Abstract
Excessive gestational weight gain is a prevalent problem and an independent predictor of future obesity in both mothers and offspring. Intervention research to prevent excessive gestational weight gain is still in its infancy but results to date have been quite modest. Research in weight control outside of pregnancy over the past 30 years has been more robust and identified several key components of effective programs, including use of caloric restriction, daily diet self-monitoring, self-weighing, behavior therapy and ongoing patient–provider contact. The aim of this article is to summarize intervention components shown to be effective in promoting successful weight control outside of pregnancy and explore potential applications in pregnancy. Available evidence suggests that the strategies shown to be effective for weight control outside of pregnancy may also promote better weight control in pregnancy, but several lines for future investigation remain.
The National Academy of Science Institute of Medicine (IOM) has reformulated specific weight gain recommendations for women during pregnancy in an effort to help prevent adverse maternal/neonatal outcomes associated with excessive gestational weight gain (GWG) [1]; however, approximately 35% of normal weight women and 60% of obese women gain more than recommended during pregnancy [2–5]. Excessive GWG is an established independent risk factor for high postpartum weight retention and future weight gain, cardiovascular disease and Type 2 diabetes in women [6–11], and is also related to several more immediate adverse obstetrical outcomes, including gestational diabetes, c-section deliveries and stillbirths [1,6–11]. The offspring of mothers with excessive GWG are also impacted and have higher rates of macrosomia and rapid early infant weight gain, future obesity and development of obesity-related comorbidities [12–14].
Unfortunately, the best strategies to prevent excessive GWG are unknown. Recent reviews [15–18] and meta-analyses [19,20] of the literature have concluded that dietary and physical activity interventions produce generally only small reductions in GWG. The reviews concurred that more research was needed examining mechanisms linking successful lifestyle interventions to reduced GWG. Interestingly, reviewers tended to diverge in recommendations for improving current treatment outcomes. Suggested strategies included greater emphasis on psychological factors, body image and motivational strategies [16], dietary, physical activity and weight monitoring strategies [19], attitudinal and motivational changes through education [20], cost-effective interventions [17], and interventions that considered the role of family, community and environmental supports [18]. Thus, there remains a wide host of potential lines of research to pursue to improve lifestyle interventions in pregnancy. The most fruitful avenues for improving treatment remain unclear, but weight control research in nonpregnant populations may prove informative in this regard. In nonpregnant populations, rapid advances over the past 30 years have identified specific approaches that are most effective in altering energy balance in favor of long-term weight control (e.g., dietary, physical activity and behavioral strategies). Findings from this literature might inform treatment strategies for preventing excessive GWG in pregnancy.
The aim of this article is to summarize key components shown to be most effective in weight loss programs for nonpregnant populations and to explore potential applications in preventing excessive GWG in pregnancy. Given the magnitude of the research literature, our article focuses on effective strategies for promoting successful weight loss, defined as 5–10% of initial body weight, in overweight/obese individuals. This magnitude of weight loss has been shown to prevent diabetes and reduce several cardiovascular disease risk factors [21]. However, it is important to note that the variables effective in promoting weight loss are the same key variables that have emerged as critical intervention targets for preventing weight gain and weight regain, as referenced in the sections that follow.
Effective weight control strategies in nonpregnant populations
The clinical research in obesity treatment and prevention has identified several dietary, physical activity and behavioral strategies effective in promoting long-term weight control
Variables associated with successful weight control in nonpregnant populations.
Calorie prescription
Meal replacements/structured meal plan
High physical activity goals (60–90 min/day)
Daily self-weighing of body weight
Daily monitoring of food intake
Behavior therapy
Continued patient-provider contact (≥2/month)
Macronutrient alteration
Breakfast
Reducing television viewing
Moderate physical activity alone (30 min/day, 5 days/week)
Social support
Cognitive strategies
Motivational interviewing
Body image, body acceptance
Education alone
Calorie restriction
Caloric restriction is the bedrock of an effective weight control program [27]. Typical lifestyle weight loss programs recommend a low calorie diet (LCD) of approximately 1200–1500 kcal/day, depending on a patient's weight. In some programs, more individualized calorie goals are prescribed. Initial intake is estimated (sometimes by multiplying weight in pounds by 12–15, depending on the person's physical activity level) and then, for weight loss, 1000 kcal/day is subtracted from this estimate in order to produce a weight loss of 2 lb/week. Tsai and Wadden performed a meta-analysis of six randomized clinical trials that compared LCDs of 1200–1500 kcal/day with more restrictive, very LCDs (VLCDs) of fewer than 800 kcal/day in obese patients [28]. They found that VLCDs induced significantly greater short-term weight losses than LCDs (16 vs 10% of initial body weight, respectively) but comparable long-term changes in weight, averaging about 8% of initial body weight. Thus, the more restrictive the calorie goal prescription, the greater the initial weight loss; however, long-term studies suggest greater weight regain with VLCD. Thus, moderate calorie restriction may be most optimal. Consistent with this evidence, the National Heart, Lung and Blood Institute (NHLBI) expert panel recommended use of LCDs over VLCDs in promoting long-term weight control [29]. Of note, effective obesity prevention programs also include specific calorie prescriptions adjusted for participants' weight status [30,31].
