Impact of age at menarche on obesity and glycemic control in Japanese patients with type 2 diabetes: Fukuoka Diabetes Registry.

AIMS/
INTRODUCTION: A younger age at menarche is associated with obesity and type 2 diabetes in adult life. The impact of early-onset menarche on obesity and glycemic control in type 2 diabetes has not been investigated. The present study examined the relationship between age at menarche and obesity and glycemic control in type 2 diabetes.
MATERIALS AND METHODS: A total of 2,133 patients with type 2 diabetes aged ≥20 years were divided into groups according to age at menarche (≤11, 12, 13, 14 and ≥15 years). A retrospective cohort study examined the association of menarcheal age with adiposity and hemoglobin A1c .
RESULTS: Age at menarche was inversely associated with body mass index (BMI) and abdominal circumference (P < 0.001). Each 1-year decrease in age at menarche was associated with a 0.25-kg/m2 and 0.6-cm increase in BMI and abdominal circumference, respectively, using a multivariate-adjusted model. Odds ratios for obesity and abdominal obesity significantly increased in participants with age at menarche ≤11 years after multivariable adjustments when age at menarche of 13 years was used as the reference (odds ratio 1.95, 95% CI 1.33-2.88, odds ratio 1.95, 95% CI 1.32-2.87, respectively). Younger age at menarche was significantly associated with higher hemoglobin A1c (P < 0.001); however, the association was not statistically significant after adjusting for BMI.
CONCLUSIONS: Age at menarche of ≤11 years was associated with obesity after adjusting for confounding factors, and poor glycemic control associated with high BMI in type 2 diabetes. Age at menarche should be considered during clinical assessments.

Introduction

There is a trend towards a younger age at which people are reaching puberty, possibly associated with the impact of Westernization. In Japan, the mean age at menarche decreased from 13.2 years in 1961 to 12.2 years in 2011 in school‐aged girls1. Earlier‐onset menarche is reported to be associated with obesity2, 3, 4, 5, type 2 diabetes mellitus4, 6, 7, 8, 9, 10, 11, cardiovascular disease4, 12, 13, 14, breast cancer15 and increased all‐cause mortality4, 13, 16, 17, 18. It is therefore possible that age at menarche can provide important information to help prevent non‐communicable diseases.

Obesity is a global health problem. Although the prevalence of obesity is lower in Japan than Western populations generally (prevalence of body mass index [BMI] ≥30: 3.7% in Japan vs 38.2% in the USA19), there has been an increase in the level of obesity in patients with type 2 diabetes in Japan (mean BMI increased from 24.1 kg/m2 in 2002 to 25.0 kg/m2 in 201320).

It is reported that a younger age at menarche is associated with obesity in later life2, 3, 4, 5. This raises the question as to the association between age at menarche and obesity in patients with type 2 diabetes, and whether early‐onset menarche is associated with plasma glucose levels and could be used to predict the future onset of type 2 diabetes4, 6, 7, 8, 9, 10, 11. Against this background, the present retrospective cohort study examined the relationship between age at menarche and obesity and glycemic control in Japanese patients with type 2 diabetes.

Methods

Study participants

The Fukuoka Diabetes Registry was designed as a prospective, multicenter, observational study to investigate the effects of modern therapy on the prognosis of Japanese patients with diabetes mellitus21. Briefly, the Fukuoka Diabetes Registry cohort comprised patients aged ≥20 years who regularly attended educational research hospitals approved by the Japan Diabetes Society or certified diabetes clinics in Fukuoka Prefecture (UMIN Clinical Trial Registry 000002627) between April 2008 and October 2010. Exclusion criteria were people with drug‐induced diabetes mellitus or receiving corticosteroid therapy; those who had undergone renal replacement therapy; those with serious diseases other than diabetes, such as advanced malignancy or decompensated liver cirrhosis; and those who were unable to visit a diabetologist regularly. Of the 2,263 women registered, after excluding those with type 1 diabetes, 2,133 female patients with type 2 diabetes mellitus remained, and were enrolled in the study. The study was approved by the Kyushu University Institutional Review Board and was carried out in accordance with the provisions contained within the Declaration of Helsinki. All patients provided written informed consent for participation in the study.

