 cm for WC, and 0.04–0.06 units for WHtR, magnitudes sufficient to modify risk classifications. These differences are consistent with previous studies that have documented specific physiological adaptations in high-altitude populations, including changes in body composition, basal metabolism, and body fat distribution [38, 39]. While we leave to the reader’s discretion the decision on implementing altitude-specific percentiles, it is undeniable that altitude plays an important role in determining normal anthropometric values. Andean populations have developed unique adaptations over millennia, which is reflected in anthropometric profiles that do not conform to standards developed in lower altitude populations [40].
The high prevalence of abdominal obesity observed when applying international criteria (44.75% for ATP III, 68.31% for IDF) in the Peruvian population raises important questions about the appropriate reference framework for anthropometric assessment. While these prevalences may reflect genuine public health concerns, the population-specific percentiles developed in this study provide a complementary tool for contextualizing individual anthropometric values within the Peruvian distribution. As Banack et al. [25] point out, the use of fixed cut-off points assumes that risk thresholds are constant for all adults regardless of age, sex, or population characteristics. Importantly, recent evidence from Peruvian workers demonstrates that population-specific percentiles (P75 and P95) showed superior predictive capacity for cardiovascular events compared to conventional cut-off points [41]. Vera-Ponce et al. [8] found that population percentiles of WC and WHtR were better predictors of myocardial infarction and stroke than ATP III, IDF, or the universal WHtR ≥0.5 criteria in this population. This finding suggests that population-specific percentiles, once validated against cardiometabolic outcomes in larger and more diverse samples, may offer improved risk stratification for the Peruvian population while still maintaining international cut-offs as valuable screening tools.
Public health implications
The implementation of anthropometric percentiles specific to the Peruvian population could provide valuable complementary tools for cardiometabolic risk assessment and public health planning. First, these percentiles offer a descriptive framework for understanding where individuals fall within their reference population distribution, which can inform clinical interpretation alongside established risk-based cut-offs. This approach allows health professionals to contextualize anthropometric measurements more appropriately, particularly in cases where values fall near conventional thresholds. Second, population-specific percentiles provide a foundation for future validation studies that can determine optimal risk thresholds for the Peruvian population specifically, as demonstrated by Vera-Ponce et al. [41] in occupational settings.
Additionally, the adoption of age- and sex-specific percentiles would facilitate longitudinal patient monitoring, allowing health professionals to identify significant changes in anthropometric status relative to the reference population. This is particularly relevant in the context of population aging, where changes in body composition are normal, but must be differentiated from pathological changes. Percentiles also provide a more effective educational tool for patient communication, as it is more understandable to explain that an individual is at the 90th percentile for their age and sex, than to use labels like “obesity” that may generate stigma.
These findings also have implications for the design of public health policies and prevention programs. Prevention strategies could be informed by both international risk-based criteria and population-specific percentiles, allowing for a more nuanced approach to risk assessment. This is especially important in the context of community interventions, where understanding population distribution patterns can help identify individuals who may benefit most from targeted interventions, where accurate identification of at-risk individuals is crucial for program success. Additionally, consideration of altitude differences could inform the development of public health strategies specific to Andean populations, recognizing their unique physiological characteristics.
Finally, the results of this study support the need to develop specific anthropometric references for other Latin American populations with similar ethnic and geographic characteristics. Recent evidence suggests considerable variability in anthropometric distributions across Latin American countries [8], and the development of population-specific percentiles could provide valuable reference frameworks for each context. Ethnic diversity, altitudinal variations, and differences in dietary and physical activity patterns characterize many Latin American countries, suggesting that the population percentiles approach could be widely applicable in the region.
Limitations
This study presents several limitations that must be considered in interpreting the results. First, although percentiles appropriately describe population distribution, longitudinal studies are required to determine their predictive capacity for cardiovascular and metabolic events in the Peruvian population specifically. Second, although the sample is nationally representative, Peru’s ethnic diversity is not stratified in the analyses, which could theoretically affect the applicability of these percentiles in specific ethnic subgroups. However, several practical considerations limit the feasibility and utility of ethnic-specific stratification. Moreover, ethnic classification in Peru is complex and largely subjective, based on self-identification that varies by context and may not reflect biological or genetic ancestry. Developing ethnic-specific percentiles would require not only adequate sample sizes for each ethnic group but also validated and objective classification methods, which remain methodologically challenging. While different ethnic groups may have distinct anthropometric profiles due to genetic, dietary, and lifestyle factors, excessive stratification risks creating overly specific references with limited practical utility. The altitude-stratified percentiles presented in this study (Supplementary Tables S1–S3) provide a pragmatic alternative, as altitude represents an objective, easily measurable proxy that captures meaningful environmental and lifestyle variations across Peru’s geography. This approach balances the need for population-appropriate references with practical applicability in clinical and public health settings. Finally, the study period (2015–2023) may not reflect secular changes in anthropometric patterns, and its extrapolation to other Latin American countries requires specific validation in each population context.