Interestingly, diet composition matters less than calorie prescription in promoting weight control. Varying compositions of fat, protein and carbohydrates in the diet all show significant weight loss when combined with caloric restriction [32]. Several studies have compared low carbohydrate diets (e.g., the Atkins diet) with low fat regimens [33,34]. These studies have found that, in the long-term, the weight losses are similar with these different diets [33]. Studies have also compared low glycemic with low fat, calorie-restricted diets and found no significant differences between the two [35]. There is also some evidence that increased consumption of ω-3 fatty acids can enhance short-term weight loss in obese females treated by a VLCD [36] but research in this area is only developing. What appears to impact weight loss is not the macronutrient composition, but the level of adherence to the prescribed diet (and the number of calories that are consumed). Although weight losses achieved with very different macronutrient distributions appear similar, those participants who adhere better to any of the diets lose more weight than those who adhere less well [37].
Structured meal plans & meal replacement programs
Several clinical trials have found that simplifying food selection and shopping for patients improves adherence to calorie goals [38,39]. For example, Jeffery et al. examined the effects of food provision on long-term weight control [38]. Overweight and obese participants in one group were randomly prescribed a 1000 kcal/day diet and were provided five prepackaged breakfasts and dinners; participants in the control group were prescribed the same number of calories but consumed a diet of self-selected table foods. Those receiving food provision lost significantly more weight than controls during the 18-month trial. Furthermore, in a follow-up randomized trial, these investigators showed that providing detailed meal plans was sufficient to improve both 6- and 12-month weight losses [39]. Thus, consuming portion-controlled servings of conventional foods via food provision or structured meals appear to improve long-term weight loss outcomes.
A diet that incorporates a partial meal replacement program can also effectively produce significant sustainable weight control [40–42]. A partial meal replacement program typically involves replacing one or two meals each day with a commercially available, energy-reduced product, and consuming one meal of regular foods. Several studies, reviews, and meta-analyses [40,43–45] have consistently shown that partial meal replacement improves short- and long-term dietary adherence to a low calorie diet and offers an effective, affordable and convenient option to promote adherence, weight control and improvements in metabolic risk factors [46–49]. Meal replacements also improve nutrient adequacy compared with a diet based on self-selected foods [50]. Partial meal replacement programs consistently receive high patient acceptability ratings [46–49,51], and have the added advantage of efficiency of delivery, as they can be prescribed by practitioners in busy healthcare settings without extensive adjunct therapy [51–53]. In a well-cited but moderately sized (n = 100) study, Ditschuneit et al. [46] found that patients who replaced two meals and two snacks a day with liquid shakes (and a snack bar) lost more weight (7 vs 1 kg, respectively) at 3 months than patients prescribed the same number of calories (1200–1500 kcal/day) but consuming a self-selected diet of conventional foods [46,47]. Both groups were then maintained on a diet of 1200–1500 kcal/day, which included one meal and one snack each day of meal-replacement products. Weight losses at the end of 2 years were excellent, averaging 10.2 and 7.7 kg, respectively. Although studies have found that adherence to a partial meal replacement regimen tends to wane over time [45], in this particular study adherence remained strong; after 4 years, 75% of the participants continued to replace one meal and one snack every day and maintained long-term weight loss [47]. Meal replacements and portion-controlled servings, such as frozen food entrées, eliminate the need to weigh and measure foods, save time planning and preparing meals, reduce contact with problem foods, lead to a more regular pattern of meal eating and greater control over eating when not hungry [38,39,54,55]. Such structured meal plans also reduce dietary variety, which is associated with increased food intake [56]. However, long-term adherence can remain problematic [45].
