Chinese Medical Journal 2014;127 (9)

1721

Original article
Breast milk macronutrient composition and the associated factors in
urban Chinese mothers
Yang Titi, Zhang Yumei, Ning Yibing, You Lili, Ma Defu, Zheng Yingdong, Yang Xiaoguang, Li Wenjun,
Wang Junkuan and Wang Peiyu
Keywords: lactation; milk composition; macronutrients; dietary intake; body mass index
Background Infancy is a critical period of growth and physiological development, in which breast milk is the best
source of nutrients. Compared to western countries, research on breast milk of Chinese population are limited. Thus, it is
necessary to measure breast milk energy and macronutrient concentrations of healthy urban Chinese mothers at different
lactation stages, to expand the database of milk composition of Chinese population, and to examine whether dietary or
other maternal factors can affect the levels of macronutrients in breast milk.
Methods Breast milk of full expression of one side breast from 436 urban Chinese lactating mothers at 511 days, 1230
days, 3160 days, 61120 days, and 121240 days postpartum was obtained at 9:00 a.m. to 11:00 a.m. Total energy,
lactose, protein, and fat contents were measured. 24-hour dietary recall was surveyed, and maternal nutrient intakes were
analyzed.
Results Milk composition changed over the course of lactation and large individual variations were documented. The
concentrations were 61.3 kcal/dl for total energy, 7.1 g/dl for lactose, 0.9 g/dl for protein, and 3.4 g/dl for fat in mature milk.
Stage of lactation was a strong factor affecting milk composition. Minimal evidence was found for associations between
maternal current dietary intake and milk macronutrient concentration, consistently with prior research. Maternal body mass
index (BMI) was positively associated with milk fat content, to a greater extent than did dietary intake. All other maternal
characteristics were not significant for milk composition.
Conclusion These findings suggest that milk composition is generally weakly associated with maternal factors except
for stage of lactation, and is likely to be more susceptible to long-term maternal nutritional status than short-term dietary
fluctuation.
Chin Med J 2014;127 (9): 1721-1725

nfancy is a critical period of growth and physiological

development, in which breast milk plays the most
important role for infant nutrition. Under ordinary
circumstances, breast milk provides all the energy
and nearly all nutrients required for infant growth and
development during the first 4 to 6 months of life, as well as
various immunological factors and bioactive components.1
When complementary foods are introduced, breast milk
can still provide half of the energy needs between 6 and 12
months of age, and one-third the energy needs between 12
and 24 months.2
Given the importance of breast milk for growing infants,
numerous research on breast milk have been conducted
internationally in the last few decades. The composition
has been found to be highly variable: it varies among and
within individuals, among different populations, within
a feeding, diurnally from feeding to feeding, and from
day to day. It is also influenced by stage of gestation and
lactation, as well as different techniques of sampling and
storage.1,3 Maternal nutrition has been found to have weak
associations with milk production, but the conclusions were
sometimes inconsistent, especially when the mothers were
under extreme conditions.3
Compared to western countries, research conducted in

China involving Chinese mothers is relatively limited, and

the majority of the comprehensive studies were performed
in the 1980s. There are so far no representative values of
milk composition in Chinese mothers and many data are
based on research on western population. In addition, some
studies were inconsistent in milk collection, composition
analysis, or interpretation. Therefore, it is critical to provide
solid research and evidence regarding Chinese breast milk.
The purposes of this study were to measure breast milk
energy and macronutrient concentrations of healthy urban
Chinese mothers at different lactation stages, to expand the
database of milk composition of the Chinese population,
and to examine whether dietary or other maternal factors
DOI: 10.3760/cma.j.issn.0366-6999.20133260
Department of Nutrition & Food Hygiene, School of Public Health,
Peking University Health Science Center, Beijing 100191, China
(Yang TT, Zhang YM, You LL, Ma DF, Zheng YD and Wang PY)
Nestl Research Center Beijing, Beijing 100095, China (Ning YB
and Wang JK)
National Institute for Nutrition and Food Safety, Chinese CDC,
Beijing 100050, China (Yang XG)
Nestl Nutrition Institute, Beijing 100102, China (Li WJ)
Correspondence to: Wang Peiyu, Department of Nutrition & Food
Hygiene, School of Public Health, Peking University Health Science
Center, Beijing 100191, China (Tel and Fax: 86-10-82802502. Email:
wpeiyu@bjmu.edu.cn)
This study was funded by Nestl, China.

