Abstract
Background
Lifestyle modification interventions among women with gestational diabetes mellitus (GDM) history have shown to be effective on preventing or delaying the onset of diabetes, however, the differentiating benefits of lifestyle interventions due to socioeconomic inequalities have remained understudied. The Intensive LifeStyle Modification Program (ILSM) was tailored for women with a history of GDM in low-resource rural areas of China. The current study aimed to examine the effect of the ILSM intervention on physiological and behavioral outcomes between subgroups of women from different socioeconomic backgrounds.
Methods
This study used the baseline, 3-month, 6-month, and 18-month data from a cluster randomized controlled trial in rural women with GDM history in Hunan Province of China. The lifestyle program consisted of a combination of in-person group sessions and telephone consultation sessions on dietary intake, physical activity, and stress management. Using generalized estimating equation, we conducted several subgroup analyses based on indicators of socioeconomic status, including age, ethnicity, income, employment, and education.
Results
A total of 320 women were analyzed. The ILSM intervention showed a consistent significance on reducing fasting blood glucose across subgroups. On the contrary, women with full-time employment and high income (≥3000RMB, approximately 425 dollars, per month) were reported to have greater benefits on intentions to eat low-glycemic foods, BMI and waist circumference, and diabetes risk. Women with younger age (≤35 years) or high education (15 years of education or above) were more likely to benefit on improving higher intentions to eat low-glycemic foods and lowering diabetes risk. Interestingly, compared to ethnic majority, ethnic minority women were more likely to increase intentions to eat low-glycemic foods.
Conclusion
The ILSM program demonstrated a consistent significance on reducing fasting blood glucose across women from different socioeconomic backgrounds along with an increase of intentions to eat low-glycemic foods for ethnic minority women, despite of showing ineffective in lower socioeconomic status women on intention to have healthy diet, BMI, waist circumference, and diabetes risk. To fully address health disparities from socioeconomic inequalities, lifestyle interventions should be tailored for individuals with older age, lower education levels, and lower income without full-time employment.
Trial registration
Chinese Clinical Trial Registry (ChiCTR2000037956), registered on 3rd Jan 2018.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-026-26854-y.
Keywords: Gestational diabetes mellitus (GDM), postpartum lifestyle intervention, health inequality, rural women
Introduction
Type 2 diabetes mellitus (T2DM) is one of the major actual public health issues consisting of chronic hyperglycemia that can damage body organs and systems. Diabetes affects approximately 425 million people worldwide in 2021 that is projected to 853 million by 2050 [1]. As an independent T2DM risk factor, gestational diabetes mellitus (GDM), defined as a glucose intolerance onset during pregnancy [2], indicates a seven-fold risk of developing T2DM in women in the following years after the pregnancy with GDM [3]. International Diabetes Federation (IDF) recommends lifestyle modification for individuals with elevated T2DM risk considering its low cost and scalability without side effects [4]. In women with GDM history, several large landmark clinical trials such as the Diabetes Prevention Program Outcomes Study have shown the effect of lifestyle interventions on preventing or delaying the onset of T2DM [5].
However, previous lifestyle interventions on diabetes prevention remain understudied in terms of its potential differentiating benefits due to socioeconomic inequalities. Health outcomes are strongly correlated with social and economic position in societies with evidence showing that individuals from deprived backgrounds tend to have a shorter life expectancy as well as a shorter disability-free life expectancy [6]. Socioeconomic inequalities exist in the prevention, diagnosis, and treatment of diabetes. The rate of self-reported diabetes was higher and also increasing more rapidly over time among individuals with lower education level than those with higher education level [7]. Further, better glycemic control, indicated by lower HbA1c levels, was strongly associated with indicators of higher socioeconomic position based on indicators of education, employment status, and equivalent household income [8]. The failure to consider potential differences in intervention effects across equity factors is one of the limitations to inform policy and practice decisions [9]. Therefore, intervention on preventing and managing diabetes needs to consider the challenge of health disparities in intervention benefits due to socioeconomic inequalities from income, employment, and education.
There have been limited interventions considering socioeconomic aspects of diabetes prevention [10]. A project from Finland showed that the effect of their lifestyle intervention (individual visits and group sessions on weight management and maintenance) on anthropometric and clinical outcomes (including BMI and waist circumference) was similar in all socioeconomic groups [11]. Another Germany project supported that the effect of their lifestyle intervention (online evidence-based patient information system for individual decision aid) on diabetes-related knowledge was not significantly impacted by socioeconomic variables [12]. However, the study from the United States reported that significant socioeconomic disparities exist in weight and behavior outcomes of their lifestyle intervention (the Lifestyle Balance Curriculum in group sessions), for example, lower household income was associated with less BMI reduction after the curriculum [13]. While these interventions implemented lifestyle interventions in different formats and reported inconsistent results regarding their similar or differentiating effects according to socioeconomic positions, more efforts are needed to reduce inequalities in diabetes prevention.
China has the largest population of people living with diabetes that increases from 98.4 million to 140.9 million during 2013 to 2021 [14]. Health disparity in diabetes care have been identified in rural areas of China. Undiagnosed T2DM prevalence was higher in rural areas than urban areas (70.5% vs. 58.0%) [15] due the lower level of medical expenditures for diabetes prevention and treatment in urban residents, with about six health care providers per thousand in rural areas compared to 16 per thousand in urban areas. In addition, due to the lower level of health literacy, rural residents tend to have difficulties in adopting and maintaining recommended behavior changes after lifestyle interventions, which may explain the rural-urban inequality of obtained benefit from lifestyle interventions [16]. To promote health equity in diabetes care, adapted lifestyle interventions are needed for rural residents with socioeconomic disadvantages in China.
Guided by the Social Cognitive Theory, our research team contextually tailored the Intensive LifeStyle Modification Program (ILSM) for women with a history of GDM in low-resource rural areas of China [17]. Our previous findings have shown the significant effect of the ILSM on reducing diabetes risk and promoting health behaviors at 6-monoth [18] and 18-month post-intervention [19], respectively. To fully understand whether the ILSM managed to address the challenge of socioeconomic inequalities in rural women across time, the current study aimed to determine the differential benefits of the ILSM intervention through comparing the effect of ILSM intervention on physiological and behavioral outcomes (fasting blood glucose, 2-hour oral glucose tolerance test, BMI, waist circumference, T2DM risk, intention to eat low glycemic foods, and physical activity) between women from different socioeconomic status.
The most frequently used indicator of SES included income, education, and employment, which lead to health gaps through different obedience of health-related guideline [20, 21]. Specifically, higher income largely represents more financial freedom that affect the behaviors of people [22]. Following the similar mechanism, the unemployed may have disadvantages in modify their lifestyle compared to full-time workers [23]. Diverse levels of education background will endow distinctive cognitive and executive abilities of health [24]. In this study, we also included age and ethnicity to indicate different SES in the Chinese population. Growing in different social, cultural, and economic environment, people from different generations may hold various attitudes to health and vary in basic health condition [25]. Compared to ethnic majority Chinese (Han), ethnic minority Chinese have obviously unique culture and lifestyle habits that may be hard to change and affect their health benefit gained from the intervention programs [20]. Therefore, we used monthly income, education, employment status, age and ethnicity as indicators of socioeconomic status (SES). Here are our hypotheses with each indicator of SES on the differentiating effect of the ILSM intervention on health outcomes.
H1: Women with higher income gain more health benefit from the ILSM intervention.
H2: Women with higher education levels gain more health benefit from the ILSM intervention.
H3: Women with full-time employment gain more health benefit from the ILSM intervention compared with people without full-time employment.
H4: Women who are younger gain more health benefit from the ILSM intervention.
H5: Ethnic minority women gain less health benefit from the ILSM intervention.
Methods
Sample and participants
This study used the baseline, 3-month, 6-month, and 18-month data of the Intensive LifeStyle Modification Program (ILSM), which was a cluster randomized controlled trial (No. ChiCTR1800015023) in women with a history of GDM in rural areas of Hunan province during 2018 to 2019. The study design and data collection of the ILSM project are described in detail elsewhere.