High intensity physical activity
Effective weight loss programs typically encourage patients to increase both programmed (e.g., jogging) and lifestyle (e.g., parking further away from store entrances) physical activity to aid weight loss and maintenance. Patients are told to both gradually accumulate 60–90 min of structured physical activity each day (~300 kcal/day) and, for lifestyle activity, to accumulate 10,000 steps each day [57]. Randomized trials comparing diet alone, exercise alone and the combination of diet plus exercise have consistently shown that long-term weight losses are better with the combination of diet plus exercise [58,59]. Jeffery and colleagues randomized obese patients to a 1000 kcal/week physical activity goal and others to a 2500 kcal/week goal [59]. The higher exercise goal led to better weight losses (7 vs 4 kg, respectively) after 18 months. Moreover, those who continued to adhere to this exercise goal at 30 months achieved excellent maintenance [60]. Home and supervised exercise approaches appear to produce comparable initial weight losses, but long-term results are better in home based models [61–63]. Also, prescribing exercise in short bouts and use of pedometers may improve adherence to both structured and lifestyle physical activity goals [64,65].
Behavior therapy
Behavioral treatment typically includes a combination of strategies including goal setting, self-monitoring, feedback, reinforcement, stimulus control and problem solving [24,66]. Behavioral treatment ‘packages’ have been evaluated in several trials with positive results. A meta-analysis [67] of ten studies comparing behavior therapy alone with control conditions found that behavioral intervention resulted in significantly higher weight losses (weighted mean difference = 2.5 kg) in all the studies identified. Also, when behavior therapy was combined with diet/exercise interventions versus diet/exercise alone, the combined interventions resulted in significantly greater weight losses [67]. The intensity of the behavioral intervention was also important. When the package of behavioral treatment utilized a higher number of strategies, more frequent clinical contact and/or a longer duration of intervention, the intervention effect was increased [67]. While it is difficult to tease apart the precise components that contribute to success, research has identified some salient features of behavioral treatment programs, including body weight monitoring, and diet and exercise self-monitoring, as noted below.
Body weight monitoring
Participants are given a weight goal and told to weigh themselves and to track progress towards their goal on a weight graph. Most studies have evaluated the effects of weight monitoring in the context of a comprehensive behavioral treatment and diet/exercise intervention program in which ‘self’-monitoring was combined with weekly or bimonthly practitioner weight monitoring at the time of treatment visits. Burke and colleagues conducted a systematic review of six studies that focused on weight monitoring [68]; two of the studies were randomized controlled trials with nonself-weighing control conditions [69,70]. Across studies, weight monitoring was found to improve weight loss treatment outcomes. Overall, several studies have now shown that frequent self-weighing is a significant and independent predictor of weight loss, less weight regain and avoidance of initial weight gain in adults [30,71]. Daily, as opposed to less frequent, self-weighing appears to promote the best weight control when combined with behavioral treatment programs in both weight loss and weight gain prevention studies [71–73], although the optimal dose and context (i.e., provider or self-weighing) for weight monitoring remain under investigation. Of note, frequent self-weighing is not associated with an increase in adverse psychological outcomes, including depressive symptoms, eating disorders or binge eating [74].
Diet monitoring
Self-monitoring of dietary intake is often considered the single most important component of a behavioral weight control program [75,76]. Patients are taught to write down everything they eat, to estimate or count calories in those foods using a resource book and to record a running total of their intake to ensure they stay within their calorie goal. Therapists review food and activity records and provide specific guidance on food and activity choices and supportive feedback to the participant [24,66]. A number of clinical trials have found self-monitoring of dietary intake to be significantly correlated with short- and long-term weight loss outcomes [76–78]. In a review of the literature on self-monitoring by Burke et al. [68], all 15 studies that focused on dietary self-monitoring found significant associations with weight loss. Self-monitoring is typically included in effective weight gain prevention trials, but few studies have directly examined associations with weight gain prevention [31,79]. Although adherence to self-monitoring tends to wane after 6 months [68], the advent of new technologies to monitor diet and exercise (e.g., personal digital assistants, accelerometers) could decrease patient burden in completing self-monitoring records and enhance adherence. However, more research is needed to determine the effects of using new monitoring technologies.