Clinical evaluation and laboratory measurements

Participants completed a self‐administered questionnaire that sought information on age at diagnosis of diabetes, family history of diabetes, smoking habits, alcohol intake, physical activity, maximum bodyweight before enrollment, age at menarche and the occurrence of menopause. Smoking habits and alcohol intake were classified as either current use or not. Metabolic equivalent hours per week were calculated using Ainsworth's methods22. A dietary survey was carried out using a brief‐type self‐administered diet history questionnaire (Gender Medical Research, Inc., Tokyo, Japan) that sought information on the frequency of consumption of 58 items to assess the participants’ dietary intakes. The dietary intake estimates for total energy and dietary fiber were calculated using an ad hoc algorithm developed for the brief‐type self‐administered diet history questionnaire based on the Standard Tables of Food Composition in Japan23. The medical records of participants were examined for medications including oral hypoglycemic agents and insulin therapy. Bodyweight and height were measured, and BMI was calculated. Obesity was defined as BMI ≥25 kg/m2 according to the Japan Society for the Study of Obesity criteria24. Abdominal circumference at the umbilical level was measured in the standing position. Abdominal obesity was defined as an abdominal circumference of ≥90 cm according to the Japan Society for the Study of Obesity criteria24.

Blood tests were carried out in either the fasting or postprandial state. Hemoglobin A1c (HbA1c) was determined by high‐performance liquid chromatography (Tosoh Corp., Tokyo, Japan), and serum C‐peptide was determined using chemiluminescent immunoassays (Kyowa Medex, Tokyo, Japan). High‐sensitivity C‐reactive protein (HS‐CRP) and serum adiponectin were determined by latex immunonephelometry (Siemens Healthcare Diagnostics, Tokyo, Japan; Mitsubishi Chemical Medience, Tokyo, Japan). β‐Cell function (homeostatic model assessment of β‐cell [HOMA2‐%β] index) and insulin resistance (homeostatic model assessment of insulin resistance [HOMA2‐IR] index) were estimated based on fasting glucose and C‐peptide concentrations using a HOMA calculator (version 2.2.2; http://www.dtu.ox.ac.uk, accessed June 2012) after exclusion of individuals with unacceptable levels of plasma glucose (<3.0 mmol/L or >25 mmol/L) or C‐peptide (<0.2 nmol/L or >3.5 nmol/L).

Statistical analysis

HOMA2‐%β, HOMA2‐IR, serum adiponectin and HS‐CRP were log‐transformed for statistical analyses because of their skewed distributions. They are presented with 95% confidence intervals (95% CI) that were back‐transformed. Participants were divided into five categories according to age at menarche: ≤11, 12, 13, 14 and ≥15 years, as reported in previous studies4, 6, 7, 9. Baseline characteristics by categories of age at menarche were explored using one‐way analysis of variance and analysis of covariance to compare means of continuous variables, and using χ2‐tests to compare proportions of categorical variables. Trend associations for age at menarche were assessed using multivariable regression analyses, and included age, duration of diabetes, current smoking, current drinking, leisure‐time physical activity, daily energy intake, oral hypoglycemic agent use and insulin use as covariates. Associations between age at menarche and the prevalence of obesity, history of obesity and abdominal obesity were tested using multivariable logistic regression analyses. All analyses were carried out using the SAS software package (version 9.4; SAS Institute Inc., Cary, NC, USA). A P‐value <0.05 was considered statistically significant.

Results

Participant characteristics

Baseline characteristics of the study participants by categories of age at menarche are shown in Table 1. The mean age at menarche was 13.7 ± 1.8 years. Those with a younger age at menarche were diagnosed with diabetes at a younger age, had a shorter duration of diabetes, were more likely to be premenopausal, had lower energy and dietary fiber intake, and exercised less. These statistical differences were absent after adjusting for age.

