Introduction

Adolescence, characterized by accelerated muscle and bone development and large changes in the pattern of endocrine secretion, is the second growth spurt of children, the first one being during infancy. About 15% of adult height, 50% of adult weight and 45% of adult bone mass are attained during this period (Matkovic et al, 1990). However, adolescent nutrition has received much less attention than infant nutrition, particularly in developing countries. Although anthropometric measures such as height, weight and body mass index are widely used in general physical examination of adolescents in these countries, evaluation of these measurements and their relationship to health has rarely been done. In a study on a total of 4254 boys and girls aged 10–18 y from different parts of China in 1991 and 1993, Wang et al (1998) reported that, in spite of an improvement in diet and nutritional status, under-nutrition, with a 12–13% prevalence rate, was still an important nutritional problem in adolescents. They defined under-nutrition as age- and gender-specific BMI less than the 5th percentile of American adolescents between 1971 and 1974, as recommended by WHO Expert Committee (1995). However, the WHO-recommended BMI cut-off might not be applicable to Asian children and adolescents, given differences in race and socio-economic conditions (Chang et al, 1994; Daniels et al, 1997; Luciano et al, 1997; Schaefer et al, 1998; De Onis et al, 2001). Furthermore, no evidence of impaired health was provided to support the selection of BMI 5th percentile as the cut-off between under-nutrition and acceptable nutrition in the report by Wang et al (1998). The WHO Expert Committee (1995) has suggested that a local cut-off be developed and used concurrently with the WHO recommended cut-off for under-nutrition in adolescents; that research is needed to determine the most appropriate cut-off values for anthropometric indicators in adolescents, based on functional and health-related outcomes; and that there is a need to determine whether identification of adolescents at risk can be improved by incorporation of maturational status into anthropometric reference data.

It is well known that body weight is positively associated with bone mass. However to our knowledge, no studies have been reported on the effect of low body weight on adolescent health when low bone mass is taken as an end-point. No data on the health effects of cut-off points corresponding to low body mass index in adolescents have been reported. The aims of this study were to measure the prevalence of low body weight in Beijing pubertal girls, and to identify the body mass index (BMI) cut-off point which was associated with delayed growth and maturation, and lower bone mineral status for this age-sex group in China.

Materials and methods

All data used in this study were drawn from a random sample of 1277 Beijing girls aged 12–14 y who participated in a survey of calcium and vitamin D status in 1995, in which sampling followed a socioeconomically stratified (urban, suburban and rural areas), systematic (selection of project schools from a school list) and cluster method. All girls who met the entry criteria in a selected school were recruited, giving a total of 1214 girls with complete anthropometric, dietary and physical activity data. Subjects with evidence of illness, especially liver, kidney or other disorders that may have caused abnormal bone development, were excluded from the study. Data on bone mineral measurement and pubertal maturation were from a random subsample of half of the subjects. The original study received written approval from the Committee on Experimental Procedures Involving Human Subjects of the University of New South Wales, and written consent from parents of all participants (Du et al, 2001).

Measurements

Body weight was measured to the nearest 0.1 kg by using calibrated lever scales (RGT-140, no. 6 Machinery Plant, Beijing), with subjects wearing only light clothes and no shoes. Height to the nearest 0.1 cm using body height measures (TG-III, no. 6 Machinery Plant, Beijing). BMI was calculated by weight in kilograms divided by square of height in metres. Bone age to the nearest 3 months was estimated by means of the Greulich and Pyle Atlas method (Greulich & Pyle, 1959), which assesses development stages of metacarpals, phalanges and carpals from a hand and wrist X-ray against a set of reference photographs. Date at menarche was recorded. Breast and pubic hair development were assessed according to the Tanner method (staging 1–5, where 1=pre-puberty, 5=adult level) by the investigators (Tanner, 1962; Adelaide Children's Hospital, 1989). Physical activity was assessed, firstly, by the School Physical Activity Score (SPAS) for the past year given by a schoolteacher for the performance level and activity of the student in a variety of sports and physical activities at school (1–4, where 1=excellent and most active), and, secondly, by spare time physical activities (PA) in minutes per week over the past year (>15 categories of PA during class breaks and out-of-school hours). Family income was classified as low (<500 yuan/month), middle (500–1500 yuan/month) and high (>1500 yuan).