Women were eligible if they met the following inclusion criteria: (a) Women with a history of GDM; (b) 18 years or older; (c) At the beginning of the study, at least within six weeks to ten years postpartum; (d) Have telephone-access either from family members, friends, or neighbors; © Have township household registration, intend to live in the research counties for at least 3 years. The exclusion criteria were as follows: (a) currently pregnant or planning to become pregnant within the next three years, (b) diabetes diagnosed before pregnancy or after delivery, (c) medications that influence glucose metabolism, (d) physical or cognitive disability. Data was collected at General Hospital at Youxian County and Maternal and Children’s Hospital at Yongding County. Trained research assistants were available to answer questions and check each questionnaire for unintentionally missing items or pages [17]. Anthropometric measures (height, weight, and waist circumference), fasting blood glucose and the 2-hour 75 g oral glucose tolerance test was administered by nurses. Figure 1 shows the sample selection process.
Fig. 1.
Flow chart of sample selection. Note: Schedule conflicts refer to situations in which participants were unable to attend the planned intervention sessions due to overlapping work, family, or medical appointments
Ethical consideration
Ethical approval was obtained from all local ethics committees. Written informed consent detailing the purposes, data confidentiality, benefits, and risks of the present study were provided by participants prior to data collection. All participant data was deidentified and accessible only to the research team members.
ILSM intervention group
During the ILSM, women in the intervention group participated in an 18-month structured lifestyle intervention program, which was aimed to reduce type 2 diabetes risk among rural women. The lifestyle program consisted of a combination of orientation and goal setting, healthy eating patterns, physical activities, stress management, family support on ILSM & family lifestyle patterns, and farewell and relapse prevention. The ILSM program included six bi-weekly in-person sessions, five bi-weekly telephone booster sessions, and 3 monthly telephone consultations to enhance maintenance of behavior change. The ILSM participants would be asked to attend bi-weekly in-person sessions at the research sites provided by project field officers and would also receive telephone consultations from them. To reduce time in traveling and to increase motivation and adherence to behavioral change recommendations, the telephone consultation would occur one week after each in-person meeting. Therefore, ILSM participants attended in-person meeting at week 1 followed by telephone booster session in week 2 then again in person meeting in week 3 followed by telephone booster session in week 4 and so on. The time commitment was about 60 min for each in-person session and 20 min for each telephone booster. The six in-person sessions were group sessions with 16–20 participants. After the intensive sessions, the participants went to the maintenance phase of intervention, which involved telephone consultation sessions monthly for three months. The maintenance telephone consultation lasted no more than 20 min per consultation. We provide $25 for each participant if they finish the all of the in-person sessions as an appreciation.
Usual care group
The usual care provided in this study adhered to current clinical guidelines and recommendations for preventing T2DM. This included educating participants on diabetes risk factors, emphasizing the significance of lifestyle modifications, and advising diabetes screening every 3 years during the first clinic visit after delivery. Women in the control group received a booklet focused on diabetes prevention. All the anthropometric measurements were conducted specifically for our research purpose.
Measures
Data were collected at baseline, 3-month, 6-month, and 18-month. Indicators of socioeconomic status were collected at baseline. Glucose intolerance, BMI and waist circumference, T2DM risk and dietary and physical activity at baseline, 3-month, 6-month, and 18-month.
Indicators of socioeconomic status
Participants’ socio-demographic information, including age, ethnicity, education, income, and employment status, was collected through a baseline questionnaire and categorized as follows. Age was categorized as < 35 years old and ≥ 35 years old, with 35 years old being the common definition of advanced maternal age. Education level was divided based on years of education: 9 years or less (equivalent to completing only China’s compulsory education or even did not finish a compulsory education), 12 years (possessing a junior high school degree), and 15 years or more (holding an associate degree or higher). Income was categorized as < 3,000RMB and ≥ 3,000RMB, with 3,000RMB (approximately 425 dollars) being the poverty threshold in China. Ethnicity (Han majority and ethnic minority) and employment status (full-time job and no full-time) were also categorized. In this study, participants self-reported their ethnicity during data collection. Income was measured at the household level, reflecting the combined earnings of all household members. Full-time employment was defined as individuals working under a signed legal labor contract with the standard number of hours required by national regulations, whereas all other employment types were categorized as non-full time.
Outcomes of physiological health
Glucose intolerance
Participants should eat and drink normally the day before, avoid strenuous physical activities, fast for 10–14 h before blood sampling, and then drew blood to test fasting blood glucose (FBG). Next, 300 milliliters of sugar water containing 75 g of glucose was consumed within 5 min, and blood was drawn 2 h later to measure 2-hour oral glucose tolerance test (2 h-OGTT). Plasma glucose was measured using a hexokinase enzyme method. FBG and 2 h-OGTT are used for the diagnosis of pre-diabetes and diabetes. According to the International Diabetes Federation guide criterion: FPG < 6.1 mmol/L and 2 h-OGTT < 7.8 mmol/L as normal, 6.1 mmol/L < FPG < 7.0 mmol/L and/or 7.8 mmol/L ≤ 2 h-OGTT < 11.1 mmol/L as pre-diabetes, and FPG 7.0 mmol/L and/or 2 h-OGTT ≥ 11.1 mmol/L as diabetes [1].
BMI and waist circumference. Anthropometric parameters (body weight, height and waist circumference) were measured and recorded by trained investigators who were graduate students of public health and nurses. Height was measured to the nearest 0.5 cm, and weight (with an empty stomach, light indoor clothing, and no shoes) was measured with a precision of 0.1 kg. Waist circumference was measured on a horizontal plane, midway between the inferior margin of the ribs and the superior border of the iliac crest. Body mass index (BMI) was calculated by dividing the weight (in kilograms) by height (in meters squared). According to Chinese guidelines for the prevention and control of overweight and obesity in adults, BMI ≥ 28 kg/m2 indicates obesity, BMI between 24 and 27.9 kg/m2 shows overweight, BMI between 18.5 and 23.9 kg/m2 shows normal, and less than 18.5 kg/m2 as underweight.
T2DM risk
The Chinese Diabetes Risk Scale (CHINARISK) was used to measure participants’ T2DM risk at baseline, 3-month, 6-month, and 18-month. The scale systematically combines modifiable and non-modifiable diabetes risk factors to identify people who are likely to develop T2D in the next 10 years, which was adapted from the Canadian Diabetes Risk Questionnaire. The test–retest reliability was 0.988. The sensitivity was 73%, with a positive predictive value of 57% and negative predictive value of 78%. The scale has 14 items, and the total score theoretically ranged from 0 to 88, with higher scores representing greater 10-year risk of T2DM, and the cutoff score was 30.
Outcomes of weight-related health intentions and behaviors
Dietary and physical activity
The Chinese version of Intentions to Eat Low-GI (IEL) Foods questionnaire was used to measure participants’ intention to eat low glycemic-index foods at baseline, 3-month, 6-month, and 18-month. The instrument has 24 items where responses are given on a seven-point Likert scale (range from 1 to7), with higher scores representing greater intentions. On the other hand, the International Physical Activity Questionnaire (IPAQ) was used to assess participants’ activity at baseline, 3-month, 6-month, and 18-month.
Data analysis
Participants’ socio-demographics are presented as count and proportion, and the χ2 test was used to assess categorical variables. The main analysis was performed using generalized estimating equations (GEE) model. The GEE solved the problem of correlation of longitudinal data, utilizing the results of each measurement in the longitudinal data, which greatly reduces the loss of information [26]. To clarify whether socioeconomic status affects the effectiveness of ILSM, we conducted several subgroup analyses for physiological health outcomes and health behaviors and added the interaction terms to generalized estimating equations model to assess the interactions between time (at baseline, 3-month, 6-month, and 18-month) and group (intervention and control). We also used GEE to analyze the health benefits of intervention and control groups among participants with different socioeconomic status. Significance level of the associations in the final model was set at P-value 0.05. The data were analyzed using SPSS 20.0 (IBM Corporation, Armonk, NY, USA).