Table 1

Clinical characteristics of the study participants by categories of age at menarche

	Total	Categories of age at menarche (years)					P for trend
		≤11	12	13	14	≥15	Unadjusted	Age‐adjusted
Age at menarche (years)	13.7 (13.0–15.0)	10.7 (10.0–11.0)	12.0 (12.0–12.0)	13.0 (13.0–13.0)	13.9 (14.0–14.0)	15.8 (15.0–16.0)
n	2,133	169	355	475	523	611
Age (years)	65.8 ± 10.4	53.2 ± 11.4	61.1 ± 10.5	64.8 ± 9.8	67.7 ± 8.1	71.3 ± 7.7	<0.001	–
Age at diabetes diagnosis (years)	51.0 ± 11.9	40.3 ± 12.9	47.5 ± 11.5	50.9 ± 11.1	52.3 ± 10.7	55.0 ± 11.1	<0.001	NS
Duration of diabetes (years)	14.3 ± 9.8	12.4 ± 9.2	13.0 ± 9.5	13.4 ± 8.9	14.9 ± 10.2	15.8 ± 10.1	<0.001	NS
Family history of diabetes (%)	61.5	68.6	65.6	60.2	61.0	58.6	0.07	NS
Menopause (%)	92.3	63.9	86.8	92.6	97.3	98.7	<0.001	NS
Age at menopause (years)	49.4 ± 5.3	48.9 ± 5.4	49.5 ± 5.0	49.4 ± 5.0	49.4 ± 5.4	49.4 ± 5.4	NS	NS
Reproductive years	35.5 ± 5.5	38.2 ± 5.4	37.5 ± 5.0	36.4 ± 5.0	35.4 ± 5.4	33.5 ± 5.6	<0.001	<0.001
Current smoker (%)	6.8	10.7	6.2	6.1	7.8	5.9	NS	NS
Current drinker (%)	16.5	17.2	18.3	16.6	17.4	14.2	NS	NS
Energy intake (kcal/day)	1,509 ± 412	1,479 ± 384	1,510 ± 466	1,491 ± 392	1,510 ± 407	1,531 ± 407	<0.05	NS
Dietary fiber intake (g/1,000 kcal)	8.3 ± 2.2	7.6 ± 2.1	8.1 ± 2.0	8.4 ± 2.4	8.4 ± 2.1	8.4 ± 2.1	<0.001	NS
LTPA (MET h/week)	16.3 ± 16.4	13.4 ± 14.0	16.4 ± 20.1	16.0 ± 14.4	17.0 ± 16.1	16.8 ± 16.4	<0.05	NS
Oral hypoglycemic agents (%)	63.5	63.9	64.8	62.7	60.6	65.8	NS	NS
Insulin (%)	30.8	36.7	32.1	29.7	30.0	29.8	NS	NS

Data are mean ± standard deviation or percentage. Numbers in parenthesis represent interquartile range. LTPA, leisure‐time physical activity; NS, not significant.

Associations between age at menarche and obesity

The age at menarche was inversely associated with BMI (P for trend <0.001; Table 2). The association remained statistically significant after multivariable adjustments including age, duration of diabetes, current smoking, current drinking, leisure‐time physical activity, daily energy intake, oral hypoglycemic agent use and insulin use. The increase in BMI per 1‐year earlier onset of menarche was 0.42 kg/m2 in the crude and 0.25 kg/m2 in the multivariate‐adjusted models. An inverse association between age at menarche and BMI was evident in both participants aged <65 years and ≥65 years. Maximum BMI was significantly associated with age at menarche in all participants and those aged <65 years and ≥65 years. The increase in maximum BMI per 1‐year earlier onset of menarche was 0.43 kg/m2 in the crude and 0.33 kg/m2 in the multivariate‐adjusted models. Abdominal circumference was inversely associated with age at menarche. The association was significant in participants aged <65 years, whereas in those aged ≥65 years, it was significant after multivariate adjustments. The increase in abdominal circumference per 1‐year earlier onset of menarche was 0.5 cm in the crude and 0.6 cm in the multivariate‐adjusted models.