Nutrient intakes over the past year were estimated by use of a specially designed and validated semi-quantitative food frequency questionnaire (SFFQ) based on the approach by Block et al (1986; Du et al, 2001). The average frequency of consumption of each of 103 food items over the past year, in specified serving sizes, was indicated by marking one of 10 frequency categories. With assistance of a set of measure models and food pictures, the SFFQs were self-administered by subjects at school under supervision of the investigator. Chinese food tables (Institute of Nutrition and Food Hygiene, 1991) were used for nutrient calculations as part of a specially designed software for this age and sex group (He et al, 1997).

Bone mineral content (BMC) and bone width (BW) at distal one-tenth (ratio of trabecular (T) vs cortical (C) bone=20:80) and one-third (T:C=1:99; Rosen, 1996) radius and ulna of non-dominant forearms were measured by using a portable bone mineral analyser (BMD-4, Beijing Broadcast Technology Institute, Beijing) utilising single photon absorptiometry (SPA). The precision error of the analyser was <2%. The short-term precision in vivo was 2.2% with intermediate repositioning. The correlation coefficient between the results from the SPA analyser and a dual energy X-ray absorptiometry (DCS-600EX, Aloka Co. Ltd, Japan, precision error <1%) was 0.603 in a small group of the subjects. The SPA analyser was calibrated against an aluminium alloy model prior to use on each day and at each project site. The measurement was performed by the same two technicians throughout.

Assessment of nutritional status

Nutritional Assessment Reference (weight for height, for girls aged 13 years), Technical Practice on Common Disease Control in Chinese schoolchildren (Ministry of Health & National Education Committee, 1993) served as the basis of the nutritional assessment. Using the average height of the subjects (154 cm), the cut-off value of body weight for different nutritional status was: <41.3 kg (under-nourished), >50.5 kg (overweight), and >55.1 kg (obese), respectively. The cut-off values in BMI were then calculated accordingly as follows: underweight, BMI<18; desirable weight, BMI 18–21; overweight, BMI 21–23; and obese, BMI>23.

The age- and gender-specific BMI cut-offs recommended by WHO (1995) were also used: thinness, BMI<15.36 (5th percentile of American adolescents, female, 13 y); at risk of overweight, BMI23.08 (85th percentile of American adolescents, female, 13 y); stunting was defined as height-for-age less than -2 s.d. from the median value (143.4 cm), or less than the 3rd percentile (144.2 cm) of American girls aged 12.9 y (WHO, 1983).

Statistical analysis

Descriptive statistics were performed for all variables by BMI group, and t-test and chi-square test were used for comparison between BMI<18 and BMI=18–21 groups. Comparison of bone mineral measurements and other variables between the BMI groups was also made by controlling for pubertal maturation. Logistic regression was conducted to determine predictors of BMC at the four sites of the forearm, and to estimate relative risk. Variables that are significantly correlated with BMC, including bone size (Prentice et al, 1994), were included into the regression analysis. BMC was converted into a dichotomous variable by the median. Age at menarche was calculated by using censored likelihood methods (Hediger & Stine, 1987). All data were analysed using SPSS (Version 10.0, SPSS Inc., Chicago, IL, USA).

Results

Using Chinese standards the prevalence rate of underweight/low body weight (BMI<18) among these Beijing girls aged 12.9 (s.d. 0.6) in 1995 was 32.2%. About half of the girls had their BMI in the desirable range (BMI=18–21). There was a significant difference (P<0.0005) of nutritional status among areas, with urban girls having lower rates of acceptable BMI and higher obesity rates (Table 1).

Table 1 - Nutritional status of 1214 Beijing girls aged 12–14 y in 1995^a.

Full table (23K)

About two thirds of the girls had achieved menarche (66.2%; Table 2), and there were no significant differences in menarcheal rates among rural, suburban and urban areas (63.4, 68.6 and 64.4%, respectively). While most girls had reached Tanner stages 3 and 4 for breast development (Table 2), they were still mainly at stages 1–3 for pubic hair development (30.1, 27.2 and 29.0%, respectively). The mean age of menarche calculated from the sample by censored likelihood methods was 13.1 (s.d. 0.1) y.

Table 2 - Characteristics and bone mineral measurements of Beijing girls^a.