Results
A total of 1,789 women from 16 towns in 2 counties were screened for the recruitment. Among these individuals, 592 (33.1%) were eligible for further assessment. Of these, a total of 320 individuals completed the survey. The CONSORT diagram in Fig. 1 illustrates the reasons for exclusion and loss of follow-up. The longitudinal effect of the ILSM intervention was reported in our previous study [19], which showed the ILSM intervention demonstrated a significant improvement in fasting blood glucose, 2 h-OGTT, BMI, waist circumference, T2D risk score, intention to eat low glycemic index food, and physical activity. In the current analysis, we reported the subgroup analyses of the effect of ILSM intervention on these physiological outcomes between women with different socioeconomic indicators.
Description and comparison of socioeconomic indicators
Among all individuals, 77.5% were younger than 35 years old, about half of them were Han nationality (55.3%), and 51.2% of them had more than 15 years of education (achieved associated degree or above). Approximately 20% of them lived with a monthly family income lower than 3000 RMB (equal to 425 dollars, considered a low family income in China) and 61.6% of them had full-time jobs. More detailed data of distribution information for all variables in this study are provided in Table 1. Independent samples t-tests were conducted to compare baseline characteristics between the intervention and control groups. Results of comparison showed that there were no significant differences on age, ethnicity, education, income per month, and employment status between intervention and control groups. Therefore, these socioeconomic indicators could be further used to conduct subgroup analyses.
Table 1.
Description and Comparison of socioeconomic indicators
| Total (n = 320) |
Control Group (n = 160) |
Intervention Group (n = 160) |
|
|---|---|---|---|
| Age | |||
| 35 years old or younger (%) | 248 (77.5%) | 123 (76.8%) | 125 (78.1%) |
| Older than 35 years old (%) | 72 (33.5%) | 37 (23.1%) | 35 (21.8%) |
| Ethnicity | |||
| Han | 177 (55.3%) | 91 (57.5%) | 85 (53.1%) |
| Minority | 143 (44.7%) | 68 (42.5%) | 75 (44.7%) |
| Education | |||
| 9 years or lower (%) | 68 (21.3%) | 30 (18.8%) | 38 (23.7%) |
| 12 years (%) | 88 (27.5%) | 42 (26.2%) | 46 (28.8%) |
| 15 years or higher(%) | 164 (51.2%) | 92 (57.5%) | 92 (57.5%) |
| Income (per month) | |||
| <3000 RMB (%) | 63 (19.7%) | 33 (20.6%) | 30 (18.8%) |
| ≥ 3000RMB or above (%) | 257 (80.3%) | 130 (81.3%) | 127 (79.4%) |
| Employment status | |||
| Full-time job (%) | 197 (61.6%) | 106 (66.2%) | 91 (56.9%) |
| Part-time job or no job (%) | 123 (38.4%) | 54 (33.8%) | 69 (43.1%) |
Trends of independent variables
The analysis indicates a clear improvement in outcomes for participants in the treat group compared to those in the control group. Across all measured indicators, the treat group consistently demonstrates more favorable results, suggesting that the intervention associated with the treat group has a positive impact. Specifically, the mean values for key outcomes in the treat group exceed those of the control group, highlighting the potential effectiveness of the implemented strategy. These findings provide evidence that the treat group experiences enhanced benefits, supporting the hypothesis that the intervention contributes to superior performance relative to the control group.
Fig. 2.
Trends of independent variables. Note: The 2-hour oral glucose tolerance test and International Physical Activity were not assessed in Phase 2 due to the unavailability of corresponding data for this phase
Subgroup analysis by ethnicity
Table 2 demonstrates that the effectiveness of the intervention between ethnic majority of Han and minority was generally consistent, although participants who were identified as ethnic minority experienced greater benefits (β = 4.06 [95%CI 1.49, 6.63]; p < 0.001) regarding their intention to eat low-glycemic index food in comparison to ethnic majority (β = 1.68 [95%CI -0.82, 4.18]; p = 0.187). Also, the intervention could significantly reduce the FGB in both ethnic minority (β=-0.15 [95%CI -0.24, -0.06]; p < 0.001) and ethnic majority (β=-0.20 [95%CI -0.24, -0.06]; p < 0.001). There were no significant differences in other variables between groups no matter the participants’ ethnicity. Moreover, Appendix Fig. 1 showed the trends of FBG and IEL by ethnic group.
Table 2.
The 3 to 18-month efficacy of the intensive lifestyle modification (ILSM) intervention by Ethnicity subgroups
| Variable | Time | Group | Group*time | ||||
|---|---|---|---|---|---|---|---|
| β | P-value | β | P-value | β | P-value | ||
| CHINARISK | Han | -0.02 (-0.75 to 0.70) | 0.955 | 0.36 (-1.72 to 2.43) | 0.736 | -0.37 (-0.88 to 0.13) | 0.149 |
| Minority | -0.10 (-0.92 to 0.71) | 0.802 | 0.01 (-2.61 to 2.62) | 0.997 | -0.36 (-0.91 to 0.18) | 0.194 | |
| BMI | Han | 0.01 (-0.33 to 0.35) | 0.956 | 0.36 (-0.78 to 1.50) | 0.539 | -0.17 (-0.39 to 0.06) | 0.144 |
| Minority | 0.023 (-0.33 to 0.38) | 0.897 | 0.25 (-1.05 to 1.55) | 0.707 | -0.10 (-0.36 to 0.16) | 0.442 | |
| Waist Circumference | Han | -0.11 (-0.47 to 0.25) | 0.558 | -0.40 (-1.14 to 0.33) | 0.281 | 0.19 (-0.05 to 0.44) | 0.122 |
| Minority | -0.01 (-0.48 to 0.47) | 0.979 | -0.15 (-1.04 to 0.73) | 0.736 | 0.24 (-0.06 to 0.54) | 0.120 | |
| FBG | Han | 0.11 (-0.04 to 0.26) | 0.144 | 0.48 (0.26 to 0.69) | < 0.001 | -0.15 (-0.24 to -0.06) | < 0.001 |
| Minority | 0.23 (0.05 to 0.41) | 0.014 | 0.41 (0.13 to 0.69) | 0.004 | -0.20 (-0.30 to -0.09) | < 0.001 | |
| OGTT-2h | Han | 0.00 (-0.26 to 0.27) | 0.985 | 0.55 (0.03 to 1.07) | 0.037 | -0.02 (-0.20 to 0.15) | 0.791 |
| Minority | 0.29 (-0.18 to 0.77) | 0.228 | 0.51 (-0.19 to 1.20) | 0.155 | -0.20 (-0.46 to 0.07) | 0.147 | |
| IEL | Han | -2.36 (-6.56 to 1.84) | 0.271 | 1.77 (-5.63 to 9.17) | 0.639 | 1.68 (-0.82 to 4.18) | 0.187 |
| Minority | -4.91 (-9.05 to-0.76) | 0.020 | -2.44 (-10.29 to 5.41) | 0.542 | 4.06 (1.49 to 6.63) | 0.002 | |
| IPAQ | Han | -0.09 (-0.63 to 0.46) | 0.758 | -0.12 (-1.10 to 0.86) | 0.809 | -0.05 (-0.41 to 0.32) | 0.797 |
| Minority | 0.39 (-0.20 to 0.98) | 0.192 | 0.47 (-0.60 to 1.53) | 0.390 | -0.35 (-0.72 to 0.02) | 0.066 | |
CHINARISK Chinese Diabetes Risk Scale, BMI Body Mass Index, FBG Fasting Blood Glucose, OGTT-2h 2-hour oral glucose tolerance test, IEL Intention to eat low-GI, IPAQ International Physical Activity Questionnaire
Subgroup analysis by age
Table 3 displays the results of the subgroup analyses by age in the outcomes in 18 months. No matter their age, participants in the intervention group experienced a significant decline in FBG, comparing with those in the control group from baseline to 18-month follow-up (age < 35: β=-0.15 [95%CI -0.22, -0.07]; p < 0.001; age > 35: β=-0.26 [95%CI -0.41, -0.11]; p < 0.001). However, for participants younger than 35 (β=-0.42 [95%CI -0.83, -0.01]; p < 0.045), there was a significant decline in T2D risk scores in the intervention group comparing to the control group over time. While the changes in T2D risk scores were not significant between groups among participants older than 35 (β=-0.42 [95%CI -0.83, -0.01]; p < 0.045). The intervention was more effective in improving the intention to eat low-glycemic index food for participants younger than 35 (β = 3.35 [95%CI 1.27, 5.43]; p = 0.002), while among participants older than 35, there were no significant differences in the intention to eat low-glycemic index food between intervention and ILSM groups (β = 0.33 [95%CI -3.27, 3.93]; p = 0.859). Moreover, Appendix Fig. 2 showed the trends of FBG, IEL, and CHINARISK by age group.