Table 2

Association of categories of age at menarche with adiposity, stratified by age in Japanese patients with type 2 diabetes

	Categories of age at menarche (years)					P for trend
	≤11	12	13	14	≥15	Unadjusted	Multivariate‐adjusted
Age at menarche (years)	10.7 (10.0–11.0)	12.0 (12.0–12.0)	13.0 (13.0–13.0)	13.9 (14.0–14.0)	15.8 (15.0–16.0)
BMI (kg/m2)
All	26.3 ± 6.1	24.7 ± 4.6	24.1 ± 4.5	23.4 ± 3.7	23.4 ± 3.8	<0.001	<0.001
<65 years	26.6 ± 6.3	25.2 ± 4.7	24.3 ± 5.0	23.5 ± 3.8	23.9 ± 4.3	<0.001	<0.001
≥65 years	24.5 ± 4.7	23.5 ± 3.9	23.9 ± 3.9	23.4 ± 3.6	23.2 ± 3.7	<0.05	<0.05
Maximum BMI (kg/m2)
All	29.7 ± 6.3	28.2 ± 4.9	27.4 ± 4.7	26.7 ± 3.8	26.7 ± 3.9	<0.001	<0.001
<65 years	29.9 ± 6.5	28.6 ± 5.3	27.4 ± 5.4	26.6 ± 3.8	26.8 ± 4.4	<0.001	<0.001
≥65 years	28.0 ± 5.3	27.4 ± 3.9	27.4 ± 3.9	26.8 ± 3.8	26.7 ± 3.8	<0.01	<0.01
Abdominal circumference (cm)
All	89.0 ± 14.3	87.0 ± 11.7	86.0 ± 12.4	84.8 ± 10.5	85.3 ± 11.2	<0.001	<0.001
<65 years	89.3 ± 14.8	87.3 ± 12.3	84.7 ± 13.0	83.6 ± 10.9	84.0 ± 12.1	<0.001	<0.001
≥65 years	87.1 ± 10.9	86.2 ± 10.3	87.2 ± 11.6	85.5 ± 10.2	85.6 ± 11.0	NS	<0.05

Data are mean ± standard deviation. Numbers in parenthesis represent interquartile range. Multivariable adjustments include age, duration of diabetes, current smoking, current drinking, leisure‐time physical activity, daily energy intake, oral hypoglycemic agent use and insulin use. BMI, body mass index; NS, not significant.

In all the participants, the prevalence of obesity was inversely related to the age at menarche, after adjusting for confounding factors (Table 3). Similarly, a history of obesity and abdominal obesity were more common in individuals who underwent menarche at an earlier age. There was a significant association between the prevalence of obesity and age at menarche in participants both aged <65 years and ≥65 years. Regarding a history of obesity and abdominal obesity, the association was significant in participants aged <65 years, but not in those aged ≥65 years. When 13 years‐of‐age was used as the reference, odds ratios (OR) for obesity, history of obesity and abdominal obesity significantly increased in those with age at menarche ≤11 years after multivariable adjustments (OR 1.95, 95% CI 1.33–2.88 for obesity; OR 2.01, 95% CI 1.31–3.15 for history of obesity; and OR 1.95, 95% CI 1.32–2.87 for abdominal obesity). In participants aged <65 years, the ORs for obesity, history of obesity and abdominal obesity were significantly higher in those who experienced menarche at ≤11 years and 12 years, whereas there were no significant differences in those aged ≥65 years.

Table 3

Association of categories of age at menarche with obesity in Japanese patients with type 2 diabetes