Chin Med J 2014;127 (9)

1722

can affect the levels of macronutrients in breast milk.

METHODS
Study design
This study is part of the MING (Mother and Infant
Nutritional Growth) Study, which was designed to survey
the current nutritional status of urban Chinese pregnant
women, lactating mothers, infants and toddlers. This paper
presents data only for the lactating mothers. Participants
were recruited from three cities (Beijing, Suzhou, and
Guangzhou); in each city, one hospital and one maternal
and childcare service center were selected randomly. A
stratified sample including 150 mothers from each of the
three cities (30 each of whom were at 511 days, 1230
days, 3160 days, 61120 days, and 121240 days
postpartum, respectively) was expressly obtained. The
selection criteria were: breastfeeding mothers aged 18 to
45, were primigravida and healthy, and had delivered a
single healthy baby. After excluding those who did not
deliver their babies at term, a total of 436 mothers with
86, 85, 88, 90, and 87 from each postpartum time point,
respectively, were examined. The study was conducted
according to the guidelines in the Declaration of Helsinki,
where all procedures involving human subjects were
approved by the Medical Ethics Research Board of Peking
University (No. IRB00001052-11042). Written informed
consents were obtained from all participants.
Anthropometric and dietary interviews
Anthropometric indices were measured. Other
characteristics, including age, social-economic status,
pregnancy history, delivery mode, physical activity status,
and smoking and drinking habits, were also taken into
consideration. Dietary information was collected using
24-hour dietary recall surveys conducted by trained
interviewers. Measuring cups and a food photograph
booklet were used to help respondents estimate portion
sizes. Nutrient intakes were then analyzed using reference
values from the 2002 China Food Composition Database
(National Institute of Nutrition and Food Safety, China
CDC).
Milk sample collection and analysis
Milk samples were collected from one breast between 9:00
and 11:00 a.m. Mothers were asked to empty their breasts in
the previous feeding session. Milk was collected using hand
expression and breast pump until the breast was completely
empty. Milk was then gently swirled and 40 ml was taken
for analyses of individual samples. The remaining milk was
fed to the child on request. The samples were placed on
freezer packs, and delivered within 2 hours to the analytical
labs where they were aliquoted and stored at 80C.
Major components (lactose, protein, fat, and energy) of
milk samples were analyzed using a MIRIS human milk
analyzer (Uppsala, Sweden), which is based on a midinfrared transmission spectroscopy technique. Two ml
aliquot from each sample was thawed and warmed to

40C, then analyzed. The quantities of the components

were measured in one single run, and total energy was
automatically calculated. The human milk analyzer has
been applied mostly in a clinical setting and was reported
to reliably measure lactose, protein, and fat, compared to
conventional methods (r=0.93, 0.92, 0.72 for fat, protein,
and lactose concentrations, respectively).4
Data analysis
Statistical analyses were carried out using SPSS 20.0
(SPSS Inc., Chicago, IL, USA). The analysis of variance
(ANOVA) and Kruskal-Wallis H test were used to
compare milk composition at each lactation stage.
Pearsons correlation coefficients were used to examine the
relationships between dietary factors and concentrations
of milk components at each lactation stage. General linear
models were used to examine the associations between
maternal variables and milk macronutrient concentrations.
Geographic location, stage of lactation, delivery mode,
and whether participating physical activity were set as
fixed factors; dietary carbohydrate, protein, and fat intake,
maternal age, body mass index (BMI), pre-pregnancy BMI,
pregnancy weight gain, and postpartum weight retention
(calculated as current weight minus pre-pregnancy weight)
were set as covariates. Backward removal method was
used to select the significantly associated factors. A P value
<0.05 was considered as statistically significant.
RESULTS
Maternal characteristics and dietary macronutrient
intakes
Maternal characteristics are listed in Table 1. The
mothers were 20 to 41 years old, the majority of whom
had received high school or higher education. 72.6% of
them had monthly household per capita income of RMB
2 0006 000 Yuan. Dietary energy and macronutrient
intakes considerably varied among individuals, as most
standard deviations were about one-half of the means.
Milk macronutrient concentration
Mean values of milk energy and macronutrients from
each lactation stage are shown in Table 2. Energy and
macronutrients levels among those at 511 days were
significantly different from those in the 1230 days group;
Table 1. Maternal descriptive characteristics (meanSD)
Characteristics
Age (years)
Height (cm)
Weight (kg)
BMI (kg/m2)
Pre-pregnancy BMI (kg/m2)
Pregnancy weight gain (kg)
Postpartum weight retention (kg)
Cesarean delivery (mean (%))
Participation of exercise (mean (%))
Dietary energy (kcal/d)
Dietary carbohydrate (g/d)
Dietary protein (g/d)
Dietary fat (g/d)

