ABSTRACT

Objective

Turkish growth reference charts are based on 1989-2002 data. Globally, positive secular trends in height have been observed, and updating growth charts every 20 years is recommended. Additionally, obesity is a rising health issue worldwide. This study investigates if there has been a further increase in young Turkish adults’ mean height and body mass index (BMI) compared to previous national data (TK2002) and Turkish-origin young adults in the Netherlands. It also explores the association between adult height and BMI with socioeconomic status (SES) and geographical region.

Methods

This cross-sectional study (2023-2024) included 217 females and 248 males, aged 18-26, voluntarily recruited randomly selected from İstanbul University, representing all regions of Türkiye. Height, weight, and SES were recorded. The top two SES groups were combined for analysis.

Results

Sample distribution aligned with Türkiye’s regional population distribution. Mean height was 1.8 cm taller in females (p=0.003) and males (p<0.001) compared to TK2002, and also taller (2.3 and 0.5 cm, respectively, p<0.001 and p=0.03) than in Netherlands measured in 2009 (NL2009). BMI was significantly higher in males than in TK2002 and NL2009 (p<0.001).

Conclusion

Final height of Turkish students increased by 1.8 cm in both sexes over two decades. Males’ BMI was alarmingly high (58% overweight or obese). A population growth study to generate updated growth charts from birth to young adulthood and prevention programs to reduce obesity are needed.

Introduction

Anthropometric measurements play a central role in tracking children’s growth and development. Growth references are benchmarks for assessing children’s health and comparing their growth with peers. A population’s height is influenced by environmental factors, particularly nutrition and socioeconomic status (SES) (1), which can result in a so-called “secular trend”. Therefore, up-to-date national references for height assessment are needed (2). Body mass index (BMI) is even more strongly affected by environmental influences, as illustrated by the obesity epidemic in the last five decades (3).

The first growth references in Türkiye (1968-1970) were collected in 0-18 year-old youngsters from families with high SES living in İstanbul (4). The subsequent study (TK2002) included 6-18 year-old students attending primary and secondary schools located in six districts of İstanbul city in 1989-2002 and showed a mean height increase of 1.4 cm and 2.7 cm in 17-year-old males and females, respectively. Notably, height differences were compared using data from 17-year-olds, as the first study lacked sufficient height data for the 18-year-old group (5). Mean weight increased by 4.5 and 1.5 kg in 18-year-old males and females, respectively.

It has been recommended that growth references should be reassessed every 5-10 or 15-20 years for populations where large (≥1 cm/decade) or little secular change is expected, respectively (6). Before embarking on an expensive nation-wide growth study, we performed a pilot study in young adults to determine whether the positive secular trend of height has continued and to assess current BMI. While a positive secular change of height is generally considered a positive indicator of a population’s health status, a BMI increase is associated with negative consequences of overweight and obesity for individuals and society (3). Young adulthood can be considered a suitable life stage to assess final height and BMI, as it reflects the cumulative outcomes of childhood growth and nutrition.

The primary aim of this study was to test our hypothesis that mean height and BMI of young Turkish adults is higher compared with previous national data (TK2002). The secondary aim was to compare current height and BMI of young Turkish adults with those of young adults of Turkish origin living in the Netherlands measured in 2009 (NL2009) (7, 8, 9). Thirdly, we aimed to explore the association between final height and BMI with SES and geographical region.

Methods

Study Setting

A prior power analysis (95% power) indicated that at least 122 subjects of each sex were needed to detect a 1.8 cm height increase (p=0.05) vs TK2002. Sample size calculation was performed using G-Power 3.1.3 (Faul, University of Kiel, Kiel, Germany). From 65,000 students at İstanbul University, originating from every region of Türkiye in 2023-2024, we recruited a sample of 529 volunteer students. Inclusion criteria for the study were: (a) age between 18 and 26 years; (b) being a student at İstanbul University; and (c) volunteering to participate in the study. Exclusion criteria were: (a) diagnosis of a chronic disease; (b) history of surgery that can affect linear growth; (c) diagnosis of growth retardation; (d) history of growth hormone administration; (e) having a parent born outside of Türkiye; (f) birth length or weight less than -2 standard deviation score (SDS); and (g) current height or BMI less than -3 SDS or greater than +3 SDS (Figure 1). The students’ medical history, birthplace, high school graduation province, parental education level and occupation and household income were recorded through a questionnaire.