	Categories of age at menarche (years)					P for trend
	≤11	12	13	14	≥15
Age at menarche (years)	10.7 (10.0–11.0)	12.0 (12.0–12.0)	13.0 (13.0–13.0)	13.9 (14.0–14.0)	15.8 (15.0–16.0)
Prevalence (%)
All
Obesity	55.0	40.3	35.0	29.8	28.3	<0.001
History of obesity	78.1	74.9	67.0	64.2	64.2	<0.001
Abdominal obesity	45.6	37.5	32.2	31.6	31.9	0.002
<65 years
Obesity	58.2	44.0	33.0	28.7	36.7	<0.001
History of obesity	81.5	75.3	63.5	62.4	61.5	<0.001
Abdominal obesity	46.6	39.5	25.7	29.3	30.3	<0.001
≥65 years
Obesity	34.8	32.1	36.7	30.4	26.5	0.014
History of obesity	56.5	74.1	70.2	65.2	64.7	0.11
Abdominal obesity	39.1	33.0	38.4	32.8	32.3	0.24
Multivariate‐adjusted OR
All
Obesity	1.95 (1.33–2.88)	1.19 (0.88–1.59)	1.00 (ref)	0.85 (0.64–1.12)	0.79 (0.60–1.03)	<0.001
History of obesity	2.01 (1.31–3.15)	1.57 (1.15–2.15)	1.00 (ref)	0.89 (0.68–1.17)	0.82 (0.63–1.07)	<0.001
Abdominal obesity	1.95 (1.32–2.87)	1.30 (0.97–1.75)	1.00 (ref)	0.97 (0.74–1.27)	0.91 (0.69–1.19)	0.002
<65 years
Obesity	2.78 (1.76–4.41)	1.58 (1.07–2.32)	1.00 (ref)	0.85 (0.54–1.31)	1.30 (0.79–2.14)	<0.001
History of obesity	2.98 (1.77–5.13)	1.81 (1.21–2.73)	1.00 (ref)	0.95 (0.62–1.44)	0.95 (0.58–1.55)	<0.001
Abdominal obesity	2.70 (1.69–4.35)	1.93 (1.29–2.89)	1.00 (ref)	1.21 (0.77–1.90)	1.31 (0.78–2.20)	<0.001
≥65 years
Obesity	0.88 (0.34–2.17)	0.80 (0.48–1.30)	1.00 (ref)	0.77 (0.54–1.01)	0.62 (0.44–0.87)	0.022
History of obesity	0.56 (0.23–1.37)	1.24 (0.75–2.10)	1.00 (ref)	0.82 (0.57–1.17)	0.74 (0.52–1.03)	0.11
Abdominal obesity	1.06 (0.42–2.54)	0.78 (0.48–1.25)	1.00 (ref)	0.80 (0.56–1.13)	0.72 (0.52–1.00)	0.37

Data are mean ± standard deviation, percentage or odds ratios (OR). Numbers in parenthesis represent interquartile range or 95% confidence interval. Obesity was defined as body mass index ≥25 kg/m2 and abdominal obesity as abdominal circumference ≥90 cm according to the Japan Society for the Study of Obesity criteria22. Multivariable adjustments include age, duration of diabetes, current smoking, current drinking, leisure‐time physical activity, daily energy intake, oral hypoglycemic agent use and insulin use.

Associations between age at menarche and glycemic control

Age at menarche was inversely associated with HbA1c (P < 0.001). The increase in HbA1c per 1‐year earlier onset of menarche was 0.04% in the crude and 0.03% in the multivariate‐adjusted models. However, the statistical significance was attenuated by age or multivariate adjustments, and absent with adjustment for BMI (Table 4). Analysis of glycemic control in participants aged <65 years yielded similar results, but there was no association in those aged ≥65 years. The insulin secretion parameter, HOMA2‐%β, did not differ among the age at menarche categories. HOMA2‐IR increased and serum adiponectin decreased at earlier onset of menarche (P < 0.001), although age adjustment removed the statistical significance. HS‐CRP, a marker of systematic microinflammation, was inversely associated with age at menarche to a marginal degree in the age‐ and multivariate‐adjustment models. For HOMA2‐%β, HOMA2‐IR, serum adiponectin and HS‐CRP, similar results were obtained in participants aged <65 years, whereas there were no associations in those aged ≥65 years.