Full table (81K)

Table 2 shows that compared with girls with desirable body weight, girls with low body weight had not only a lower bone age, a delayed breast development, a lower rate of menarche, but also a lower distal one-third radius and ulna BMC, BMD and BW. At distal one-tenth radius and ulna, differences in BMC and BW between the two BMI groups were also significant. No differences in age, height, nutrient and milk intakes, SPAS, PAs and family income were observed between the two groups.

Characteristics and bone mineral measurements of Beijing girls achieving menarche by BMI group and girls at high pubertal maturation level (Tanner stage 4 for breast development) by BMI group are presented in Table 3. Controlling for pubertal maturation status, girls with low body weight again had lower BMD at distal one-third radius and ulna than their counterparts with desirable weight while BW was comparable. SPAS, one of the two indicators of physical activity, showed that both low body weight groups were less active than the group with desirable body weight. In addition, subjects at Tanner stage 4 and with low body weight had lower protein and energy intakes than their desirable weight counterparts.

Table 3 - Characteristics and bone mineral measurements of Beijing girls achieving menarche/at high pubertal maturation level, by BMI group^a.

Full table (80K)

Logistic regression analyses of BMC at distal one-third and one-tenth radius and ulna on variables that correlated significantly with BMC or were bone size variables, including bone age, Tanner breast stage, socioeconomic area, SPAS, milk intake, family income, BW, age, height, and weight were performed for the low and desirable body weight subjects. PA was also included in the analysis even though its association with BMC did not achieve a significant level. BMC was converted into a dichotomous variable by the median at each site of the forearm. BMI was coded as a categorical variable (<18 or 18–21). The results showed that BW, bone age, BMI, socioeconomic area and milk intake were predictors of BMC, with models varying by bone site. For the total of four BMC models, bone age was a predictor in three, with an odds ratio up to 1.55 (95% CI 1.25–1.93); socioeconomic area was a predictor in the BMC at distal one-third radius (odds ratio 2.94, 95% CI 1.56–5.29); milk intake was in the BMC at distal one-tenth radius (odds ratio 1.007, 95% CI 1.003–1.01, for rural vs other areas); BMI was a predictor in the BMC at distal one-third ulna. The model is shown in Table 4. When comparing BMI<18 vs BMI=18–21, the risk of the BMC being less than the median increased by 82%.

Table 4 - Logistic regression model of bone mineral content at distal one-third ulna of Beijing girls aged 12–14 y (n=380)^a.

Full table (30K)

Using the age- and gender-specific BMI recommended by WHO, rates of thinness (BMI<15.36) and risk of overweight (BMI23.08) were 5.7 and 13.5%, respectively, for these Beijing girls. Comparison of percentile BMI values between the 1214 Beijing girls (13 y, 1995) and American girls (13 y, 1968) are presented in Figure 1. The mean BMI values of Beijing girls were lower than those of American girls at each percentile (5th, 15.27 vs 15.36, 15th, 16.29 vs 16.43; 50th, 18.51 vs 18.95; 85th, 22.68 vs 23.08; 95th, 26.35 vs 27.07); however, Beijing urban girls had higher BMI values than American girls at 85th and 95th percentiles, showing a trend towards obesity among Beijing urban girls. The stunting rate was 4.0% using the less than -2 s.d. cut-off and 4.8% using the less than 3rd percentile cut-off, respectively.

Figure 1.

Comparison of percentile BMI values between the 1214 Beijing girls (13 y, 1995, by rural, suburban and urban area) and American girls (13 y, 1968; WHO Expert Committee, 1995). The mean BMI values of Beijing girls are lower than those of American girls at each percentile (5th, 15.27 vs 15.36; 15th, 16.29 vs 16.43; 50th, 18.51 vs 18.95; 85th, 22.68 vs 23.08; 95th, 26.35 vs 27.07, however, Beijing urban girls had higher BMI values than American girls at 85th and 95th percentiles

Full figure and legend (36K)