Results (n=436)
27.23.9
160.34.6
59.78.5
23.33.1
20.82.7
16.25.7
6.65.3
45.5
39.9
1 991.2857.1
240.4125.4
81.051.4
82.747.7

Chinese Medical Journal 2014;127 (9)

1723

Table 2. Breast milk composition in different stage of lactation (meanSD)

Composition
Energy (kcal/dl)
Lactose (g/dl)
Protein (g/dl)
Fat (g/dl)

511 days (n=86)

58.610.4*
6.80.5*
1.30.3*
2.91.1*

1230 days (n=85)

64.112.4
7.00.4
1.10.2
3.61.4

3160 days (n=88)

62.014.0*
7.20.3
0.90.2
3.51.6 *

61120 days (n=90)

58.013.3*
7.20.3
0.80.1
3.11.4 *

121240 days (n=87)

58.814.2*
7.30.3
0.80.1
3.21.6 *

Values with different superscript symbols differ significantly, P<0.05. 1 Kcal/dl=41.84 kJ/L.

energy, lactose, and fat concentrations were significantly

higher, while protein concentration was significantly
lower among the 1230 days participants. Compared to
1230 days, energy densities at 3160 days presented
no difference, but decreased at 61120 days. Lactose
concentration at 3160 days was significantly higher
compared to that of the 1230 days participants, and was
maintained thereafter. Protein concentration gradually
decreased and remained stable in 61240 days. Fat
concentrations did not change significantly after 1230
days. The inter-individual CVs ranged from 37.9% to
50.0% for fat, 12.5% to 23.1% for protein, and 4.1% to
7.4% for lactose for all of the 5 stages combined. Of note is
that while the data in the present study represents values of
milk constituents at various lactation stages, it is suboptimal
to reflect trends of longitudinal changes.
Correlations of maternal dietary intake and milk
composition
Since milk energy and fat concentrations were highly
correlated (r=0.98, P <0.001), only analyses of milk
fat, as well as lactose and protein were presented. In
Table 3, energy intake was correlated only with milk fat
concentration in 61120 days (P=0.041). There were no
one-to-one correlations between dietary macronutrient
intakes and the concentrations of the breast milk
counterparts.
Milk composition and maternal factors
Results of the general linear models examining the
associations between maternal variables and milk
macronutrients concentrations are shown in Table 4. Stage
of lactation had a major impact on all milk macronutrients.
The regression coefficients of 61120 days and 121240
days were not significantly different, thus were combined
and used as a reference; the regression coefficients of
the other three stages were the relative values compared
to the 61240 days reference. Milk composition also
varied among cities; the regression coefficients of Beijing

Table 3. Correlation coefficients of dietary intake and milk

composition
Stage of lactation
511 days (n=86)
Diet. energy
Diet. carbohydrate
Diet. protein
Diet. fat
1230 days (n=85)
Diet. energy
Diet. carbohydrate
Diet. protein
Diet. fat
3160 days (n=88)
Diet. energy
Diet. carbohydrate
Diet. protein
Diet. fat
61120 days (n=90)
Diet. energy
Diet. carbohydrate
Diet. protein
Diet. fat
121240 days (n=87)
Diet. energy
Diet. carbohydrate
Diet. protein
Diet. fat

Lactose

Protein

Fat

0.057
0.083

0.099

0.076

0.081

-0.122

0.006
0.047

-0.139

-0.097

0.159

0.065

0.127
0.169

0.072

0.053

0.068

0.083

0.196
0.125

0.173

0.077

0.216*

0.203

0.186
0.018

0.047

0.13

0.117

0.166

and Suzhou were relative to that of Guangzhou. Dietary

fat intake had significant but minimal effect on milk
lactose and protein concentrations, but not on milk fat
concentration. Moreover, although the associations were
weak, maternal BMI was positively associated with
protein and fat concentrations, and negatively with lactose
concentration. All other maternal variables were excluded
from the models as insignificant factors.
DISCUSSION
The purposes of this study were to measure breast milk
energy and macronutrient concentrations of healthy urban
Chinese mothers at various lactation stages, and to examine