Increased cognitive restraint
Restrained eating may be defined as a self-initiated, cognitive attempt to restrict food intake (e.g., ‘consciously control my intake’ and ‘count calories’). Using this definition, restraint has been shown consistently to increase during behavioral treatment and to be related to greater weight loss [80–82], less weight regain [83,84] and less weight gain in general populations [85,86]. Although concerns have been raised that restraint may precipitate binge eating, clinical trials have also shown increased dietary restraint is related to reductions in binge eating [87] and bulimic symptoms [88] in women. Control over food intake appears crucial for successful weight control.
Ongoing patient-provider contact
Contact with a therapist appears to be the most effective approach to helping patients maintain their behavior changes. Several studies have shown the benefits of continuing to attend weight control classes in the long term [89–92], which has been systematically studied in a series of maintenance studies by Perri and colleagues [91,92]. Continued contact (typically weekly for 6 months and every 2 weeks thereafter) seems to provide patients with the support and motivation needed to continue to practice weight control behaviors, such as diet monitoring and exercising regularly. The nature of contact (i.e., individual vs group meetings) appears to be less important than continuing to see patients, but one randomized controlled trial found that group treatment induced a significantly larger weight loss than individual care, even in patients who indicated a preference for and received individual treatment [93].
Weight control strategies that work in combination with other strategies
Several other intervention targets are often integrated into treatment packages to improve overall outcome. These variables alone are not considered strong enough to promote sustained weight control, but in combination with effective strategies may prove beneficial.
Decreasing fat intake alone (without caloric restriction)
Historically, dietary fat restriction has been emphasized in weight gain prevention [30] and lifestyle weight control programs [94–96] due in part to the evidence that overweight individuals have higher consumption of fat [97,98]. Moreover, fat is energy dense compared with other macronutrients (i.e., fat has 9 kcal/g and carbohydrate and protein have only 4 kcal/g), and thus it is easy to overeat when consuming a high-fat diet. Modifying fat intake without calorie restriction has been show in clinical trials to induce only a 2 kg weight loss [99]. However, several clinical trials [94–96] found that weight losses were approximately 5–7 kg greater when subjects were given goals for both calories and fat compared with fat alone. Thus, most behavioral treatment programs now focus on reducing both calorie and fat intake, providing patients with a fat gram goal that is the equivalent of 25–35% kcal from fat.
Limiting fast food, soft drinks, television viewing & eating breakfast
Observational research suggest that consumption of fast food and sugar-sweetened beverages, skipping breakfast and frequent television viewing are related to weight gain in the general population [100–103]. Also, improvements in these variables have been related to successful long-term weight loss maintenance [104–107]. Randomized controlled trials in children have shown beneficial impacts on weight of reducing television and computer viewing [108], fast food consumption and sugar-sweetened beverages [109], and of adding breakfast [110]. In adults, behavioral treatment programs typically include strategies to address these, but few randomized controlled trials have specifically examined independent effects of modifying each component [111,112].
Physical activity alone (without caloric restriction)
Randomized controlled trials have shown that increasing exercise alone, without also changing dietary intake, produces only modest weight losses of approximately 2 kg; similarly the short-term effect of adding moderate (30 min most days per week) physical activity to a diet program is not very pronounced [113,114]. However, as noted above, more intensive levels of physical activity are critical for long-term weight management.
Social support
Social support has long been studied as a way to improve motivation for long-term weight loss. However, findings in this area have generally been mixed. Wing and Jeffery examined the effects of the participation of friends of recruited patients and providing a social support-focused treatment [115]. Specifically, in one group, patients were asked to identify three friends to participate, while another group attended sessions alone. Within each condition (with friends vs alone) half were randomly assigned to lifestyle weight loss behavioral treatment while half were assigned to a lifestyle weight loss program with social support strategies (e.g., team building, assertiveness in social situations). Findings indicated that both recruitment with a friend and the social support intervention improved weight loss and maintenance. Successful 10-month weight loss and maintenance was achieved by only 24% of the group recruited alone and receiving standard lifestyle weight loss treatment versus 66% of those recruited with friends and given the social support-enhanced intervention. However, other studies have shown more mixed results for social support. A meta-analysis of 21 papers examined the role of family support in weight loss and found that spouse involvement increased effectiveness [116], but findings were inconclusive. Overall, findings on the effects of social support suggest a small positive effect [117]. Support from social context is often helpful but can, for some patients, interfere with long-term weight management.
