Data were derived from the NSPGDC in nine cities of China [1, 2, 8,9,10], which was a repeated cross-sectional national special survey undertaken since 1975, once every 10 years, and a subproject of the NSPGDC, namely, the National Epidemiological Survey on Simple Obesity in Childhood (NESSOC) in the same nine cities [11]. To exhibit a more complete picture of the growth and nutrition of Chinese children, we used all data of the series of the NSPGDC from 1975 to 2015 to exhibit secular trends in height and BMI trajectories and the latest data of the NESSOC in 2016 to examine the updated prevalence of stunting, underweight, overweight and obesity among children under 7 years old in Chinese cities.

The NSPGDC was carried out in five rounds in nine cities of China in 1975, 1985, 1995, 2005 and 2015. Of these nine cities, Beijing and Shanghai are province-level municipalities that have almost the highest level of socioeconomic development in China, and the other seven cities are provincial capital cities that have almost the highest level of socioeconomic development in their respective provinces, including Harbin (Heilongjiang’s provincial capital), Xi’an (Shaanxi), Nanjing (Jiangsu), Wuhan (Hubei), Guangzhou (Guangdong), Fuzhou (Fujian) and Kunming (Yunnan). On the other hand, these nine cities covered all seven geographical regions of China, including East China (Shanghai, Nanjing, Fuzhou), South China (Guangzhou), North China (Beijing), Central China (Wuhan), Southwest China (Kunming), Northwest China (Xi’an), and Northeast China (Harbin). The NSPGDC used stratified cluster sampling according to urban/suburban rural areas and administrative districts in each survey city. Children under 3 years old were from the street community, and children aged 3 to 6.9 years were from kindergarten. Urban children are determined as permanently living in urban areas in the surveyed city, or these children moved into the surveyed city from other large cities and lived for more than two-thirds of the children’s age; and suburban rural children, as either one or both parents are farmers and children are brought up in the suburban rural area (surrounding the surveyed city). Children under 7 years old are classified into 22 age groups: newborn to 3 days, monthly for 1 to 5 months, bimonthly for 6 to 11 months, tri-monthly for 12 to 23 months, half-yearly for 2 to 5.9 years, and yearly for 6 to 6.9 years. The classification of age groups follows the criteria of ‘‘low limit’’; for example, 3 months represents children aged 3.0–3.9 months. Each sex-age subgroup consists of about 200 children in urban/rural areas of each survey city. About 1800 children were included in each age subgroup for urban boys, urban girls, rural boys and rural girls. Exclusion criteria were temporary residents, premature birth, birth weight below 2.5 kg, multiple births, acute illness within a month, chronic illness, and obvious malnourishment and physical handicap.

The NESSOC was one of the subprojects of the NSPGDC and the most representative epidemiological special survey on childhood obesity in China [11, 12], and the surveyed cities and sampling method were the same as the NSPGDC. One or more districts were selected as the study units in each survey city, and the estimated numbers of children under 7 years old were more than 10 thousand, with about half in urban areas and half in rural areas. Children below 3 years were selected from the street community, and children aged 3 to 6.9 years were selected from kindergarten. The survey number of communities and kindergartens was estimated based on the total number of children under 7 years old and the age distribution of children in the selected districts of each survey city. The participation proportion of the study subjects was not less than 95% in the selected communities and kindergartens, and the total number of participants was not less than 10 thousand in each survey city.

A total of 904,263 children under 7 years were included in this present study, with 793,772 children from five rounds of the NSPGDC between 1975 and 2015 and 110,491 children from the NESSOC in 2016. Table 1 shows the detailed sample sizes by gender, urban/rural area, and survey year.

Using unified measuring tools/instruments in a standardized way by specially trained technicians or nurses [2, 13], weight was measured with children wearing the lightest vest, shorts or underwear by newborn scales (accurate to 10 g) for newborns, and lever scales (accurate to 50 g, in 1975–2005) or electronic scales (T-scale M303, accurate to 50 g, in 2015) for children aged 1 month to 6.9 years, and height (not in shoes) was measured as recumbent length with a horizontal infantometer (accurate to 0.1 cm) for children under 3 years old and as standing height with a metal column height measuring device (accurate to 0.1 cm) for children aged 3 to 6.9 years in 1975‒2015. All height measurements were recorded to the nearest 0.1 cm. Measurement errors of weight and height were not more than 0.05 kg or 0.5 cm between two repeated measurements.

The NSPGDC and its subproject (NESSOC) were undertaken in nine cities of China. The steering committee was located in Beijing (capital of China), and the nine coordinating subgroups in each survey city worked simultaneously following a uniform program, organizing and training local survey teams, carrying out the measurements and checking the questionnaires under the direction and supervision of the Beijing steering committee. All the field investigators participated in rigorous training and passed professional examinations before starting the field investigation. Unified measuring tools/instruments were equipped for each field site of each survey city. The field investigation started in May and finished in October of the same year. All physical measurements were carried out at least 1 hour after a meal between approximately 8 a.m. and 4 p.m. The questionnaire design, questionnaire entry and data verification were completed by the Beijing steering committee.

Data on gross domestic product (GDP) per capita, Engel’s coefficient, health professionals including doctors and nurses, and under-5 and infant mortality rates between 1978 and 2020 were obtained from the China Statistical Yearbook 2022 [14].

BMI was calculated as weight (kg) divided by the square of length/height (m). A 40-year change in height from 1975 to 2015 was examined, but only a 30-year change in BMI from 1985 to 2015 was analyzed due to the scarcity of the raw dataset in 1975, and BMI was not calculated and documented in 1975, as BMI was not widely used. We defined prevalence of stunting as < − 2 standard deviations (SD) from median for length/height-for-age, underweight as < − 2 SD, overweight as ≥ + 1 SD but < + 2 SD, and obesity as ≥ + 2 SD from median for BMI-for-age based on the China growth standards for 0–18 years [4] or the World Health Organization (WHO) growth standards for 0–5 years and the WHO growth references for 5–19 years [15, 16]. A comparable definition of the prevalence according to the same cutoffs when using the China and WHO criteria may be more feasible and reasonable for the under-7-year data in this present study. Differences in the prevalence based on the China and WHO criteria were tested by McNemar’s test. A two-sided P value of less than 0.05 was considered statistically significant. Data analysis was performed in SAS v9.4 (SAS Institute, Cary, North Carolina).