Table 4. General linear model for milk macronutrient content

Items
Intercept
Diet. Fat (g)
BMI
Stage 511 days
Stage 1230 days
Stage 3160 days
Stage 61240 days
Beijing
Suzhou
Guangzhou

B
7.481
0.001
0.012
0.465
0.266
0.079
0*
0.020
0.104
0*

Lactose
95% CI
(7.123, 7.320)
(0.000, 0.002)
(0.024, 0.000)
(0.562, 0.369)
(0.362, 0.170)
(0.175, 0.016)

(0.068, 0.108)
(0.191, 0.017)

P values
0.000
0.026
0.047
0.000
0.000
0.102

0.654
0.019

B
0.616
0.000
0.007
0.511
0.268
0.107
0*
0.072
0.079
0*

Protein
95% CI
(0.463, 0.769)
(0.001, 0.000)
(0.000, 0.014)
(0.457, 0.565)
(0.214, 0.322)
(0.054, 0.161)

(0.023, 0.121)
(0.031, 0.128)

This parameter is set to zero because it is redundant.

P values
0.000
0.020
0.037
0.000
0.000
0.000

0.004
0.001

B
1.882

0.054
0.326
0.418
0.251
0*
0.275
0.342
0*

Fat
95% CI
(0.876, 2.889)

(0.009, 0.100)
(0.691, 0.039)
(0.053, 0.783)
(0.112, 0.614)

(0.608, 0.059)
(0.012, 0.671)

P values
0.000
NS
0.019
0.080
0.025
0.175

0.106
0.042

1724

if the concentrations were associated with dietary or other

maternal factors. Although there was no biochemical data,
the anthropometric measurements, screening information,
and socio-economic data suggested that these mothers
were likely in good nutritional status, and had little
problem accessing adequate food. In this study, full-term
pregnancies and primigravidas (due to Chinas one child
policy) with single birth prevented the interference from
pre-term delivery and multiparity.3,5
Milk composition
Measurement was relied on a single milk sample instead
of pooled 24-hour samples due to the limitations in field
condition. Currently, there is no agreement regarding
the time of day for collecting the milk sample that best
represents the pooled 24-hour samples.6,7 Therefore, a
standardized time in the morning was selected for milk
collection. Before milk was collected, the mothers were
asked to empty their breasts in the previous feeding session
as the amount of milk removed at last feed would impact
milk fat content.8 Full expression of milk was collected to
obtain an early to late feed of milk as hind milk contains
more fat than fore milk.9
As was expected, considerable individual variation in
milk composition presented, with fat content being the
most variable. Following trends previously described in
the literature, milk composition rapidly changed during
lactation establishment.3 There are different cutoff points
to separate colostrum, transitional milk, mature milk, and
extended milk. Here, with milk from 12120 days regarded
as mature milk, it comprised 61.3 kcal/dl (1 Kal/dl=41.84 kJ/L)
for energy, 7.1 g/dl for lactose, 0.9 g/dl for protein,
and 3.4 g/dl for fat. Compared with the USA reference
values of mature milk summarized by Picciano (6.7 g/dl
for lactose, 0.9 g/dl for protein and 6570 kcal/dl for
energy), protein concentration was within range, while
lactose concentration was significantly higher and energy
density was lower.10 To compare with a study conducted in
1987 in Wuhan, China, which also collected a single milk
sample by full expression (mean lactose, protein, and fat
concentrations of 31180 days were 6.4, 1.1, and 3.8 g/dl,
respectively), milk was higher in lactose, but lower in
protein and fat.11 It should be noted that the protein content
measured in that study included non-protein nitrogen,
which overestimated the protein content. Such differences
in values were within the expectations as constituents were
summarized to widely range from 6088 kcal/dl for energy,
2.56.0 g/dl for fat, 0.831.30 g/dl for protein, 6.38.1 g/dl
for carbohydrate in mature milk.12
Milk composition and maternal factors
Few maternal factors were found to be associated with
milk macronutrients in this sample. Stage of lactation was
the strongest factor affecting milk composition, which
is in accordance with prior literature. Results of Table
2 and Table 4 suggest that milk lactose concentration
increased with stage of lactation, and maintained
throughout 18 months. Such positive association between