Cognitive therapy
Cognitive approaches to weight loss focus on identifying and modifying thoughts or self-perceptions associated with maladaptive diet and exercise patterns, prior relapses or prior treatment failures. A meta-analysis by Shaw et al. found that cognitive treatments compared with no treatment or behavioral treatment were generally disappointing, showing behavioral therapy to be superior in three studies and one study showing no effect of cognitive therapy [67]. However, cognitive therapy when combined with diet and exercise interventions was found to produce superior weight losses than diet and exercise treatment alone [67]. Thus, a cognitive component is often added to effective treatment packages.
Motivational interviewing
Miller and Rollnick have defined motivational interviewing (MI) as a “client centered, directive method for enhancing intrinsic motivation to change by exploring and resolving ambivalence” [118]. Studies using MI combined with various diet and exercise strategies have generally demonstrated improved weight loss and treatment adherence [119]. A review of the literature by VanWormer and Boucher [119] concluded that MI can exert a positive influence on enhancing motivation to improve diet and exercise – particularly in the short term. However, it remains unclear whether MI performs better as a standalone or adjunct to cognitive behavioral treatment packages [119].
Stimulus control
Given the role of environmental factors in influencing weight, lifestyle programs typically teach stimulus control techniques to reduce the cues for inappropriate behaviors and strengthen the cues for appropriate eating and activity. Although few randomized controlled trials have been conducted in this area, Jakicic and colleagues [58,120] found in a randomized trial that access to exercise equipment at home facilitated the maintenance of short-bout physical activity, which may improve long-term weight loss. Cohort studies have also reported that availability of physical activity equipment and less exposure to televisions and high-fat foods in the home were associated with better long-term weight control [121]. Modification of ‘cues’ in the home environment to improve physical activity and healthy eating is typically integrated into standard behavioral treatment recommendations, but additional randomized trials are needed to document effects in this area.
Lifestyle intervention ‘packages’ can be effective across multiple populations
Extensive research has examined whether certain individuals respond better than others to the effective lifestyle change and behavioral strategies described above. For the most part, treatment works regardless of demographic background. Thus, gender [81,122,123], age [81,123,124], dieting history [81,123], marital status [81], education [81] and employment [125] do not appear to predict behavioral treatment outcomes. Similarly, most research suggests that treatment works regardless of pre treatment levels of depression and anxiety [78,81,126,127], binge eating [128] and personality [129] variables. Findings are mixed on the role of ethnicity; some clinical weight loss trials have shown minorities to be somewhat less successful than nonminorities at weight loss [130,131] but as successful [132] or more successful [133] at weight loss maintenance.
Weight control strategies that do not work for weight control
Education alone
Providing patients with general information about calories, eating, activity and weight loss does little to produce weight loss. This was observed in several early studies [134], and, today, ‘education alone’ is used as a control condition in studies testing the effects of various interventions [25].
‘Nondieting’ approaches
Nondieting interventions have typically focused on helping patients improve their body image, body acceptance, mindfulness and ‘intuitive’ eating and de-emphasized caloric restriction and weight loss goals. Such interventions encourage patients to lead as full a life as possible, regardless of body weight, and teach patients how to disentangle self-worth from body weight, and/or regulate eating based on internal cues of hunger, appetite and satiety. A review of the literature of six randomized controlled trials concluded that such approaches were associated with improvements in physical activity, eating disorder pathology, blood pressure, mood, self-esteem and body image [135], but such strategies rarely promote significant weight loss [136,137].
Summary & limitations
The most effective lifestyle weight control programs appear to be those that combine calorie restriction, high levels of physical activity and behavioral strategies, including frequent self-monitoring of diet and body weight. Structured meal plans, partial meal-replacement programs and ongoing patient–provider contact also appear to improve both short- and long-term weight loss outcomes. As noted earlier, the same strategies that promote weight loss in overweight and obese individuals also appear to prevent weight gain in the general population, but the literature in this area is less robust. Also of note is that even with comprehensive behavioral weight loss programs, gradual weight regain is often observed over time, resulting in modest (~5%) overall weight losses at 3–4 years [21,25]. Although contributors to weight regain remain poorly understood, possible reasons include physiological responses to weight loss [138], daily exposure to an environment that encourages consumption of foods that are high in fat and sugar [139], and/or patient ‘burnout’ in continuing to attend weight maintenance sessions. Even with weight regain, however, improvements in long-term health are observed [21].
Weight control before, during and after pregnancy would, of course, be most optimal. However, GWG alone is an independent predictor of future health outcomes [53,140]. Thus, successful weight control during pregnancy alone could have beneficial impacts on the long-term health of both the mother and baby. Efforts to promote weight control during the window of pregnancy might benefit from the positive findings of lifestyle interventions outside of pregnancy.