Table 3.
The 3 to 18-month efficacy of the intensive lifestyle modification (ILSM) intervention by age subgroups
| Variable | Time | Group | Group*time | ||||
|---|---|---|---|---|---|---|---|
| β | P-value | β | P-value | β | P-value | ||
| CHINARISK | <35 | -0.08 (-0.68 to 0.52) | 0.797 | 0.50 (-1.37 to 2.36) | 0.604 | -0.42 (-0.83 to -0.01) | 0.045 |
| ≥35 | 0.01 (-1.19 to 1.20) | 0.992 | -0.61 (-3.68 to 2.47) | 0.698 | -0.17 (-1.00 to 0.66) | 0.689 | |
| BMI | <35 | 0.03 (-0.26 to 0.33) | 0.815 | 0.44 (-0.57 to 1.46) | 0.389 | -0.16 (-0.36 to 0.04) | 0.109 |
| ≥35 | -0.10 (-0.53 to 0.33) | 0.654 | 0.06 (-1.55 to 1.66) | 0.946 | -0.04 (-0.37 to 0.29) | 0.827 | |
| Waist Circumference | <35 | 0.00 (-0.33 to 0.33) | 0.979 | 0.22 (-0.41 to 0.85) | 0.491 | -0.17 (-0.39 to 0.05) | 0.126 |
| ≥35 | 0.20 (-0.42 to 0.81) | 0.531 | 0.65 (-0.56 to 1.86) | 0.293 | -0.30 (-0.71 to 0.11) | 0.146 | |
| FBG | <35 | 0.11 (-0.02 to 0.24) | 0.093 | 0.41 (0.21 to 0.60) | < 0.001 | -0.15 (-0.22 to -0.07) | < 0.001 |
| ≥35 | 0.32 (0.09 to 0.54) | 0.006 | 0.60 (0.22 to 0.98) | 0.002 | -0.26 (-0.41 to -0.11) | < 0.001 | |
| OGTT-2h | <35 | 0.15 (-0.15 to 0.45) | 0.333 | 0.52 (0.04 to 1.00) | 0.034 | -0.11 (-0.29 to 0.07) | 0.214 |
| ≥35 | 0.01 (-0.39 to 0.40) | 0.973 | 0.54 (-0.31 to 1.39) | 0.212 | -0.02 (-0.27 to 0.24) | 0.898 | |
| IEL | <35 | -4.70 (-8.10 to -1.29) | 0.007 | -0.84 (-7.18 to 5.49) | 0.794 | 3.35 (1.27 to 5.43) | 0.002 |
| ≥35 | 0.87 (-5.14 to 6.87) | 0.777 | 3.75 (-6.75 to 14.24) | 0.484 | 0.33 (-3.27 to 3.93) | 0.859 | |
| IPAQ | <35 | 0.13 (-0.35 to 0.62) | 0.594 | 0.16 (-0.68 to 1.01) | 0.705 | -0.17 (-0.48 to 0.14) | 0.286 |
| ≥35 | 0.10 (-0.60 to 0.79) | 0.784 | 0.12 (-1.25 to 1.40) | 0.869 | -0.21 (-0.69 to 0.26) | 0.379 | |
Subgroup analysis by income
When conducting subgroup analysis based on income levels, interventions were more effective for individuals with a monthly income greater than $425 in the risk of diabetes (β=-0.52 [95%CI -0.94, -0.10]; p = 0.015), BMI (β=-0.21 [95%CI -0.40, -0.01]; p = 0.036) waist circumference (β = 0.26 [95%CI 0.04, 0.48]; p = 0.023), and intention to consume a low glycemic index diet (β = 2.59 [95%CI 0.41, 4.76]; p < 0.020). Regardless of individual income levels, the FBG levels of the intervention group were significantly reduced comparing to the control group (Income <$425: β=-0.19 [95%CI -0.32, -0.07]; p = 0.003; Income≥$425: β=-0.17 [95%CI -0.25, -0.09]; p < 0.001). The results are revealed in Table 4. Moreover, Appendix Fig. 3 showed the trends of CHINARISK, BMI, Waist Circumference, FBG, IEL by income group.
Table 4.
The 3 to 18-month efficacy of the intensive lifestyle modification (ILSM) intervention by income subgroups
| Variable | Time | Group | Group*time | ||||
|---|---|---|---|---|---|---|---|
| β | P-value | β | P-value | β | P-value | ||
| CHINARISK | Income<3000 | -0.87 (-2.08 to 0.33) | 0.156 | 0.25 (-3.10 to 3.60) | 0.884 | 0.026 (-0.76 to 0.81) | 0.949 |
| Income >3000 | 0.25 (-0.34 to 0.84) | 0.403 | 0.24 (-1.62 to 2.10) | 0.801 | -0.52 (-0.94 to -0.10) | 0.015 | |
| BMI | Income<3000 | -0.30 (-0.80 to 0.21) | 0.247 | 0.09 (-1.59 to 1.76) | 0.919 | 0.06 (-0.28 to 0.40) | 0.732 |
| Income >3000 | 0.12 (-0.15 to 0.40) | 0.385 | 0.40 (-0.61 to 1.41) | 0.434 | -0.21 (-0.40 to -0.01) | 0.036 | |
| Waist Circumference | Income<3000 | 0.33 (-0.31 to 0.97) | 0.315 | -0.40 (-1.56 to 0.75) | 0.493 | 0.02 (-0.38 to 0.43) | 0.903 |
| Income >3000 | -0.17 (-0.50 to 0.16) | 0.307 | -0.26 (-0.91 to 0.39) | 0.432 | 0.26 (0.04 to 0.48) | 0.023 | |
| FBG | Income<3000 | 0.19 (-0.01 to 0.39) | 0.068 | 0.45 (0.14 to 0.77) | 0.005 | -0.19 (-0.32 to -0.07) | 0.003 |
| Income >3000 | 0.15 (0.01 to0.29) | 0.032 | 0.45 (0.24 to 0.65) | < 0.001 | -0.17 (-0;25 to -0.09) | < 0.001 | |
| OGTT-2h | Income<3000 | 0.22 (-0.46 to 0.89) | 0.531 | 0.50 (-0.35 to 1.36) | 0.249 | -0.10(-0.48 to 0.29) | 0.616 |
| Income >3000 | 0.08 (-0.16 to 0.32) | 0.508 | 0.54 (0.06 to 1.02) | 0.028 | -0.09 (-0.24 to 0.06) | 0.247 | |
| IEL | Income<3000 | -3.08 (-0.16 to 2.00) | 0.235 | -0.31 (-10.37 to 9.75) | 0.952 | 2.80 (-0.44 to 6.04) | 0.090 |
| Income >3000 | -3.51 (-7.14 to 0.13) | 0.059 | 0.19 (-6.30 to 6.67) | 0.955 | 2.59 (0.41 to 4.76) | 0.020 | |
| IPAQ | Income<3000 | 0.22 (-0.39 to 0.83) | 0.482 | 0.07 (-1.19 to 1.32) | 0.918 | -0.31 (-0.74 to 0.12) | 0.153 |
| Income >3000 | 0.11 (-0.40to 0.61) | 0.685 | 0.17 (-0.70 to 1.05) | 0.701 | -0.14 (-0.47 to 0.18) | 0.390 | |
Subgroup analysis by employment status
Table 5 reveals effectiveness of interventions on reducing waist circumference (β = 0.28 [95%CI 0.03, 0.52]; p = 0.027) and diabetes risk (β=-0.54 [95%CI -1.02, 0.63]; p = 0.027). This intervention improved intention to consume low GI foods (β = 2.48 [95%CI 0.13, 4.83]; p = 0.039) among intervention participants who were employed full-time. Regardless of whether individuals were employed full-time (β=-0.19 [95%CI -0.27, -0.11]; p < 0.001) or not (β=-0.15 [95%CI -0.27, -0.22]; p = 0.020), FBG levels of the intervention group were effectively reduced comparing to that of the control group. Moreover, Appendix Fig. 4 showed the trends of FBG, IEL, Waist Circumference, and CHINARISK by occupation group.
