Diet and lifestyle interventions in postpartum women in China: study design and rationale of a multicenter randomized controlled trial

This multicenter RCT is designed to assess the effectiveness of a diet and lifestyle intervention in postpartum women in China, including a short-term effect on reducing health problems during puerperium period and a long-term effect on improving health status in women and their offspring over a two-year period.

The current multicenter study will be conducted in three representative areas in China, Shandong province, Hubei province and Guangdong province, which locate in northern, central and southern parts of China, respectively. Women who attend routine pregnancy diagnosis in hospitals or maternal healthcare centers will be invited to take part in the randomized controlled intervention trial.

Sample size has been calculated as the following formula [18]:

, in which p ₁ indicates incidence rate in control group, p ₂ indicates incidence rate in intervention group and indicates average of p ₁ and p ₂ . Z_α/2 is 1.96 when significance level (α) is 0.05 and Z_β is 1.64 for power (1 - β) is 0.90.

Take constipation, a typical diet-related digestive disorder, for example. The incidence rate of constipation in our previous study is 20.95% in the control group [16]. We aim to reduce the incidence rate by 50%, which is 10.475%, the sample size N = = 309.89 ≈ 310 for each group. On the basis of the above calculation, a sample size of 800 participants (400 in the intervention group and 400 in the control group) has been determined, considering a certain attrition rate. This sample size is adequate, with statistical power >80%, for nearly all the major outcomes, including changes in nutritional status and health status.

The flowchart of participant's recruitment process and trial design is summarized in figure 1. Pregnancy women that are willing to take part in our study will be checked by our eligibility criteria. The eligibility criteria for enrollment includes: 1) healthy pregnant women; 2) at their third trimester; 3) had at least three routine examinations at these antenatal clinics [16]. A written consent form will be assigned by each woman to confirm her willingness to participate, and then baseline data will be collected by a questionnaire survey (main components shown in Table 1) and a consequent anthropometric measurement.

Recruited women will be randomly allocated to either intervention group or control group according to a random number generated by a computer-based procedure. Then, a unique code for each participating woman and her allocation will be recorded.

Although it is impossible to ensure blinding for intervention conductor and the participating women, we will ensure blinding for the outcome measurements and biochemical analyses by independent examiners unaware of group allocation. Participants cannot be blinded for the intervention, but they are asked not to reveal information about their allocation to the examiners. The key of coding concerning group assignment is only known by the programmer of the database that is used during the study.

On the basis of our previous study in postpartum women from Hubei province of China and the latest WHO recommendation [19], we have formed a set of recommendations for women's diet and lifestyle during the postpartum period, which is the guideline for the current intervention study. A comparison between traditional practices for "doing the month" and expected diet and lifestyle that we advocated during the postpartum period are shown in Table 1. It is obvious that our guideline emphasizes the importance of increasing consumption of vegetables and fruits, which is regard as a diet taboo for postpartum women in traditional Chinese practice. In addition, it also emphasizes the importance of maintaining energy balance through reducing excessive energy from meat, eggs and brown sugar and increasing energy expenditure from appropriate physical activity, such as doing maternal keep-fit exercises.

In the past decades, social ecological models have been increasingly used to formulate intervention strategies in comprehensive consideration, because simple interventions are unlikely to work on their own and the development of effective interventions requires strategies that affect multiple settings simultaneously [20]. As shown in Figure 2, a panel of environmental factors that influence the knowledge, belief and practice of postpartum women (the core) has been identified. Microsystem, mainly referred to family members of the individual woman, is crucial immediate environmental factors that should be considered in priority. Exosystem, including surrounding people, such as peers, friends, colleagues, neighbors and care physicians, and sources of knowledge, such as book, magazines, TV show, radio and internet, does not usually directly interact, but that can still affect the individual woman. In addition, we have to consider macrosystem, including history, culture, laws, ethnics, social conditions and economic systems, to give a personalized guide to individual postpartum woman. Previous studies by us and other groups have suggested that main influence factors on the diet and lifestyle of a postpartum woman include the woman's knowledge about postpartum nutrition and health care, the woman or her husband's educational level and family income levels, and the traditional concept learned from her mother in-law (49.4%), mother (34.0%), books and magazines (16.5%), relatives (14.3), friends or colleagues (12.6%) [15, 21].

Therefore, we have formulated a three-dimension comprehensive intervention strategy (Table 3), which incorporates intervention measures simultaneously to individual postpartum woman, their family members and community environment to maximize the effectiveness of our multicenter trial. For individual postpartum woman, we will set antepartum participatory training courses, provide specially prepared brochures and supporting VCD, set up a specialized counseling hotline and give individual bedside guidance after birth. For their family members, we will invite every postpartum woman's husband and mother to take part in workshops or seminars, making them know some basic knowledge of postpartum care, and persuading them to promote adoption of the diet and lifestyle we advocated by the postpartum woman. For community environment, we will put up posters about basic knowledge of postpartum care in the community bulletin boards for general population educating and mobilize community leaders and heads from and maternal and child health centers for maximum support.

Major measurements before and after the intervention trial are shown in Table 4. The primary outcome for the current study is the changes in the knowledge and belief about postpartum practice among the women, which expect the women in intervention group know more about the nutritional requirements in postpartum period and the potential benefit from a balanced diet with appropriate amount of animal food and abundant amount of plant-derived food. The secondary outcome for the current study is the changes in the nutritional status, including both macronutrients and micronutrients level, which reflects the real action of adopting our recommendation in their everyday food intake. The tertiary and the most important outcome is the changes in postpartum recovery and health status, including a promotion for uterine involution and a reduction of postpartum diseases such as puerperal infection, fever, constipation, hemorrhoids, anal fissure, breast disease, oral ulcers, gum bleeding, anemia, pains (headache, heel pain, back pain, leg pain, joint pain, leg cramps, etc.) and postpartum depression.

Data entry will be independently done by two researchers, and then a computer software-based error detection will be carried out to check the consistency of these data. If inconsistency occurs, the original questionnaire will be referred and the error will be corrected via a re-entry.

This study will be carried out in accordance with requirements documented in the Declaration of Helsinki. Ethics approval has been obtained from the local Health Department and the research ethics boards of Tongji Medical College, China.

Descriptive statistics will be calculated and checked for balance in intervention and control groups in demographic, health and outcome measurements at baseline. Comparisons between intervention group and control group will be performed by Chi-square (categorical variables), t test (continuous variables, normal distribution) or Mann-Whitney U test (continuous variables, skewed distribution). Final analyses will be undertaken using generalized linear mixed models to investigate changes over time in the two groups in the major outcomes. These models allow for repeated measures and can be used for normally distributed, binomial, and ordinal data [22]. All statistical analyses will be performed using SPSS 12.0 statistical software package (SPSS Inc, Chicago, IL, USA).