Chin Med J 2014;127 (9)

lactose concentration and duration of lactation has been

documented,13,14 though some studies reported otherwise.15
Protein content declined with duration of lactation, which
was constantly reported previously.13-15 Similar to the results
of the present study, fat content has been shown to decline
with infant age,14 while conflicting findings were still
indicated in some studies.15
The effects of maternal diet on milk composition have
been wildly investigated; no convincing evidence from
observational or experimental research has been shown
that dietary factors can affect total fat, protein or lactose
contents.3,16,17 Consistently, our results found no one-to-one
association between current dietary macronutrient intakes
and milk macronutrient concentrations. In the general
linear models, dietary fat was weakly associated with
milk lactose and protein, which, however, has rarely been
reported. From the models, after adjusting the other factors
associated with milk composition, a 50 g increase of dietary
fat intake resulted in increases of 0.05 g/dl and 0.025 g/dl
in lactose and protein concentrations, respectively. Such
changes do not seem to have a biological significance. As
for dietary energy intake, it did not appear to correlate with
milk energy levels or macronutrients except fat content.
Studies showed that at less than recommended energy
intakes, successful lactation could still be maintained.18,19 In
order to explain the circumstance of normal lactation when
energy intake is limited, compensatory mechanisms have
been suggested which involve adaptions in metabolism,
thermogenesis, and physical activity.20,21
Maternal body composition has also been evaluated for
its effects on milk composition. In some studies, maternal
BMI and triceps skinfold thickness have been shown to be
a positive predictor of milk fat content.14,22,23 Here, we found
a similar relationship. Interestingly, it has been reported
the total amount of fat secreted into milk as well as infant
growth velocities in the lower BMI (BMI <23) group did
not differ from those in the higher BMI group (BMI 23)
due to the inverse relation between milk production and
milk fat concentration.23 In the present study, maternal BMI
was also associated negatively with lactose and positively
with protein content. From the general linear models,
with the other factors being adjusted, a 5 kg/m2 increase
in maternal BMI led to lactose and protein concentration
changes of 0.06 g/dl and 0.035 g/dl, respectively. Again,
no biological significances were indicated from these
associations. Regarding other maternal characteristics, no
associations with milk macronutrients were found at this
time.
Overall, the associations between maternal factors and
milk composition were generally not robust, except for
stage of lactation. Although exerting weakly, maternal
BMI influenced milk composition to a greater extent
than did current dietary intake, suggesting that milk
composition is more susceptible to long-term maternal
nutritional status than short-term diet fluctuation. It is
important to note that the issue of using a single feed

Chinese Medical Journal 2014;127 (9)

milk sample instead of pooled 24-hour samples might

compromise the representativeness of the values and
obscure the associations between maternal factors and
milk composition. Moreover, information of breastfeeding
behaviors such as breastfeeding intensity (exclusive or
partial) and nursing frequency was not available, and their
likely influence on milk composition variation cannot be
ruled out at this time.
In summary, we documented breast milk macronutrients
and energy concentrations of lactating mothers at 5
lactation stages from urban China. Compared with certain
reference values for mature milk, they were different, but
within reported ranges. Consistent with previous results,
milk macronutrient contents changed over the course of
lactation. Maternal current dietary intake showed to have
little effect on milk composition, which is in agreement
with current knowledge. Maternal BMI had a weak positive
association with milk fat content, having been approved by
some previous studies. Our findings support the previous
evidence that current maternal diet or nutritional status
does not have important effects on milk macronutrient
concentrations, resulting in a child being buffered from
fluctuations in maternal nutrition. Further research
documenting longitudinal changes in milk composition
and maternal status is needed during which influencing the
aforementioned factors should be controlled and thoroughly
examined.
REFERENCES
1. Picciano MF. Nutrient composition of human milk. Pediatr Clin
North Am 2001; 48: 53-67.
2. World Health Organization. Infant and young child feeding.
(Accessed July 10, 2010 at http://www.who.int/mediacentre/
factsheets/fs342/en/index.html)
3. Lawrence RA, Lawrence RM. Breastfeeding: a guide for the
medical profession. 7th ed. Maryland Heights, MO: Elsevier
Mosby; 2011: 99-101, 283-292, 107.
4. Menjo A, Mizuno K, Murase M, Nishida Y, Taki M, Itabashi K,
et al. Bedside analysis of human milk for adjustable nutrition
strategy. Acta Pediatr 2009; 98: 380-384.
5. Prentice A, Prentice AM, Whitehead RG. Breast-milk fat
concentrations of rural African women. 2. Long-term variations
within a community. Br J Nutr 1981;45: 495-503.
6. Ruel MT, Dewey KG, Martnez C, Flores R, Brown KH.
Validation of single daytime samples of human milk to estimate
the 24-h concentration of lipids in urban Guatemalan mothers.
Am J Clin Nutr 1997; 65: 439-444.
7. Garza C, Butte NF. Energy concentration of human milk
estimated from 24-h pools and various abbreviated sampling
schemes. J Pediatr Gastroenterol Nutr 1986; 5: 943-948.
8. Daly SE, Di Rosso A, Owens RA, Hartmann PE. Degree of breast
emptying explains changes in the fat content, but not fatty acid