Table 5.
The 3 to 18-month efficacy of the intensive lifestyle modification (ILSM) intervention by occupation subgroups
| Variable | Time | Group | Group*time | ||||
|---|---|---|---|---|---|---|---|
| β | P-value | β | P-value | β | P-value | ||
| CHINARISK | Full-time job | 0.10 (-0.60to 0.80) | 0.780 | 0.80 (-1.30 to 2.90) | 0.457 | -0.54 (-1.02 to -0.63) | 0.027 |
| No full-time job | -0.28 (-1.12 to 0.56) | 0.518 | -0.48 (-3.00 to2.04) | 0.711 | -0.12 (-0.70 to 0.46) | 0.683 | |
| BMI | Full-time job | 0.01 ( -0.31 to 0.32) | 0.976 | 0.50 (-0.63 to 1,64) | 0.386 | -0.20 (-0.42 to 0.03) | 0.085 |
| No full-time job | 0.04 (-0.32 to0.40) | 0.818 | 0.13 (-1.18 to 1.43) | 0.846 | -0.05 (-0.29 to 0.18) | 0.665 | |
| Waist Circumference | Full-time job | -0.13 (-0.48 to 0.23) | 0.490 | -0.51 (-1.23 to 0.20) | 0.159 | 0.28 (0.03 to 0.52) | 0.027 |
| No full-time job | 0.05 (-0.46 to 0.56) | 0.846 | 0.00 (-0.93 to 0.92) | 0.993 | 0.10 (-0.22 to 0.42) | 0.555 | |
| FBG | Full-time job | 0.17 90.04 to 0.30) | 0.011 | 0.45 (0.23 to 0.67) | < 0.001 | -0.19 (-0.27 to -0.11) | < 0.001 |
| No full-time job | 0.14 (-0.08 to 0.36) | 0.211 | 0.45 (-0.16 to 0.74) | 0.002 | -0.15 (-0.27 to -0.22) | 0.020 | |
| OGTT-2 h | Full-time job | 0.27 (-0.05 to 0.59) | 0.104 | 0.69 (0.16 to 1.22) | 0.011 | -0.18 (-0.38 to 0.02) | 0.076 |
| No full-time job | -0.17 (-0.54 to 0.20) | 0.375 | 0.20 (-0.46 to 0.86) | 0.545 | 0.07 (-0.15 to 0.29) | 0.547 | |
| IEL | Full-time job | -3.38 (-7.15 to 0.38) | 0.078 | 2.27 (-4.57 to 9.11) | 0.516 | 2.48 (0.13 to 4.83) | 0.039 |
| No full-time job | -3.14 (-8.04 to 1.77) | 0.210 | -3.55 (-12.58 to 5.49) | 0.442 | 2,80 (-6.66 to 22.54) | 0.287 | |
| IPAQ | Full-time job | 0.00 (-0.51 to 0.50) | 0.990 | 013 (-0.75 to 1.02) | 0.768 | -0.16 (-0.48 to 0.16) | 0.340 |
| No full-time job | 0.42 (-0.25 to 1.09) | 0.222 | 0.14 (-1.11 to 1.39) | 0.822 | -0.26 (-0.72 to 0.20) | 0.267 | |
Subgroup analysis by education
When conducting subgroup analysis based on educational attainment, we found that compared to the control group, the FBG levels of intervention group were effectively reduced among participants receiving 9 years of education or below (β=-0.19 [95%CI -0.34, -0.05]; p = 0.008), as well as among participants with 15 years of education or above (β=-0.19 [95%CI -0.33, -0.05]; p = 0.008). Individuals with 15 years of education in the intervention group benefited the most on improving the intention to eat low GI foods (β = 5.64 [95%CI 2.67, 8.62]; p < 0.001) (Table 6). Moreover, Appendix Fig. 5 showed the trends of FBG, and IEL by education group.
Table 6.
The 3 to 18-month efficacy of the intensive lifestyle modification (ILSM) intervention by education subgroups
| Variable | Time | Group | Group*time | ||||
|---|---|---|---|---|---|---|---|
| β | P-value | β | P-value | β | P-value | ||
| CHINARISK | 9 years or lower (%) | 1.94(-1.79 to 5.68) | 0.205 | 0.76(-0.41 to 1.92) | 0.308 | -0.80(-1.67 to 0.07) | 0.071 |
| 12 years (%) | 0.42(-2.28 to 3.12) | 0.759 | -0.23(-1.17 to 0.71) | 0.630 | -0.353(-0.955 to 0.248) | 0.250 | |
| 15 years or higher (%) | -0.67(-3.05 to 1.70) | 0.579 | -0.27(-1.06 to 0.53) | 0.509 | -0.21(-0.76 to 0.34) | 0.451 | |
| BMI | 9 years or lower (%) | 0.03 (-0.41 to 0.48) | 0.884 | 1.01 (-0.94 to 2.95) | 0.309 | -0.21 (-0.54 to 0.12) | 0.216 |
| 12 years (%) | -0.06 (-0.50 to 0.38) | 0.779 | -0.13 (-1.66 to 1.40) | 0.868 | -0.13(-0.44 to 0.19) | 0.430 | |
| 15 years or higher (%) | -0.08 (-0.31 to 0.49) | 0.678 | 0.32 (-0.94 to 1.57) | 0.620 | -0.124 (-0.374 to 0.126) | 0.330 | |
| Waist Circumference | 9 years or lower (%) | 0.58(-1.51 to 2.66) | 0.587 | 2.58(-2.56 to 7.71) | 0.325 | -0.638(-2.10 to 0.82) | 0.392 |
| 12 years (%) | -0.22(-1.74 to 1.30) | 0.777 | -0.41(-4.18 to 3.37) | 0.833 | -0.44(-1.42 to 0.54) | 0.378 | |
| 15 years or higher (%) | -0.25(-1.87 to 1.36) | 0.759 | 0.46(-3.19 to 4.11) | 0.803 | -0.39(-1.40 to 0.62) | 0.447 | |
| FBG | 9 years or lower (%) | 0.14(-0.11 to 0.38) | 0.265 | 0.54(0.18 to 0.89) | 0.003 | -0.192(-0.33 to -0.05) | 0.007 |
| 12 years (%) | 0.03(-0.17 to 0.22) | 0.807 | 0.17(-0.11 to 0.44) | 0.231 | -0.08(-0.20 to 0.04) | 0.182 | |
| 15 years or higher (%) | 0.27(0.10 to 0.43) | 0.001 | 0.60(0.34 to 0.87) | < 0.001 | -0.23(-0.331 to -0.128) | < 0.001 | |
| OGTT-2 h | 9 years or lower (%) | 0.04(-0.33 to 0.41) | 0.843 | 0.69(-0.13 to 1.51) | 0.099 | -0.06(-0.32 to 0.21) | 0.665 |
| 12 years (%) | 0.04(-0.47 to 0.55) | 0.872 | 0.11(-0.622 to 0.839) | 0.771 | -0.03(-0.32 to 0.26) | 0.817 | |
| 15 years or higher (%) | 0.21 (-0.16 to 0.58) | 0.271 | 0.71(0.08 to 1.33) | 0.027 | -0.15(-0.37 to 0.07) | 0.198 | |
| IEL | 9 years or lower (%) | -0.51(-6.80 to 5.78) | 0.874 | 7.85(-4.15 to 19.84) | 0.20 | 1.88(-1.90 to 5.66) | 0.330 |
| 12 years (%) | -2.58(-8.09 to 2.93) | 0.358 | 3.275(-6.04 to 12.59) | 0.491 | 1.07(-2.17 to 4.30) | 0.517 | |
| 15 years or higher (%) | -5.74(-9.89 to -1.59) | 0.007 | -6.1(-14.07 to 1.87) | 0.134 | 4.33(1.73 to 6.92) | 0.001 | |
| IPAQ | 9 years or lower (%) | 0.90(-0.84 to 2.64) | 0.309 | 0.68(-0.29 to 1.65) | 0.169 | -0.62(-1.31 to 0.07) | 0.077 |
| 12 years (%) | -0.21(-0.78 to 0.37) | 0.475 | -0.45(-1.47 to 0.58) | 0.394 | 0.05(-0.40 to 0.40) | 0.797 | |
| 15 years or higher (%) | 0.16(-0.48 to 0.79) | 0.626 | 0.233(-0.92 to 1.38) | 0.691 | -0.18(-0.59 to 0.23) | 0.394 | |
Within-group health benefits based on socioeconomic indicators
We also analyzed the health benefits of intervention and control groups among populations with different socioeconomic characteristics (Appendix Tables 1 to 4). The results show that only in the control group, unemployed women have significantly better BMI control effects compared to employed women. In other cases, there were no significant differences in health benefits obtained from the intervention across women with different socioeconomic characteristics.