Conclusions and recommendations
This study establishes the first specific anthropometric references for the Peruvian adult population, demonstrating substantial differences from currently used international criteria. Age- and sex-specific percentiles more appropriately reflect normal physiological variations in the population, while the observed differences by altitude suggest the importance of considering environmental factors in anthropometric assessment. The discrepancy between prevalences estimated with international criteria versus population percentiles evidences the urgent need for specific references to avoid systematic overestimation of cardiometabolic risk in the Peruvian population.
We recommend the adoption of the population-specific percentiles presented in this study as complementary reference tools alongside established international criteria. These age- and sex-specific percentiles provide a valuable framework for contextualizing individual anthropometric measurements within the Peruvian population distribution. While international cut-offs (ATP III, IDF, WHtR ≥0.5) remain important as screening tools derived from cardiovascular risk studies, population percentiles offer additional information about relative standing within the reference population. The superior predictive capacity of population-specific percentiles suggests that future validation studies against cardiometabolic outcomes in larger, nationally representative samples could establish Peru-specific risk thresholds. Until such validation is completed, we recommend that researchers and health professionals use both approaches: international cut-offs for risk screening based on established evidence, and population percentiles for contextualizing measurements within the Peruvian distribution. These percentiles can be implemented through digital tools that facilitate their calculation in clinical practice, contributing to a more comprehensive assessment of nutritional status and cardiometabolic risk in the Peruvian population.
Supplementary information
Acknowledgements
Special thanks to the members of Universidad Nacional Toribio Rodríguez de Mendoza de Amazonas (UNTRM), Amazonas, Peru, for their support and contributions throughout the completion of this research.
Author contributions
Víctor Juan Vera-Ponce: Conceptualization, Investigation, Formal analysis, Methodology, Resources, Writing - Original Draft, Writing - Review & Editing. Jhosmer Ballena-Caicedo: Conceptualization, Investigation, Methodology, Writing - Original Draft, Writing - Review & Editing. Lupita Ana Maria Valladolid-Sandoval: Methodology, Data Curation, Writing - Original Draft, Writing - Review & Editing. Fiorella E. Zuzunaga-Montoya: Software, Investigation, Data Curation, Methodology, Writing - Original Draft, Writing - Review & Editing. Rossmery Leonor Poemape Mestanza: Investigation, Methodology, Resources, Writing - Original Draft, Writing - Review & Editing. Marcos Francisco Garavito Castillo: Methodology, Software, Writing - Original Draft, Writing - Review & Editing. Carmen Inés Gutierrez De Carrillo: Formal analysis, Funding acquisition, Supervision, Writing - Original Draft, Writing - Review & Editing
Funding
This study was funded by the Vicerectorado de Investigación of the Universidad Nacional Toribio Rodríguez de Mendoza de Amazonas.
Data availability
Data are available upon request to the corresponding author.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
The online version contains supplementary material available at 10.1038/s41366-026-02035-0.
References
- 1.Després J-P, Lemieux I. Abdominal obesity and metabolic syndrome. Nature. 2006;444:881–7. 10.1038/nature05488. [DOI] [PubMed] [Google Scholar]
- 2.Lear SA, James PT, Ko GT, Kumanyika S. Appropriateness of waist circumference and waist-to-hip ratio cutoffs for different ethnic groups. Eur J Clin Nutr. 2010;64:42–61. 10.1038/ejcn.2009.70. [DOI] [PubMed] [Google Scholar]
- 3.Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA. 2001;285:2486–97. 10.1001/jama.285.19.2486. [DOI] [PubMed] [Google Scholar]
- 4.International Diabetes Federation. IDF Diabetes Atlas. 10th ed. Brussels: International Diabetes Federation; 2021. https://diabetesatlas.org/idfawp/resource-files/2021/07/IDF_Atlas_10th_Edition_2021.pdf.