Table 4

Association of categories of age at menarche with glycemic control, insulin secretion, insulin resistance, adiponectin and microinflammation in Japanese patients with type 2 diabetes

	Categories of age at menarche (years)					P for trend
	≤11	12	13	14	≥15	Unadjusted	Multivariate	Multivariate + BMI
Age at menarche (years)	10.7 (10.0–11.0)	12.0 (12.0–12.0)	13.0 (13.0–13.0)	13.9 (14.0–14.0)	15.8 (15.0–16.0)
HbA1c (%)
All	7.77 ± 1.29	7.63 ± 1.16	7.59 ± 1.09	7.48 ± 0.97	7.47 ± 1.02	<0.001	<0.05	NS
<65 years	7.78 ± 1.31	7.72 ± 1.27	7.62 ± 1.12	7.49 ± 0.99	7.59 ± 1.18	<0.05	0.050	NS
≥65 years	7.70 ± 1.16	7.45 ± 0.85	7.56 ± 1.06	7.47 ± 0.97	7.45 ± 0.99	NS	NS	NS
HbA1c (mmol/mol)
All	59.8 ± 13.1	58.4 ± 11.8	58.0 ± 11.1	56.8 ± 9.9	56.8 ± 10.4	<0.001	<0.05	NS
<65 years	59.9 ± 13.3	59.2 ± 12.9	58.3 ± 11.3	56.9 ± 10.0	57.9 ± 12.0	<0.05	0.050	NS
≥65 years	59.1 ± 11.8	56.6 ± 8.7	57.7 ± 10.8	56.7 ± 9.8	56.5 ± 10.1	NS	NS	NS
HOMA2‐%β
All	39.6 (36.5–43.0)	38.9 (36.8–41.1)	39.5 (37.6–41.6)	40.8 (38.9–42.8)	41.1 (39.4–43.0)	NS	NS	NS
<65 years	42.3 (34.3–52.1)	37.6 (34.3–41.3)	39.6 (36.9–42.4)	41.6 (39.4–44.1)	41.4 (39.5–43.4)	NS	NS	NS
≥65 years	39.2 (35.7–43.0)	39.5 (36.8–42.4)	39.5 (36.7–42.5)	39.4 (36.2–42.8)	39.9 (35.8–44.5)	NS	NS	NS
HOMA2‐IR
All	1.12 (1.05–1.20)	1.02 (0.97–1.06)	0.98 (0.94–1.02)	0.95 (0.91–0.99)	0.98 (0.94–1.01)	<0.001	NS	NS
<65 years	1.14 (1.06–1.23)	1.06 (1.00–1.13)	0.99 (0.93–1.05)	0.96 (0.90–1.03)	1.03 (0.95–1.13)	<0.05	NS	NS
≥65 years	1.04 (0.88–1.24)	0.93 (0.86–1.00)	0.97 (0.92–1.03)	0.94 (0.90–0.99)	0.97 (0.93‐1.00)	NS	NS	NS
Adiponectin (μg/mL)
All	9.0 (8.2–9.8)	10.0 (9.5–10.7)	10.6 (10.1–11.2)	10.7 (10.2–11.2)	11.4 (10.9–11.9)	<0.001	NS	NS
<65 years	8.9 (8.1–9.8)	9.3 (8.6–10.0)	10.4 (9.6–11.2)	10.6 (9.7–11.5)	9.8 (8.7–10.9)	<0.05	NS	NS
≥65 years	9.6 (7.7–12.0)	11.9 (10.8–13.2)	11.0 (10.2–11.7)	10.7 (10.1–11.4)	11.8 (11.2–11.4)	NS	NS	NS
HS‐CRP (mg/L)
All	0.64 (0.52–0.79)	0.49 (0.42–0.56)	0.48 (0.43–0.55)	0.50 (0.44–0.56)	0.50 (0.45–0.56)	NS	<0.05	NS
<65 years	0.67 (0.54–0.84)	0.49 (0.41–0.58)	0.42 (0.35–0.50)	0.45 (0.37–0.55)	0.48 (0.37–0.62)	<0.05	NS	NS
≥65 years	0.48 (0.28–0.82)	0.49 (0.38–0.63)	0.56 (0.47–0.66)	0.53 (0.45–0.61)	0.51 (0.45–0.57)	NS	NS	NS

Data are mean ± standard deviation, percentage or odds ratios. Numbers in parenthesis represent interquartile range or 95% confidence. Multivariable adjustments include age, duration of diabetes, current smoking, current drinking, leisure‐time physical activity, daily energy intake, oral hypoglycemic agent use and insulin use. BMI, body mass index; HbA1c, hemoglobin 1c; HS‐CRP, high‐sensitivity C‐reactive protein; HOMA2‐%β homeostasis model assessment of β‐cell function; HOMA2‐IR, homeostasis model assessment of insulin resistance; NS, not significant.