Discussion
Our study revealed that the postpartum period after GDM diagnosis could be an ideal time for the prevention of T2DM and further, diverse intervention formats and intensive lifestyle modification could be considered to benefit women with socioeconomic inequalities. The ILSM program was demonstrated a consistent significance on reducing fasting blood glucose (FBG) across ethnicity, age, income, employment status, and education, although greater effects on increasing intentions to eat low-glycemic foods, decreasing BMI and waist circumference, or reducing diabetes risk were seen in women with full time employment, high income (≥$425 per month), younger age (≤ 35 years), high education level (15 years of education or above), and interestingly, ethnic minority.
Glucose markers, including FBG, are identified as key indicators of delaying progression to T2DM among the high-risk population [27]. This significant effect on reducing FBG is consistent with prior studies showing that evidence-based lifestyle interventions could improve glucose intolerance in the population at high risk for developing T2DM [28, 29]. While previous interventions on the effect of lifestyle interventions according to socioeconomic positions did not include glucose markers in their outcome evaluation [11–13], our study could provide valuable evidence concluding that the implementation of the current postpartum lifestyle intervention could contribute to health equity of diabetes prevention on glucose management in Chinese women with history of GDM. This postpartum lifestyle intervention aimed to encourage behavioral change at the individual level using strategies of in-person group classes and telephone consultation provided by project officers [17]. Compared to previous interventions addressing health disparities of diabetes prevention among high-risk individuals [11–13], this intervention focused on high-risk women with history of GDM who might have experienced formal diabetes management during pregnancy and provided more intensive lifestyle modification through both in-person and telephone formats.
However, the differentiating benefits of the ILSM intervention on weight and behavior outcomes should be noted. First, women with full-time employment and higher income (≥$425 per month) were reported to have greater benefits on intentions to eat low-glycemic foods, BMI and waist circumference, and T2DM risk. Stable income from full-time employment is a strong indicator of high socioeconomic status. Employment-related factors may partly explain the less favorable BMI control among working women. Occupational sitting and reduced physical activity during work hours can contribute to greater weight gain, while work-related stress may trigger emotional eating and disturb sleep, both of which are associated with adverse metabolic changes [30, 31]. Women with lower socioeconomic status face multiple barriers that limit the effectiveness of lifestyle interventions. However, there is a lack of evidence which indicated the waist circumstance and employment status.
Individuals from higher socioeconomic positions are likely to live in the neighborhood with exercise facilities and have easier access to higher quality of food, resulting in physical activity, less intake of fat and simple carbohydrates, and more intake of fruits, vegetables and whole wheat bread [32]. Time poverty restricts opportunities for exercise and meal planning, as low‑income mothers often juggle long working hours, multiple jobs, and caregiving responsibilities, spending more time in sedentary activities [33]. Limited access to healthy foods in disadvantaged neighborhoods further constrains dietary behavior change, even when motivation is present [34]. Additionally, cognitive and psychosocial factors, including lower health literacy, weaker self-efficacy, and less favorable outcome expectations, hinder the adoption of healthy behaviors. These mechanisms extend beyond mere material resource limitations and collectively reduce the benefits of interventions in low‑SES populations [35]. Therefore, when enrolled either in the ILSM intervention or in control groups, women from higher socioeconomic positions could have more resources to support their health-enhancing intentions and behaviors.
Additionally, women with younger age (≤ 35 yrs) or higher education level (15 years of education or above) were more likely to benefit from the ILSM intervention on improving higher intentions to eat low-glycemic foods and lowering T2DM risk. Prior evidence supported that lower disease-related knowledge or the capacity to obtain, process, and understand disease-related knowledge tends to be seen in people with older age [36] and lower education backgrounds [37], which was further associated with poorer health outcomes [38]. Specifically in Chinese adults, younger age and higher education have been associated with higher levels of diabetes knowledge [39] which subsequently lead to better diabetes self-care intentions and behaviors. Aligning with our hypothesis, it could be speculated that younger people growing in a more modern environment might have more access to social, cultural, and economic resources which could support them to get longer education year and understand more health-related information.
Interestingly, against our theoretical hypothesis on ethnicity, our findings showed that ethnic minority women were more likely to gain benefits from the ILSM intervention on increasing intentions to eat low-glycemic foods, compared to ethnic majority (Han) women. This result suggests that, compared with Han women, ethnic minority women may experience a larger shift in motivation toward adopting low-glycemic or healthier dietary patterns during the postpartum intervention. Prior studies have shown that certain ethnic minority groups often start with less healthy baseline dietary habits and lower exposure to formal health education, leaving more room for improvement once targeted guidance is provided [40]. Although a Chinese study showed that ethnic minority women tended to have lower rate of formal education and higher rate of low health literacy [41], there have been limited studies exploring the mechanisms of health behavior change in ethnic minority Chinese living in the rural areas. Considering culturally-specific unhealthy habits (e.g., more alcohol drinking, consuming more carbohydrates, and/or physical inactivity) [42] before preconception, pregnancy and the postpartum period might indicate a special period for receiving education and having regular visits to health care for these ethnic minority women, particularly for women diagnosed with GDM. Based on this, we speculated that ethnic minority women have more space for behavioral change from preconception to postpartum, compared to ethnic minority women who are more likely to have healthier behaviors before pregnancy. In addition, the Social Cognitive Theory provides a useful framework for understanding why women with lower socioeconomic status may experience fewer benefits from the intervention. Specifically, these women may face greater barriers related to outcome expectations, such as limited awareness of the long-term health benefits of dietary control and physical activity, which can weaken motivation for sustained behavior change. They may also have lower self-efficacy, including reduced confidence in their ability to adopt and maintain healthier routines amid financial constraints, limited social support, or competing family responsibilities. These psychosocial barriers could partially explain the diminished intervention effects observed in lower SES groups and highlight the need for future programs to incorporate strategies that explicitly strengthen both outcome expectations and self-efficacy. Therefore, this might provoke a great opportunity for individual behavioral change during the postpartum lifestyle modification.