Application of behavioral weight control strategies to pregnancy
There are notable differences in factors impacting weight control in pregnant and nonpregnant women. Pregnancy represents a unique physiologic state, affecting several systems that influence weight, including cardiovascular (increases in cardiac output, heart rate, stroke volume), hematologic (e.g., blood volume), urinary (e.g., renal function), respiratory (oxygen consumption), gastrointestinal (e.g., constipation, heartburn) and endocrine (edema, insulin resistance) functions. Psychosocial (body image, unwanted pregnancies, affronts to self-concept) and environmental (bed rest) changes also pose unique challenges to maintaining healthy activity levels and eating patterns during pregnancy. Moreover, unlike weight control outside of pregnancy, weight control efforts during pregnancy must be carefully regulated to ensure optimal growth and safety for the fetus.
Despite these differences, promoting weight control – whether in or out of the pregnant state – requires targeting the same basic variables. Weight gain reflects an energy imbalance wherein energy intake exceeds expenditure. Reducing intake relative to expenditure will slow or reverse this weight gain. Thus, the dietary, exercise-related and restraint-related variables that predict weight gain in nonpregnant populations [141–145] also predict weight gain in pregnant populations [146–152].
We conducted a literature search (via search engines and references lists from other papers) of English language articles published between 1985 and 2011 that evaluated interventions to reduce excessive GWG in adult (≥18 years) women. Clearly, the literature on interventions to prevent excessive GWG is still in its infancy but growing at a rapid pace. Of the ten published randomized controlled trials found in the area, seven were published between 2009 and 2011, and several more are underway
Characteristics of studies targeting gestational weight gain with open recruitment of pregnant women listed in the clinical trials registry.
GDM: Gestational diabetes; GWG: Gestational weight gain; IOM: Institute of Medicine gestational weight gain guidelines; MI: Motivational interviewing
Use of effective weight control strategies in pregnant populations
Calorie restriction & structured meal plans
Most striking, perhaps, is that every study that used calorie goals also reported a significant positive effect of the intervention on reducing GWG. As shown in
Summary of interventions to reduce excessive gestational weight gain and inclusion of intervention targets shown to be ‘effective’ or ‘moderately effective’ in nonpregnant populations
GDM: Gestational diabetes; GI: Glycemic index; GWG: Gestational weight gain; IOM: Institute of Medicine gestational weight gain guidelines; MI: Motivational interviewing; NW: Normal weight; OB: Obese; OW: Overweight.
The amount of caloric prescription was not always reported. However, in the studies specifying their prescriptions, overweight/obese women were provided goals of 18–25 kcal/kg of early pregnancy weight [157,159,161]; Mottola et al. used approximately 2000 kcal/day with a restriction not to exceed 33% of current estimated energy intake [158]. Higher calorie goals were prescribed for normal weight women based on 36 kcal/kg [159] and lower goals for morbidly obese based on 15 kcal/kg [157].
The rationale behind the selection of calorie goals was not well specified. Wolff and colleagues reported computing goals based on estimated energy expenditure that took into account energy costs of physical activity and fetal growth (Basal metabolic rate × 1.4 [physical activity level factor of 1.2 + 0.2 added to cover cost of fetal growth]), but the amount of caloric restriction was not reported [153].
Behavior therapy
Nearly all studies reported the use of behavioral strategies, typically involving goal setting and feedback based on self–monitoring records of diet, physical activity and/or weight. However, other elements of behavioral treatment (e.g., stimulus control, problem solving, reinforcement) were lacking from most studies [153,155,157–159,163–168]. Consistent with research in nonpregnant populations, the more behavioral strategies utilized, the greater the reported effects — especially when combined with a strong dietary plan. The studies that included a greater number of behavioral strategies and calorie goals found positive [154,160] effects; those with more extensive behavioral approaches but lacked calorie goals reported mixed [169,170] or nonsignificant [171] effects.