- 5.Ardern CI, Janssen I, Ross R, Katzmarzyk PT. Development of health-related waist circumference thresholds within BMI categories. Obes Res. 2004;12:1094–103. 10.1038/oby.2004.137. [DOI] [PubMed] [Google Scholar]
- 6.Lean ME, Han TS, Morrison CE. Waist circumference as a measure for indicating need for weight management. BMJ. 1995;311:158–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Browning LM, Hsieh SD, Ashwell M. A systematic review of waist-to-height ratio as a screening tool for the prediction of cardiovascular disease and diabetes: 0.5 could be a suitable global boundary value. Nutr Res Rev. 2010;23:247–69. 10.1017/S0954422410000144. [DOI] [PubMed] [Google Scholar]
- 8.Vera-Ponce VJ, Loayza-Castro JA, Zuzunaga-Montoya FE, Vásquez-Romero LEM, Sanchez-Tamay NM, Bustamante-Rodríguez JC, et al. Abdominal obesity prevalence in Latin America: a systematic review and meta-analysis comparing ATP III and IDF criteria. Front Endocrinol. 2025;16:1562060 10.3389/fendo.2025.1562060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Vásquez-Romero LEM, Zuzunaga-Montoya FE, Loayza-Castro JA, Vigil-Ventura E, Ramos W, Vera-Ponce VJ. Prevalence of obesity according to body mass index, waist circumference, and waist-to-height ratio in Peru: a systematic review and meta-analysis. Obesity Pillars. 2025;13:100166 10.1016/j.obpill.2025.100166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bernabé-Ortiz A, Carrillo-Larco RM, Gilman RH, Checkley W, Smeeth L, Miranda JJ, et al. Contribution of modifiable risk factors for hypertension and type-2 diabetes in Peruvian resource-limited settings. J Epidemiol Community Health. 2016;70:49–55. 10.1136/jech-2015-205988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Instituto Nacional de Estadística e Informática. Censos Nacionales 2017: XII de Población, VII de Vivienda y III de Comunidades Indígenas. Lima: INEI; 2018.
- 12.Woolcott OO, Castillo OA, Gutierrez C, Elashoff RM, Stefanovski D, Bergman RN. Inverse association between diabetes and altitude: a cross-sectional study in the adult population of the United States. Obesity. 2014;22:2080–90. 10.1002/oby.20800. [DOI] [PMC free article] [PubMed]
- 13.Zhang B, Xu R, Shi Z, Wang T, Jian W, Wang H, et al. Associations of abdominal obesity-related dietary patterns with prediabetes and type 2 diabetes: exploring the mediating effects of body composition and altitude in Tibetan adults. Public Health Nutr. 2025;1–31. 10.1017/S1368980025100724. [DOI] [PMC free article] [PubMed]
- 14.Zila-Velasque JP, Grados-Espinoza P, Challapa-Mamani MR, Sánchez-Alcántara F, Cedillo-Balcázar J, Cs A-D, et al. Prevalence of metabolic syndrome and its components according to altitude levels: a systematic review and meta-analysis. Sci Rep. 2024;14:27581. 10.1038/s41598-024-77928-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Maxfield A, Hadley C, Hruschka DJ. The relationship between altitude and BMI varies across low- and middle-income countries. Am J Hum Biol. 2024;36:e24036 10.1002/ajhb.24036. [DOI] [PubMed] [Google Scholar]
- 16.Centers for Disease Control and Prevention. Growth Charts - CDC Growth Charts. 2024. https://www.cdc.gov/growthcharts/cdc-growth-charts.htm.
- 17.Instituto Nacional de Estadística e Informática. Perú: Encuesta Demográfica y de Salud Familiar, Endes 2023. Lima: INEI; 2024. https://www.gob.pe/institucion/inei/informes-publicaciones/5601739-peru-encuesta-demografica-y-de-salud-familiar-endes-2023.
- 18.Wechsler H, Nelson TF. Binge drinking and the American college student: what’s five drinks? Psychol Addict Behav. 2001;15:287–91. 10.1037/0893-164x.15.4.287. [DOI] [PubMed] [Google Scholar]
- 19.Lumley T, Gao P, Schneider B. Survey: analysis of complex survey samples. 2025. https://cran.r-project.org/web/packages/survey/index.html.
- 20.Rigby RA, Stasinopoulos DM. Generalized additive models for location, scale and shape. J R Stat Soc Ser C Appl Stat. 2005;54:507–54. 10.1111/j.1467-9876.2005.00510.x. [Google Scholar]
- 21.Sjoberg DD, Whiting K, Curry M, Lavery JA, Larmarange J. Reproducible summary tables with the gtsummary Package. R J. 2021;13:570 10.32614/RJ-2021-053. [Google Scholar]
- 22.Wickham H. ggplot2. Cham: Springer International Publishing; 2016. 10.1007/978-3-319-24277-4.
- 23.U.S. Department of Health and Human Services. Code of Federal Regulations, Title 45 Public Welfare, Part 46 Protection of Human Subjects (45 CFR 46). Washington, DC: U.S. Government Publishing Office; 2018. [PubMed]
- 24.Canadian Institutes of Health Research, Natural Sciences and Engineering Research Council of Canada, Social Sciences and Humanities Research Council of Canada. Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (TCPS2). Ottawa: Government of Canada; 2018.