Research and political implications
Our study underscored the complex interplay between socioeconomic factors and health outcomes, emphasizing the need for tailored and nuanced approaches in public health interventions aimed at diverse demographic groups. There are important implications for future research and policy making on improving health equity of diabetes prevention. Interventions development needs to pay more attention to lower SES women with tailored strategies informed by socioeconomic resources, for example, recommending physical activities that require less facilities and personalized dietary plan that utilizes less expensive foods. Future studies could first understand the demands of these people based on their real-life experiences, and then tailor interventions that could better fit into their daily life routines. Concrete examples of such interventions may include developing flexible, app-based lifestyle coaching programs that accommodate irregular schedules; offering community-based physical activity sessions, such as short, guided walks or low-impact exercise classes near workplaces or residential areas; and providing individualized dietary counseling that incorporates affordable, culturally appropriate meal plans for families with limited time or resources.
To gain the same level of health benefits in diabetes prevention, educational materials and consultation support might need to be tailored for women with lower disease-related knowledge related to diabetes, indicated by older age and lower education backgrounds. Prior evidence from a review recommended interventions shown to be effective in lower SES participants primarily included community-based strategies or policies aimed at structural changes to the environment [43]. While existing food and nutrition resources are insufficient to support healthy eating which is crucial for diabetes prevention and management, people with less socioeconomic resources endorse lifestyle interventions focusing on food resource management, nutritional counseling, and health coaching [44]. Therefore, including community-level strategies and provoking environmental changes through policy making for diabetes prevention may benefit women with socioeconomic inequalities.
Limitations
There are several limitations to be noticed. First, the sample size was calculated based on the purpose of detecting the effect of ILSM program on the primary outcome, therefore, the power might not be adequate to conduct subgroup analyses in terms of socioeconomic characteristics. Post hoc power analysis was conducted using the difference between two dependent means (matched pairs) to determine the likelihood of a type 2 error. Based on current information (beta = 0.28), the sample size was not adequately powered at 28% to detect no difference between the intervention and control groups. Under this circumstance, we still detected significant effects of the ILSM intervention on improving some diabetes-related outcomes in different subgroups. We suggested future research using adequate power to potential detect more positive results on addressing health disparities on diabetes prevention. Second, the self-report measurements for physical activity may have recall bias, which might be the reason of non-significant results in the International Physical Activity Questionnaire.
Conclusion
This study provided valuable evidence supporting the effect of lifestyle intervention on glucose intolerance according to socioeconomic characteristics. Women enrolled in the ILSM program tailored for high-risk women living in the rural areas of China are likely to obtain similar levels of health benefits on glucose intolerance even though socioeconomic inequalities. However, there was a differentiating effect on increasing intentions to eat low-glycemic, decreasing BMI and waist circumference, and further reducing T2DM risk based on women’s different socioeconomic backgrounds. Therefore, interventions need to address lower socioeconomic status women with tailored strategies informed by insufficient socioeconomic resources.
Supplementary Information
Acknowledgements
The authors would like to thank all participants for their time and cooperation, as well as the volunteers who helped with data collection.
Authors’ contributions
J.W., Y.B., J. Z., and J.G. conceptualized the study design, developed the analytical framework, and led the manuscript writing. Y.B. and Y.L. was responsible for data collection, data cleaning, and statistical analysis, and contributed to the interpretation of empirical results. J.Z. and J.G provided methodological guidance, reviewed the analytical strategies, and critically revised the manuscript for intellectual content. M.L., R.L., and Y.W. reviewed the manuscript for technical and linguistic accuracy. J.W., Y.B., J. Z., and J.G. refined the clarity of expression, corrected inconsistencies in terminology, and ensured compliance with journal formatting requirements, including citation style, tables, and figure layout.All authors reviewed and approved the final manuscript and agree to be accountable for the accuracy and integrity of all aspects of the work.
Funding
National Social Science Fund Major Projects (No.22&ZD144); 2022 top-notch innovative talent cultivation funding plan of Renmin University of China; Joint Funds of the National Natural Science Foundation of China (No.72061160491); and the China Medical Board (grant number:16–256 and 21–424).
Data availability
The datasets used during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the ethical committee of Xiangya Nursing School of Central South University (No. 2016034). And this trial registered at the Chinese Clinical Trial Registry (No. ChiCTR1800015023) on 1st March 2018, http://wwwchictr.org.cn/showproj.aspx?proj-25569. At the start of the study, local nurses explained the program to interested women, confirmed their eligibility, and obtained consent.
Consent for publication
N/A.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Jun Wang and Yang Bai contributed equally to this work and are co-first authors.
Contributor Information
Jie Zhong, Email: jiezhong@hku.hk.
Jia Guo, Email: guojiacsu@csu.edu.cn.
References
- 1.Facts & figures. International Diabetes Federation. Available from: https://idf.org/about-diabetes/diabetes-facts-figures/. Cited 10 Dec 2025.
- 2.CDC. Centers for Disease Control and Prevention. 2019. Gestational Diabetes. Available from: https://www.cdc.gov/diabetes/basics/gestational.html. Cited 8 Feb 2022.
- 3.Bellamy L, Casas JP, Hingorani AD, Williams D. Type 2 diabetes mellitus after gestational diabetes: a systematic review and meta-analysis. Lancet. 2009;373(9677):1773–9. [DOI] [PubMed] [Google Scholar]
- 4.Hoskin MA, Bray GA, Hattaway K, Khare-Ranade PA, Pomeroy J, Semler LN, et al. Prevention of Diabetes Through Lifestyle Intervention: Lessons Learned from the Diabetes Prevention Program and Outcomes Study and its Translation to Practice. Curr Nutr Rep. 2014;3(4):364–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Aroda VR, Christophi CA, Edelstein SL, Zhang P, Herman WH, Barrett-Connor E, et al. The Effect of Lifestyle Intervention and Metformin on Preventing or Delaying Diabetes Among Women With and Without Gestational Diabetes: The Diabetes Prevention Program Outcomes Study 10-Year Follow-Up. J Clin Endocrinol Metab. 2015;100(4):1646–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bull ER, Dombrowski SU, McCleary N, Johnston M. Are interventions for low-income groups effective in changing healthy eating, physical activity and smoking behaviours? A systematic review and meta-analysis. BMJ Open. 2014;4(11):e006046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Espelt A, Kunst AE, Palència L, Gnavi R, Borrell C. Twenty years of socio-economic inequalities in type 2 diabetes mellitus prevalence in Spain, 1987–2006. Eur J Public Health. 2012;22(6):765–71. [DOI] [PubMed] [Google Scholar]
- 8.Reisig V, Reitmeir P, Döring A, Rathmann W, Mielck A, KORA Study Group. Social inequalities and outcomes in type 2 diabetes in the German region of Augsburg. A cross-sectional survey. Int J Public Health. 2007;52(3):158–65. [DOI] [PubMed] [Google Scholar]
- 9.Welch V, Dewidar O, Ghogomu ET, Abdisalam S, Ameer AA, Barbeau VI et al. How effects on health equity are assessed in systematic reviews of interventions. Cochrane Database Syst Rev. 2022;(1). Available from: https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.MR000028.pub3/full. Cited 5 Mar 2024. [DOI] [PMC free article] [PubMed]