Body weight monitoring
Weight monitoring was utilized in most studies and, most commonly, monitoring was done by practitioners at the time of women s regularly scheduled prenatal visits. Every study that found a positive effect on GWG also included weight monitoring as part of the treatment package. Claesson et al. found positive effects on GWG with an intervention that did not include a strong dietary component but included frequent, weekly weight monitoring at the time of provider visits [162]. Nonetheless, some studies promoted weight monitoring and only found effects in subpopulations (i.e., low income, overweight or normal weight women only) [163,169,170,172]. In general, studies that failed to find significant effects appeared to lack weight monitoring elements [164–168,171]. Jeffries et al. specifically studied the impact of weight monitoring alone (without adjunct dietary or activity advice) in a randomized trial [163]. Pregnant women were asked to weigh and record their weight a total of seven times throughout pregnancy versus a no–advice control group. Findings indicated a positive effect for self–weighing on GWG in overweight women but not in obese or normal weight women. Whether more frequent weight monitoring would have yielded more positive results is unknown. Although this study focused on promoting an increase in ‘self–monitoring of weight, most but not all [170,172,173] of the interventions monitored weight exclusively at the time of practitioner visits and did not include a strong ‘self–monitoring’ component.
Diet monitoring
Studies that included daily diet monitoring also uniformly reported positive effects on GWG and/or better adherence to IOM guidelines in both randomized [153,156,161] and nonrandomized [157–160] trials. With the exception of two studies [154,162], studies that included less frequent (e.g., 2 days/week) or no diet monitoring showed mixed [163,169,170,172] or nonsignificant [164–168,171] effects on GWG.
Ongoing patient–provider contact
Studies with more frequent patient—provider contact during pregnancy (above and beyond usual prenatal care) — whether face to face or via telephone — also appeared to have better outcomes. The three studies that included weekly patient–provider visits [158,159,162] reported positive outcomes on reducing GWG. However, less frequent contact was more common, typically in frequency of prenatal visits, and this frequency generally yielded mixed [154,169,170] or nonsignificant [166] results.
Use of less effective weight control strategies in pregnant populations
Most studies utilized weaker intervention elements, such as modifications of specific foods/macronutrient composition, such as reducing fast food or soft drinks [163–166,168,170–173], and moderate physical activity (without caloric prescription). Studies utilizing these weaker intervention elements failed to find significant main effects on GWG (
Of note, moderate levels of physical activity were promoted in nearly all studies. The most common recommendation was 30 min/day, 3—4 days/week, which is consistent with American College of Obstetricians and Gynecologists (ACOG) [174] recommendations developed to ensure the safety of mothers and fetus. Several reviews [175–181] and an earlier meta-analysis [182] have summarized the effects of interventions to increase physical activity (and not necessarily to impact weight gain) in pregnancy. Overall, these summaries concluded that regular physical activity (30 min, 3—5 times/week) in low–risk pregnancies did not lead to adverse outcomes for the fetus or mother and improved overall maternal fitness and wellbeing. However, data on the independent effects of physical activity on maternal weight gain were insufficient [177,179], and few studies have confirmed a positive effect of exercise in controlling weight gain during pregnancy [175]. Overall, including moderate levels of physical activity did not appear to impact GWG, unless combined with a strong dietary plan. However, additional research in adequately powered trials is clearly needed.
Motivational interviewing was included in two intervention studies: one found a positive [162] and one found a nonsignificant [166] effect on GWG. Claesson et al. conducted a nonrandomized evaluation of a program that included MI with weekly or more frequent meetings, self–weighing and supervised weekly activity [162]. The results indicated that the intervention significantly reduced GWG. By contrast, Jackson and colleagues incorporated MI strategies into a less intensive personalized electronic Video doctor' intervention that targeted diet and exercise but failed to find a significant effect on GWG [166].
Limitations
Our review and analysis are exploratory and limited by several factors. Some pregnancy studies might not have described the intervention in full (and utilized intervention components not described in the published manuscript). Also, we categorized studies based on randomized versus nonrandomized designs but did not otherwise ‘weigh’ the scientific merit of studies, such as their representativeness of the US population, method of computing GWG, sample size and statistical power. Other recent reviews have addressed these topics more thoroughly [15–20]. Moreover, nearly all studies failed to measure behavioral outcomes, which clearly differ from intervention prescriptions.
Conclusion & future perspective
Interventions to prevent excessive GWG appeared most effective when they most closely mirrored effective lifestyle programs in nonpregnant populations. Calorie goals and structured meal plans, the two strongest approaches that effectively promote weight control outside pregnancy, also uniformly reduced GWG in published trials. As in nonpregnant populations, use of daily diet self–monitoring, frequent weight measurement, behavioral strategies and ongoing contact characterized the studies that were most successful in promoting weight control during pregnancy. On the Clinical Trials Database, 14 studies were identified that reported targeting GWG and are currently recruiting (Table 1). These trials will contribute valuable information; however, several lines for future investigation remain.