- 25.Banack HR, Kim CD, Cook CE, Wasser A, Kaufman JS, Stovitz SD. BMI-for-age percentile curves for older adults. Obesity. 2025. 10.1002/oby.24189. [DOI] [PMC free article] [PubMed]
- 26.Blaak EE, Goossens GH. Metabolic phenotyping in people living with obesity: implications for dietary prevention. Rev Endocr Metab Disord. 2023. 10.1007/s11154-023-09830-4. [DOI] [PMC free article] [PubMed]
- 27.Di Angelantonio E, Bhupathiraju S, Wormser D, Gao P, Kaptoge S, Berrington de Gonzalez A, et al. Body-mass index and all-cause mortality: individual-participant-data meta-analysis of 239 prospective studies in four continents. Lancet. 2016;388:776–86. 10.1016/S0140-6736(16)30175-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Aune D, Sen A, Prasad M, Norat T, Janszky I, Tonstad S, et al. BMI and all cause mortality: systematic review and non-linear dose-response meta-analysis of 230 cohort studies with 3.74 million deaths among 30.3 million participants. BMJ. 2016;353:i2156 10.1136/bmj.i2156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet. 2004;363:157–63. 10.1016/S0140-6736(03)15268-3. [DOI] [PubMed] [Google Scholar]
- 30.Ross R, Neeland IJ, Yamashita S, Shai I, Seidell J, Magni P, et al. Waist circumference as a vital sign in clinical practice: a Consensus Statement from the IAS and ICCR Working Group on Visceral Obesity. Nat Rev Endocrinol. 2020;16:177–89. 10.1038/s41574-019-0310-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lee CMY, Huxley RR, Wildman RP, Woodward M. Indices of abdominal obesity are better discriminators of cardiovascular risk factors than BMI: a meta-analysis. J Clin Epidemiol. 2008;61:646–53. 10.1016/j.jclinepi.2007.08.012. [DOI] [PubMed] [Google Scholar]
- 32.Zhang C, Rexrode KM, van Dam RM, Li TY, Hu FB. Abdominal obesity and the risk of all-cause, cardiovascular, and cancer mortality: sixteen years of follow-up in US women. Circulation. 2008;117:1658–67. 10.1161/CIRCULATIONAHA.107.739714. [DOI] [PubMed] [Google Scholar]
- 33.Ashwell M, Gunn P, Gibson S. Waist-to-height ratio is a better screening tool than waist circumference and BMI for adult cardiometabolic risk factors: systematic review and meta-analysis. Obes Rev. 2012;13:275–86. 10.1111/j.1467-789X.2011.00952.x. [DOI] [PubMed] [Google Scholar]
- 34.Villareal DT, Apovian CM, Kushner RF, Klein S. American Society for Nutrition, NAASO, The Obesity Society. Obesity in older adults: technical review and position statement of the American Society for Nutrition and NAASO, The Obesity Society. Am J Clin Nutr. 2005;82:923–34. 10.1093/ajcn/82.5.923. [DOI] [PubMed] [Google Scholar]
- 35.St-Onge M-P, Gallagher D. Body composition changes with aging: the cause or the result of alterations in metabolic rate and macronutrient oxidation? Nutrition. 2010;26:152–5. 10.1016/j.nut.2009.07.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Hughes VA, Roubenoff R, Wood M, Frontera WR, Evans WJ, Fiatarone Singh MA. Anthropometric assessment of 10-y changes in body composition in the elderly. Am J Clin Nutr. 2004;80:475–82. 10.1093/ajcn/80.2.475. [DOI] [PubMed] [Google Scholar]
- 37.Schwartz AV, Kelsey JL, Maggi S, Tuttleman M, Ho SC, Jónsson PV, et al. International variation in the incidence of hip fractures: cross-national project on osteoporosis for the World Health Organization Program for Research on Aging. Osteoporos Int. 1999;9:242–53. 10.1007/s001980050144. [DOI] [PubMed] [Google Scholar]
- 38.Beall CM. Two routes to functional adaptation: Tibetan and Andean high-altitude natives. Proc Natl Acad Sci USA. 2007;104:8655–60. 10.1073/pnas.0701985104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Huicho L, Pawson IG, León-Velarde F, Rivera-Chira M, Pacheco A, Muro M, et al. Oxygen saturation and heart rate in healthy school children and adolescents living at high altitude. Am J Hum Biol. 2001;13:761–70. 10.1002/ajhb.1122. [DOI] [PubMed] [Google Scholar]
- 40.Pawson IG, Huicho L. Persistence of growth stunting in a Peruvian high altitude community, 1964-1999. Am J Hum Biol. 2010;22:367–74. 10.1002/ajhb.21001. [DOI] [PubMed] [Google Scholar]
- 41.Vera-Ponce VJ, Zuzunaga-Montoya FE, Vasquez-Romero LEM, Loayza-Castro JA, Valladolid-Sandoval LAM, Ballena-Caicedo J, et al. Anthropometric markers as cardiovascular predictors: a comparison between conventional cut-off points and population percentiles. J Endocrinol Metabol. 2025;14:24–33. 10.14740/jem1054. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available upon request to the corresponding author.