Discussion

Assessed by Chinese criteria, Beijing girls aged 12.9 y had a low body weight rate (BMI<18) as high as 32.2%, in contrast to a 5.7% of thinness rate if the 5th percentile BMI value (15.36) of American girls aged 13 y in 1968 was used for the assessment as recommended by the WHO Expert Committee (1995). Given the similar distribution of BMI in Beijing girls and the American reference group, and given that mean menarche age of Beijing girls was only 0.3 y later than that of American girls (13.1 vs 12.8 y, WHO Expert Committee, 1995), it seems reasonable to use the WHO reference in assessment of nutritional status of Beijing girls for this age group. The 5.7% of thinness rate of Beijing girls aged 13 y in 1995 was much lower than the 12–13% thinness rate of Chinese adolescents aged 10–18 y from eight provinces of China in 1991–1993 reported by Wang et al, and stunting rate was also much lower in the Beijing girls than in those Chinese girls in the study by Wang et al (4.0 vs 16–19%), indicating that undernutrition in adolescents was less serious in Beijing, the capital city of China, than in other areas of China. The rates of overweight and obesity by the Chinese criteria (BMI=22–23, and BMI>23) were 7.7 and 12.4%, respectively, and the combined rate of 21.1% was higher than the 13.5% rate of risk of overweight by the WHO reference (BMI23.08). International cut-off points for BMI for overweight and obesity in children have been suggested recently by Cole et al (2000): cut-off points of BMI for 13-y-old girls are 22.58 for overweight and 27.76 for obesity, defined to pass through BMI of 25 and 30 at age 18. Using these cut offs, the rates of overweight and obese would be 12.4 and 3.1%. The combined rate was 15.5%, which fell between the Chinese standard and that of the WHO reference. Although the upper limit of the desirable BMI for these Chinese girls needs validating, preferably against markers of disease risk, more attention needs to be paid to the trend towards obesity among urban girls. The mean menarcheal age of 13.1 (s.d. 0.1) y calculated from current study sample is slightly higher than the median ages of menarche in urban girls (12.61 (1.11) y) and rural girls (12.97 (1.05) y) of Beijing from a national survey on children and adolescents' health in 1985 (Lin et al, 1992). In the study of Lin et al, probit analysis was used and the subjects were aged 7–22 y. This may have resulted in a less accurate estimate than the method and subjects in the study reported here. Compared with a median menarcheal age of 14.16 y in Beijing girls in 1962, using probit analysis again (Ye et al, 1981), with a mean menarcheal age of 13.1 y in 1995, Beijing girls appear to experience pubertal maturation earlier than 30 y ago. In contrast, using the WHO reference data, De Onis et al (2001) reported a 50.5% prevalence of thinness and a 11.2% prevalence of stunting in 818 Indian boys aged 7–16 y (both figures being much higher than those from our current study of Chinese girls), and indicated that the reference data seemed inadequate for their study sample of Indian children.

Our current study, however, shows a significant difference in growth, maturation and bone mineral parameters by setting the cut-off for low body weight at BMI of 18. Girls with BMI less than 18, compared with those with BMI of 18–21, not only had delayed growth and development, as indicated by lower bone age, delayed breast and pubic hair development, and a lower rate of menarche, but they also had lower BMD at the distal one-third radius and ulna, with other confounding factors being comparable, especially the stage of pubertal maturation (Tables 2 and 3). The greatest difference in BMD between the two BMI groups was 0.044 g/cm² (44% of the s.d.) and, if this difference persisted, those with a lower BMD and BMI could have lower BMD later in life. A change of 1 s.d. in bone mass may reduce the risk of fracture in later life by as much as 100% (Hui et al, 1989; Wasnich et al, 1989). Furthermore, logistic regression of BMC at distal one-third ulna confirmed that BMI was a predictor of BMC, after controlling for other possible confounding factors. When comparing BMI<18 vs BMI = 18–21, the risk of BMC less than the median increased by 82% (Table 4). It was reported that Japanese women with earlier menarche had higher peak bone mass than those with later menarche (Ito et al, 1995). In our study, girls with a low body weight also had delayed menarche and breast development and it may also have had an adverse effect on BMC/BMD (Table 3), despite the fact that pubertal maturation was not shown to be an independent determinant of BMC in the logistic analysis. Girls with low body weight need more primary health care since the phenomenon of BMI tracking (maintenance of steady BMI status over time) has been demonstrated in Chinese children (Wang et al, 2000). Furthermore, attention should be paid to post-menarcheal girls and those at higher pubertal maturation stages with low body weight (accounting for 14.3% of the study sample). They have less chance to catch up to the BMD levels of their acceptable weight counterparts than do pre-menarcheal or less mature low-body-weight girls, since menarche and advanced pubertal maturation indicate the end of the growth spurt, and the gap in BMD is more likely to be continued into adulthood. Menarcheal status (more easily assessed than pubertal stage) therefore can be used together with BMI to assess the risk of lower BMD in this population.