Measurements

Height was measured with bare feet using a portable stadiometer (SECA^TM 213, Hamburg, Germany) and documented to the nearest 0.1 cm. Height measurements were repeated twice, and the mean value was calculated. If the difference exceeded 0.3 cm, a third measurement was taken, and the average of the two closest values was used. Individuals were weighed wearing light clothing using a digital scale (SECA^TM 813, Hamburg, Germany), recorded to the closest 0.1 kg. BMI was calculated as kg/m². Two trained technicians performed all measurements.

Classification of SES

After collecting all data, the participants were classified into four groups according to SES, based on educational level of both parents and father’s occupation. The SES classification ranged from SES 1 (highest scoioeconomic group) to SES 4 (lowest). For example, SES 1 included individuals with university-educated parents and professional occupations, while SES 4 included those with limited education and unskilled jobs (Supplementary Material 1) (7, 8, 9). Compared with the TK2002 questionnaire, the educational level of parents was increased by one level because of the 26% average increase in the duration of education in Türkiye between 2008 and 2022 (10). Since height and BMI were not different between the top two SES groups in previous (4, 5) nor in the current study (all p>0.2), students from both classes (151 females and 157 males) were combined for this analysis.

Geographical Subgroup Analysis

Geographical subgroups were defined in two different ways: based on the participant’s birthplace and based on the region where they completed high school. Each participant was categorized into one of the seven official geographical regions of Türkiye (1: Marmara, 2: Aegean, 3: Mediterranean, 4: Black Sea, 5: Central Anatolia, 6: Eastern Anatolia, 7: Southeastern Anatolia). Analyses were conducted separately for both definitions, and corresponding results were reported in tables and Supplementary Materials 2 and 3.

Comparative Analyses

Height and BMI of the present study population were compared with those of the TK2002 study, in which data for 18-year-olds were collected from measurements of senior high school students in six affluent districts of İstanbul. The technical specifications and measurement methods of the instruments used in this study are identical to those used in TK2002, ensuring comparability and consistency across both studies. Final height was compared with 21-year-olds of Turkish origin living in the NL2009, while BMI was compared with 18-year-olds from the same cohort (7, 8, 9).

Statistical Analysis

Statistical analyses were performed using SPSS, version 29.0 (IBM Corp., Armonk, NY, USA). Normality was assessed using the Shapiro-Wilk test. Descriptive data are presented as frequencies, percentages, means ± SDs, or ranges. Normally distributed data were compared using Student’s t-tests, while non-normally distributed data were analysed with the Mann-Whitney U test. To investigate the association between the geographical regions and height and BMI, a Multivariate Analysis of Variance (MANOVA) was performed. Multiple group comparisons were conducted using one-way ANOVA. For significant ANOVA results, Levene’s test for homogeneity of variances and Tukey’s honestly significant difference test for post-hoc comparisons were used. Age, SES, and geographical region were used to conduct subgroup analyses. Age groups were divided into 18-22 and 23-26 years.

Results

The study included 465 young adults (217 females, 248 males) (1). Table 1 shows the results of anthropometric measurements for SES groups. In females, mean height tended to decrease with lower SES (p=0.4) but no such pattern was noted for weight and BMI. In males, mean height of the SES 4 group was lower (by 4 cm) than that of the other three SES groups (p=0.002), weight and BMI tended to decrease with lower SES. Data obtained from the SES 1 and 2 groups were used for further analyses.

Supplementary Materials 3 and 4 show the percentages of the study sample for region of birth and of high school completion, respectively, in comparison with Türkiye’s overall population distribution (11). In general, the regional distribution of our sample is consistent with that of the Turkish population, although percentages of individuals born in the Marmara and Mediterranean regions are slightly over-represented and Aegean and Central Anatolia regions slightly under-represented.