- 10.Spencer Bonilla G, Rodriguez-Gutierrez R, Montori VM. What We Don’t Talk About When We Talk About Preventing Type 2 Diabetes—Addressing Socioeconomic Disadvantage. JAMA Intern Med. 2016;176(8):1053–4. [DOI] [PubMed] [Google Scholar]
- 11.Rautio N, Jokelainen J, Oksa H, Saaristo T, Peltonen M, Niskanen L, et al. Socioeconomic position and effectiveness of lifestyle intervention in prevention of type 2 diabetes: One-year follow-up of the FIN-D2D project. Scand J Public Health. 2011;39(6):561–70. [DOI] [PubMed] [Google Scholar]
- 12.Genz J, Haastert B, Müller H, Verheyen F, Cole D, Rathmann W, et al. Socioeconomic factors and effect of evidence-based patient information about primary prevention of type 2 diabetes mellitus - are there interactions? BMC Res Notes. 2014;7(1):541. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Jiang L, Huang H, Johnson A, Dill EJ, Beals J, Manson SM, et al. Socioeconomic Disparities in Weight and Behavioral Outcomes Among American Indian and Alaska Native Participants of a Translational Lifestyle Intervention Project. Diabetes Care. 2015;38(11):2090–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Liu J, Liu M, Chai Z, Li C, Wang Y, Shen M et al. Projected rapid growth in diabetes disease burden and economic burden in China: a spatio-temporal study from 2020 to 2030. Lancet Reg Health – West Pac. 2023;33. Available from: https://www.thelancet.com/journals/lanwpc/article/PIIS2666-6065(23)00018-4/fulltext. Cited 26 Feb 2024. [DOI] [PMC free article] [PubMed]
- 15.Dong Y, Gao W, Nan H, Yu H, Li F, Duan W, et al. Prevalence of Type 2 diabetes in urban and rural Chinese populations in Qingdao, China. Diabet Med. 2005;22(10):1427–33. [DOI] [PubMed] [Google Scholar]
- 16.Song S, Yuan B, Zhang L, Cheng G, Zhu W, Hou Z, et al. Increased Inequalities in Health Resource and Access to Health Care in Rural China. Int J Environ Res Public Health. 2018;16(1):49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Guo J, Tang Y, Wiley J, Whittemore R, Chen JL. Effectiveness of a diabetes prevention program for rural women with prior gestational diabetes mellitus: study protocol of a multi-site randomized clinical trial. BMC Public Health. 2018;18(1):809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chen Y, Zhong Q, Luo J, Tang Y, Li M, Lin Q, et al. The 6-Month Efficacy of an Intensive Lifestyle Modification Program on Type 2 Diabetes Risk Among Rural Women with Prior Gestational Diabetes Mellitus: a Cluster Randomized Controlled Trial. Prev Sci. 2022;23(7):1156–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zhong Q, Chen Y, Luo M, Lin Q, Tan J, Xiao S, et al. The 18-month efficacy of an Intensive LifeStyle Modification Program (ILSM) to reduce type 2 diabetes risk among rural women: a cluster randomized controlled trial. Glob Health. 2023;19(1):6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wang D, Dai X, Mishra SR, Lim CCW, Carrillo-Larco RM, Gakidou E, et al. Association between socioeconomic status and health behaviour change before and after non-communicable disease diagnoses: a multicohort study. Lancet Public Health. 2022;7(8):e670–82. [DOI] [PubMed] [Google Scholar]
- 21.Levy M, Chen Y, Clarke R, Bennett D, Tan Y, Guo Y, et al. Socioeconomic differences in health-care use and outcomes for stroke and ischaemic heart disease in China during 2009–16: a prospective cohort study of 0·5 million adults. Lancet Glob Health. 2020;8(4):e591–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chokshi DA. Income, Poverty, and Health Inequality. JAMA. 2018;319(13):1312–3. [DOI] [PubMed] [Google Scholar]
- 23.Silver SR, Li J, Quay B. Employment status, unemployment duration, and health-related metrics among US adults of prime working age: Behavioral Risk Factor Surveillance System, 2018–2019. Am J Ind Med. 2022;65(1):59–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zajacova A, Lawrence EM. The Relationship Between Education and Health: Reducing Disparities Through a Contextual Approach. Annu Rev Public Health. 2018;39(1):273–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Savelli E, Murmura F, Bravi L. Healthy and quality food attitudes and lifestyle: a generational cohort comparison. TQM J. 2023. Available from: 10.1108/TQM-05-2023-0156. Ahead-of-print(ahead-of-print). Cited 14 Mar 2024.
- 26.Ballinger GA. Using Generalized Estimating Equations for Longitudinal Data Analysis. Organ Res Methods. 2004;7(2):127–50. [Google Scholar]
- 27.Roglic G. WHO Global report on diabetes: A summary. International Journal of Noncommunicable Diseases. 2016;1(1):3–8.
- 28.Mensink M, Blaak EE, Corpeleijn E, Saris WH, de Bruin TW, Feskens EJ. Lifestyle Intervention According to General Recommendations Improves Glucose Tolerance. Obes Res. 2003;11(12):1588–96. [DOI] [PubMed] [Google Scholar]
- 29.Epton T, Keyworth C, Goldthorpe J, Calam R, Armitage CJ. Are interventions delivered by healthcare professionals effective for weight management? A systematic review of systematic reviews. Public Health Nutr. 2022;25(4):1071–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health. 1999;89(9):1322–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Fujishiro K, Lawson CC, Hibert EL, Chavarro JE, Rich-Edwards JW. Job strain and changes in the body mass index among working women: a prospective study. Int J Obes. 2015;39(9):1395–400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Volaco A, Cavalcanti AM, Filho RP, Précoma DB. Socioeconomic Status: The Missing Link Between Obesity and Diabetes Mellitus? Curr Diabetes Rev. 2018;14(4):321–6. [DOI] [PubMed] [Google Scholar]
- 33.Gough M, Lippert AM, Martin MA. The role of time use behaviors in the risk of obesity among low-income mothers. Womens Health Issues. 2019;29(1):23–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Evans A, Banks K, Jennings R, Nehme E, Nemec C, Sharma S, et al. Increasing access to healthful foods: a qualitative study with residents of low-income communities. Int J Behav Nutr Phys Act. 2015;12(Suppl 1):S5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Pampel FC, Krueger PM, Denney JT. Socioeconomic disparities in health behaviors. Annu Rev Sociol. 2010;36(1):349–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Chesser AK, Keene Woods N, Smothers K, Rogers N. Health Literacy and Older Adults: A Systematic Review. Gerontol Geriatr Med. 2016;2:2333721416630492. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Mantwill S, Monestel-Umaña S, Schulz PJ. The Relationship between Health Literacy and Health Disparities: A Systematic Review. PLoS ONE. 2015;10(12):e0145455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Berkman ND, Sheridan SL, Donahue KE, Halpern DJ, Crotty K. Low Health Literacy and Health Outcomes: An Updated Systematic Review. Ann Intern Med. 2011;19(2):97–107. [DOI] [PubMed] [Google Scholar]
- 39.Hu J, Gruber KJ, Liu H, Zhao H, Garcia AA. Diabetes knowledge among older adults with diabetes in Beijing, China. J Clin Nurs. 2013;22(1–2):51–60. [DOI] [PubMed] [Google Scholar]
- 40.Mosdøl A, Lidal IB, Straumann GH, Vist GE. Targeted mass media interventions promoting healthy behaviours to reduce risk of non-communicable diseases in adult, ethnic minorities. Cochrane Database Syst Rev. 2017;2(2):CD011683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Wang C, Li H, Li L, Xu D, Kane RL, Meng Q. Health literacy and ethnic disparities in health-related quality of life among rural women: results from a Chinese poor minority area. Health Qual Life Outcomes. 2013;11(1):153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Anderson NB, Bulatao RA, Cohen B, National Research Council (US). Panel on Race E. Racial/Ethnic Disparities in Health Behaviors: A Challenge to Current Assumptions. In: Critical Perspectives on Racial and Ethnic Differences in Health in Late Life. National Academies Press (US); 2004. Available from: https://www.ncbi.nlm.nih.gov/books/NBK25518/. Cited 14 Mar 2024. [PubMed]
- 43.Beauchamp A, Backholer K, Magliano D, Peeters A. The effect of obesity prevention interventions according to socioeconomic position: a systematic review. Obes Rev. 2014;15(7):541–54. [DOI] [PubMed] [Google Scholar]
- 44.Stotz SA, Ricks KA, Eisenstat SA, Wexler DJ, Berkowitz SA. Opportunities for Interventions That Address Socioeconomic Barriers to Type 2 Diabetes Management: Patient Perspectives. Sci Diabetes Self-Manag Care. 2021;47(2):153–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used during the current study are available from the corresponding author on reasonable request.