Discussion

In the present study, age at menarche was inversely associated with adiposity in patients with type 2 diabetes. To the best of the authors’ knowledge, this is the first study to report on the association between age at menarche in patients with type 2 diabetes. It is reported that earlier‐onset menarche is associated with obesity and the development of type 2 diabetes mellitus in later life2, 3, 4, 5, 6, 7, 8, 9, 10, 11. The present results extend the association between age at menarche and obesity to patients with type 2 diabetes. Although people with obesity and type 2 diabetes are strongly urged to modify their lifestyle habits, including diet and exercise, there were twice as many patients in the ≤11 years‐of‐age at menarche group compared with the 13 years‐of‐age at menarche group. It is possible that obesity contributes to worsened glycemic control in those with age at menarche ≤11 years.

In a meta‐analysis of 10 studies including 246,671 women12, mostly from Western populations, early menarche (<12 vs 12 years) was associated with 0.34 kg/m2 higher BMI. Of 303,000 women in the China Kadoorie Biobank, increases in BMI and abdominal circumference per 1‐year earlier onset of menarche were reported to be 0.19 kg/m2 and 0.38 cm, respectively5. In the current study, increases in BMI and abdominal circumference per 1‐year earlier onset of menarche were 0.25 kg/m2 and 0.6 cm in the multivariate‐adjusted model. It appears that the impact of earlier‐onset menarche on obesity is greater in the current study. This difference might be explained by the populations studied. For example, the general population vs patients with type 2 diabetes, or Western populations vs Asian populations, in whom those with type 2 diabetes are not typically obese.

The association between age at menarche and BMI was observed in both the <65 years and ≥65 years age groups. One of the mechanisms that has been proposed to explain the association between early age at menarche and obesity is that earlier‐onset menarche might result in longer exposure to estrogen25 and adrenal steroids26, which tend to maintain adiposity. Furthermore, an overlap between single‐nucleotide polymorphisms implicated in the timing of puberty and in determining BMI in adulthood has been reported25. These single‐nucleotide polymorphisms include TCF, which was shown to be a risk factor for obesity and type 2 diabetes27, and LIN28B, which is associated with insulin sensitivity and oxidative stress‐related β‐cell apoptosis28. Individuals who have these single‐nucleotide polymorphisms tend to gain weight faster during infancy and early childhood, and show earlier‐onset menarche29.

Earlier age at menarche is associated with the future development of type 2 diabetes in Western and Asian populations4, 6, 7, 8, 9, 10, 11, and was associated with elevated HbA1c, secondary to high BMI in the current study. These associations can be largely explained by increased adiposity. In the current study, earlier age at menarche was not associated with insulin secretion, but was associated with increased insulin resistance, reduced serum adiponectin and increased systemic microinflammation (Table 4). However, adjusting for age and BMI removed these associations, being in line with HbA1c.

Analyses stratified according to age (<65 years and ≥65 years) showed that there were significant associations between age at menarche and history of obesity or abdominal obesity in participants aged <65 years, but not in those aged ≥65 years (Table 3). This might be due to the small number of participants who went through menarche at ≤11 years (n = 23) among those aged ≥65 years. In addition, there were no significant interactions between the age of the participant (<65 and ≥65 years) and their age at menarche.

The present study had some limitations. First, there is the possibility of recall bias regarding age at menarche, although previous studies report that recalling age at menarche is reliable over many years27. Second, we could not clarify the influence of childhood or pubertal BMI on adulthood obesity because of a lack of information. Finally, the conclusions of the study should not be generalized to other ethnic populations, especially with high BMI, without caution.

In conclusion, age at menarche of ≤11 years was associated with obesity after adjusting for confounding factors, and poor glycemic control associated with BMI in type 2 diabetes. As obesity can accelerate the development and progression of diabetic complications, age at menarche should be a factor for consideration when determining clinical management of patients with obese type 2 diabetes.

Disclosure

The authors declare no conflict of interest.