Table 2 presents mean height, weight and BMI in the high SES group by region. In females, birthplace was not associated with height and BMI [MANOVA, Wilks’ Lambda, F(12, 298)=1.572, p=0.099], in contrast to the region of high school completion [Wilks’ Lambda, F(12, 288)=2.120, p=0.016]. Further analysis revealed an effect of the region of high school completion on height [F(6, 150)=2.91, p=0.005] but not on BMI [F(6, 150)=0.97, p=0.412]. Post-hoc analysis showed differences in height between Marmara and Central Anatolia regions (p=0.011) and between Central Anatolia and Eastern Anatolia (p=0.04). In males, results indicated no association between birthplace or of high school completion and height and BMI [Wilks’ lambda, F(12, 298)=1.572, p=0.099 and Wilks’ Lambda, F(12, 298)=0.933, p=0.514, respectively].

Supplementary Material 4 shows mean ± SD of anthropometric measurements in the high SES group according to two age groups (18-22 years versus 19-26 years). No notable differences were observed, but BMI in males showed a trend of increasing with age (p=0.12).

Table 3 and Figures 2a and 2b show 3-10-25-50-75-90-97^th percentiles of TK2024, TK2002 and NL2009 for height in both sexes. Mean height was 1.8 cm taller in females (p=0.003) and males (p<0.001) compared to TK2002.

Table 4 and Figures 2c and 2d show these percentiles for BMI versus earlier growth studies. In females, BMI was similar to TK2002 and lower than NL2009 (p<0.001). BMI was higher in males than in TK2002 and NL2009 (p<0.001).

Discussion

Our study shows a notable increase in mean height of young adult males and females by 1.8 cm over two decades, showing a continuing positive secular trend in both sexes. Compared to previous national growth studies, secular growth change continues at a similar rate (4, 5). Notably, males show a concerning increase in BMI, with 58% classified as overweight or obese, while females’ BMI has remained stable. Compared to the 2003 Turkey Demographic Health Survey (TDHS), the weight-for-height ratio of 0-5 year-olds increased in the 2018 TDHS, suggestive of an increasing BMI in childhood (18, 19). Regional variations in height were present among females but not males. We speculate that the ongoing secular trend of height is associated with improved living and nutritional conditions, even in the best-off segments of the population, and improved child health and immunization programs.

İstanbul (population>15 millions) has the highest number of migrants from all regions of Türkiye, predominantly young individuals aged 20 to 24 moving for educational purposes (12). Previous national growth reference studies used data from children in affluent neighbourhoods of İstanbul (4, 5, 13). İstanbul University, with a student body of 63,790 from across Türkiye, accurately reflects the country’s demographic composition (14). The alignment between our sample distribution and Türkiye’s regional demographics ensures the generalizability of our findings. The only regional height variation was observed among females when classified by the region where they completed high school (shorter height in Central Anatolia compared with Marmara and Eastern Anatolia), consistent with a 2007-2008 study in Kayseri, Central Anatolia (15).

The main factors that influence the assessment of secular change of final height are the SES, geography and age of the male study sample. While SES and geography of TK2002 and TK2024 were comparable, the age of measurement of males was 18 and 18-26 years, respectively. Since in the NL2009 study mean male height increased by 0.5 cm between 18 and 21 years (9), we cannot exclude that the unbiased secular trend in males is 1.3 cm. Current mean height of Turkish young adult males is also taller than that of offspring of Turkish immigrants living in the Netherlands 15 years earlier (NL2009) (8), probably mainly associated with the time interval.