1725
composition, of human milk. Exp Physiol 1993; 78: 741-755.
9. Hytten FE. Clinical and chemical studies in human lactation: II.
Variation in major constituents during a feeding. Br Med J 1954;
1: 176-179.
10. Picciano MF. Representative values for constituents of human
milk. Pediatr Clin North Am 2001; 48: 263-264.
11. Zhou YZ, Shu YX, Lin M, Cheng ZS, Hu FX, Qi ZY, et al.
Nutritional status of lactating mothers and the composition of
their breast milk (in Chinese). Acta Nutrimenta Sinica 1987; 9:
227-233.
12. Miller EM, Aiello MO, Fujita M, Hinde K, Milligan L, Quinn
EA. Field and laboratory methods in human milk research. Am J
Hum Biol 2013; 25: 1-11.
13. Allen JC, Keller RP, Archer P, Neville MC. Studies in human
lactation: milk composition and daily secretion rates of
macronutrients in the first year of lactation. Am J Clin Nutr
1991; 54: 69-80.
14. Brown KH, Akhtar NA, Robertson AD, Ahmed MG. Lactational
capacity of marginally nourished mothers: relationships
between maternal nutritional status and quantity and proximate
composition of milk. Pediatrics1986; 78: 909-919.
15. Mitoulas LR, Kent JC, Cox DB, Owens RA, Sherriff JL,
Hartmann PE. Variation in fat, lactose and protein in human milk
over 24 h and throughout the first year of lactation. Br J Nutr
2002; 88: 29-37.
16. Lnnerdal B. Effects of maternal dietary intake on human milk
composition. J Nutr 1986; 116: 499-513.
17. Subcommittee on Nutrition during Lactation: Committee on
Nutritional Status during Pregnancy and Lactation, Food and
Nutrition Board, Institute of Medicine, National Academy of
Sciences. Nutrition during lactation.Washington, DC: National
Academy Press; 1991: 122-124.
18. Butte NF, Garza C, Stuff JE, Smith EO, Nichols BL. Effect of
maternal diet and body composition on lactational performance.
Am J Clin Nutr 1984; 39: 296-306.
19. Dusdieker LB, Hemingway DL, Stumbo PJ. Is milk production
impaired by dieting during lactation? Am J Clin Nutr 1994; 59:
833-840.
20. Prentice AM, Prentice A. Energy costs of lactation. Annu Rev
Nutr 1988; 8: 63-79.
21. Tigas S, Sunehag A, Haymond MW. Metabolic adaptation to
feeding and fasting during lactation in humans. J Clin Endocrinol
Metab 2002; 87: 302-307.
22. Prentice AM, Goldberg GR, Prentice A. Body mass index
and lactation performance. Eur J Clin Nutr 1994; 48 Suppl 3:
S78-S86.
23. Barbosa L, Butte NF, Villalpando S, Wong WW, Smith
EO. Maternal energy balance and lactation performance of
Mesoamerindians as a function of body mass index. Am J Clin
Nutr 1997; 66: 575-583.

(Received December 19, 2013)

Edited by Chen Limin