Our finding that girls with same higher Tanner stage but lower BMI had significantly lower BMD (Table 3) would seem to suggest that these girls may have a lower peak bone mass in adulthood. However, the current study, being a cross-sectional study, cannot determine whether their BMD would increase if they gained weight, or whether they are simply slower to attain peak bone mass. Such questions could only be answered by a follow-up study of these girls to full maturity and even into their late twenties.

The bone mineral measurement in this study was conducted using SPA densitometry which showed a relatively low correlation coefficient of 0.603 against a dual-energy X-ray absorptiometry (DXA) in a group of eight subjects. This correlation coefficient would be raised to 0.837, with SPA accounting for 70% of the variability in DXA measurements (improved from 36.4%), if one subject with a pair of extreme values was removed from the analysis. However, the association between the BMD results of the two methods may be strong enough and the SPA reliable enough for an epidemiological survey such as this, which focused on comparison of group values and was not designed to set up reference BMD values.

While there were no differences in nutrient intakes and SPAS between the low body weight and the desirable weight groups (Table 2) we did find that higher SPAS (meaning poorer performance in school physical education classes) and lower protein and energy intakes characterized girls with low body weight when comparison was made by stratification of pubertal status (Table 3). These results suggest protein and energy intakes and school physical education activity might also contribute to BMI, though that BMI is mainly genetically determined (Pietilainen et al, 1999). Socioeconomic area was another environmental factor associated with BMI (Table 1), probably reflecting the uneven rate of development in different areas of China in recent years.

In conclusion, the high prevalence of low body weight (BMI<18) among Beijing pubertal girls is still a major health problem. BMI<18 can be recommended as the cut-off for diagnosing delayed growth and development in girls aged 12–14 y and for screening girls at risk of lower BMD. Together with BMI, menarcheal status and pubertal maturation stage can be used to assess the risk of low BMD and to target intervention groups. Primary health care and appropriate interventions (nutrition and physical activity) need to be implemented among Chinese school age girls to improve their growth and bone health.