In developing countries, it is advised to exclude malnourished children of low socioeconomic backgrounds from population growth studies, as they may not accurately represent the overall population’s growth potential. We therefore included students with higher SES in the study, in line with the previous national growth references (TK2002) and WHO growth standards (5, 21). This “prescriptive” or “normative” approach aims to create references that can be used to identify inappropriate growth rather than to define the current population averages (6). In the present study, the mean height of males in the lowest SES group was lower than that of the other SES groups. In addition, a trend of a height decrease from high to low SES was observed in females. We realize that this approach has disadvantages, for example that it obscures disparities within the country, overestimates the “real” mean final height and secular trend (if mean height in the highest SES cohort increases more than in lower SES groups) or underestimates it if the opposite phenomenon would occur. Furthermore, it causes a bias if the height of the Turkish population is compared with that of other countries. 

A very concerning result of our study is the increase in male BMI (58% overweight or obese). Male BMI had increased compared to TK2002 and the NL2009 data (7, 13), indicating that the Turkish population is at risk of a severe obesity epidemic that starts at least in young adulthood, but likely even before this. Factors such as sedentary lifestyles and unhealthy dietary patterns may contribute to this increase. The two TDHSs (2013 and 2018) in 0-5 year-olds (19, 22) reported overweight prevalences of 11 and 8%, respectively, while the prevalence peaked at 6-36 months and 12-17 months of age, respectively. For future BMI reference charts, it is important to present normative data derived from growth studies performed before the overweight epidemic, as done in the current USA, WHO and Dutch charts (16, 23, 24, 25, 26).

Interestingly, while males showed a marked increase in BMI, the BMI of females remained relatively stable over the same period. This sex-specific difference may be influenced by societal pressure regarding body image and stronger motivation for weight control among females, as reported in previous studies (27). Moreover, females often show healthier dietary preferences than males (28). These behavioral differences might partly explain the stable BMI trend observed in females in our study.

When the sample was divided into age groups (18-22 years and 23-26 years), no differences were found in height. This suggests that the chosen age range of young adults for assessing final height was appropriate and did not introduce bias related to age within the specified range. BMI in males showed a non-significant age-related increase (p=0.12); however, this finding should be interpreted cautiously and does not indicate a strong trend.

Our study has notable strengths, such as using a clearly defined sample from İstanbul University, which includes participants from all regions of Türkiye. Thus, we believe that our findings can be generalized to the Turkish population. Moreover, using standardized anthropometric measurements and detailed statistical analyses ensures the reliability and accuracy of the data. Furthermore, the SES questionnaire used in this study was based on scales from previous national studies, ensuring comparability and relevance for the analysis. This methodological consistency allows for reliable comparisons to previous research, strengthening the validity of our findings. The study’s design meets the criteria recommended by Waterlow and recognized by the World Health Organization. These criteria necessitate that the reference population be adequately nourished, the sampling procedure be clearly defined and reproducible, the sample size be sufficient, the measurements be relevant and high quality, and the data be appropriately processed (21, 29).

Study Limitations

A potential limitation of our study is that the sample consists only of university students. Although we acknowledge that this does not fully represent all young adults in Türkiye, the large and diverse student body of İstanbul University from all regions of Türkiye significantly reduces this concern. Furthermore, the current national growth curves are based on measurements taken from students in schools in well-off İstanbul districts with high SES (13). Since the sample that best reflects the growth potential of the society should be selected when constructing growth references, we believe that the study sample was appropriate for our purpose and that the results can be generalized to the Turkish population.

Another potential limitation is the age range of 18-26 years in our study. The previous national study (TK2002) with which we compared our results includes data only up to 18 years old. Despite these limitations, the study provides valuable insights into the anthropometric measurements of Turkish young adults.

Conclusion

Our results have demonstrated that the average height of Turkish young adults has increased by 1.8 cm over the past two decades. Therefore, we suggest that this increase be considered when interpreting target height, to reflect current trends in during growth evaluation. Future research is needed to assess the age period in which this height increase has mainly occurred. Furthermore, while the BMI of young female adults did not change, males have reached alarmingly high BMI levels, with 58% being classified as overweight or obese, highlighting a rising public health issue. This indicates the importance of enhancing prevention programs to promote healthier habits and lower obesity rates, especially among boys.