References

1. Adelaide Children's Hospital (1989). Stages of Puberty, Sydney: Pharmacia & Upjohn
2. Block, G, Hartman, AM, Dresser, CM, Carroll, JG & Gardner, L (1986). A data-based approach to diet questionnaire design and testing. Am. J. Epidemiol., 124, 453–469. | PubMed | ISI | ChemPort |
3. Chang, Y, Zhai, F, Li, W, Ge, K, Jin, D & de Onis, M (1994). Nutritional status of preschool children in poor rural areas of China. Bull. WHO., 72, 105–112.
4. Cole, TJ, Bellizzi, MC, Flegal, KM & Dietz, WH (2000). Establishing a standard definition for child overweight and obesity worldwide: international survey. Br. Med. J., 320, 1240 | Article | ISI | ChemPort |
5. Daniels, SR, Khoury, PR & Morrison, JA (1997). The utility of body mass index as a measure of body fatness in children and adolescents: differences by race and gender. Pediatrics, 99, 804–807. | Article | PubMed | ISI | ChemPort |
6. De Onis, M, Dasgupta, P, Saha, S, Sengupta, D & Blössner, M (2001). The National Center for Health Statistics reference and the growth of Indian adolescent boys. Am. J. Clin. Nutr., 74, 248–253. | PubMed | ChemPort |
7. Du, X, Greenfield, H, Fraser, DR, Ge, K, Trube, A & Wang, Y (2001). Vitamin D deficiency and associated factors in adolescent girls in Beijing. Am. J. Clin. Nutr., 74, 494–500. | PubMed | ISI | ChemPort |
8. Greulich, WW & Pyle, SI (1959). Radiographic Atlas of the Skeletal Development of the Hand and Wrist, 2nd ednCalifornia: Stanford University Press
9. He, W, Du, X & Greenfield, H (1997). CAVD, a Survey System Using Epi Info, Beijing: Chinese Academy of Preventive Medicine (Diskette)
10. Hediger, ML & Stine, RA (1987). Age at menarche based on recall information. Ann. Hum. Biol., 14, 133–142.
11. Hui, SL, Slemenda, CW & Johnston, CC (1989). Baseline measurement of bone mass predicts fracture in white women. Ann. Intern. Med., 111, 355–361.
12. Institute of Nutrition and Food Hygiene (1991). Food Composition Tables, Beijing: People's Health Publishing House
13. Ito, M, Yamada, M, Hayashi, K, Ohki, M, Uetani, M & Nakamura, T (1995). Relation of early menarche to high bone mineral density. Calcif. Tissue Int., 57, 11–14.
14. Lin, WS, Chen, CN, Su, JZX, Zhu, FC, Xing, WH, Li, JY & Te, GS (1992). The menarcheal age of Chinese girls. Ann. Hum. Biol., 19, 503–512. | PubMed |
15. Luciano, A, Bressan, F & Zoppi, G (1997). Body mass index reference curves for children aged 3–19 years from Verona, Italy. Eur. J. Clin. Nutr., 51, 6–10. | Article | PubMed |
16. Matkovic, V, Fontana, D, Tominac, C, Goel, P & Chesnut, CHIII (1990). Factors that influence peak bone mass formation: a study of calcium balance and the inheritance of bone mass in adolescent females. Am. J. Clin. Nutr., 52, 878–888. | PubMed |
17. Ministry of Health and National Education Committee (1993). Chinese Reference Standards for Anthropometry Assessment of School Children, Beijing: MH & NEC
18. Pietilainen, KH, Kaprio, J, Rissanen, A, Winter, T, Rimpela, A, Viken, RJ & Rose, RJ (1999). Distribution and heritability of BMI in Finnish adolescents aged 16y and 17y: a study of 4884 twins and 2509 singletons. Int. J. Obes. Relat. Metab. Disord., 23, 107–115. | Article | PubMed | ChemPort |
19. Prentice, A, Parsons, TJ & Cole, TJ (1994). Uncritical use of bone mineral density in absorptiometry may lead to size-related artifacts in the identification of bone mineral determinants. Am. J. Clin. Nutr., 60, 837–842. | PubMed | ChemPort |
20. Rosen, CJ (1996). Osteoporosis: Diagnostic and Therapeutic Principles, Totowa, NJ: Humana Press
21. Schaefer, F, Georgi, M, Wuhl, E & Scharer, K (1998). Body mass index and percentage fat mass in healthy German schoolchildren and adolescents. Int. J. Obes. Relat. Metab. Disord., 22, 461–469. | Article | PubMed | ChemPort |
22. Tanner, JM (1962). Growth at Adolescence, pp30–36, Oxford: Blackwell Scientific Publications
23. Wang, Y, Popkin, B & Zhai, F (1998). The nutritional status and dietary pattern of Chinese adolescents, 1991 and 1993. Eur. J. Clin. Nutr., 52, 908–916. | Article | PubMed | ISI | ChemPort |
24. Wang, Y, Ge, K & Popkin, BM (2000). Tracking of body mass index from childhood to adolescence: a 6-y follow-up study in China. Am. J. Clin. Nutr., 72, 1018–1024. | PubMed | ISI | ChemPort |
25. Wasnich, RD, Ross, PD, Davis, JW & Vogel, JM (1989). A comparison of single and multi-site BMC measurements for assessment of spine fracture probability. J. Nucl. Med., 30, 1166–1171.
26. WHO (1983). Measuring Change in Nutritional Status, Geneva: WHO
27. WHO Expert Committee (1995). Physical Status, the Use and Interpretation of Anthropometry, WHO Technical Report Series no. 854Geneva: WHO
28. Ye, GS, Lin, ZZ, Lin, WS, Li, BW, Tang, XW & Bian, GN (1981). A study of Beijing student adolescence. Chin. Med. J., 94, 101–108, (in Chinese)

Acknowledgements

The authors are indebted to Professor GS Ma, Mr XW Li, Ms GS Liu, Mr XX Li, Miss JJ Tan, Mr JM Zhu and his staff, Professor W He, Mr BM Guo, Ms JX Gao, Miss H Wang, Professor A Baumann, Dr D Mackerras, Ms E Emmerson for their help and technical assistance in the project. The authors are grateful to all principals, staff and health workers involved for their effort and contribution to the fieldwork and all subjects and their parents for their cooperation.