Use of calorie restriction
Concern that overly aggressive restriction of dietary energy may harm the fetus makes the use of calorie goals in pregnancy a thorny issue. In the past, practitioners were reluctant to place a nondiabetic, pregnant women on a calorie–controlled diet for fear of fetal growth restriction, low birth weight and starvation ketosis, which were suggested by studies examining the effects of inadequate weight gain, famine and poor nutrition in nonobese pregnant women [183–190]. However, other research suggests that reduced calories to minimize excessive weight gain in the context of a well-balanced nutritional program appeared not to increase ketonuria or result in adverse effects on perinatal morbidity, birth weight or fetal growth restriction [154,161,163,170,171,191]. Furthermore, studies suggest positive effects on reducing excessive GWG, hyperglycemia, triglycerides, incidence of gestational diabetes, c-sections, gestational hypertension and postpartum weight retention [154,161,163,170,171,191]. Moderate calorie restriction in the context of a nutritionally sound lifestyle intervention during pregnancy may be safe and effective at reducing excessive GWG. However, more research using randomized trial designs with sufficient sample sizes and longitudinal biomarkers of ketonemia are needed to rule out potential neurocognitive deficits in offspring and any adverse effects on infant and maternal body composition (bone, muscles) [1]. Estimation of the energy needs of pregnant women is complex and an accurate equation based on trimester of pregnancy has not yet been fully developed [1,192]. Nonetheless, as is evident in nonpregnant populations, precise estimation of calories need not be necessary for developing effective calorie goals that promote weight control.
Use of meal replacements
None of the GWG randomized or nonrandomized trials reviewed to date have tested the effects of a partial meal replacement program. Although liquid milk–based fortified meals have long been prescribed during pregnancy as a means of preventing inadequate weight gain and nutrient deficiencies [193–198], future research is needed to examine whether incorporating meal replacements into treatments to prevent excessive GWG can improve calorie control and nutritional adequacy and reduce excessive GWG.
Physical activity dose
None of the studies promoted physical activity levels consistent with recommendations for maintaining long-term weight loss in nonpregnant populations (60–90 min/day). Most studies recommended 30 min of moderate intensity physical activity on most days per week, which is consistent with recommendations of ACOG [174]. More well-controlled, clinical trial research with adequate sample sizes is needed on the effects of more frequent physical activity, and the optimal types and intensities during pregnancy to improve GWG and other health outcomes [175–182,199–201].
Frequency of self-weighing
Frequency of self-weighing during pregnancy also remains a topic of debate. In the UK, practitioners are advised against weighing women during pregnancy based on anecdotal evidence of potential negative emotional impacts. Also, outside of pregnancy, concerns have been raised over the potential adverse effects of self-weighing on mood and increased risk of developing eating disorders [202,203]. However, there is little evidence to support these concerns in or outside of pregnancy [204,205,301]. Studies in nonpregnant populations have found no adverse impact of frequent self-weighing on depressive symptoms, binge eating or eating disorder psycho-pathology, and found positive effects on weight control [68,74]. In our review, none of the nine studies that used weight monitoring reported adverse effects; moreover, all reported either positive effects on GWG overall [153,154,157,160,162] or in subgroups of low income [172], overweight [163] or normal weight [169,170] women. Similarly, in a recent meta–analysis of the literature, Streuling et al. concluded that weight monitoring appeared to help promote successful lowering of GWG when combined with dietary and physical activity strategies [19]. However, the emotional impact of self-weighing was not directly measured in studies of pregnant women to date, and the optimal frequency of weighing (daily, weekly) and context (part of larger treatment package vs alone) remain to be determined.
Increasing cognitive restraint
Few pregnancy intervention trials to date have reported directly targeting and/or measuring intervention-related changes in cognitive restraint [173]. As in nonpregnant populations, women with higher restraint early in pregnancy tend to be heavier [206] and gain more weight than less restrained eaters [152,207,208]. Restraint declines during pregnancy [152] but whether such declines relate to GWG is unclear. Given the strong relationship between increases in restraint and better weight control in nonpregnant populations [80–84,209,210], more research is needed to determine whether preventing pregnancy-related declines in restraint or enhancing restraint during pregnancy will improve GWG outcomes.
Frequency of patient-provider contact
As in nonpregnant populations, more frequent visits with practitioners appeared to promote better weight control. Pregnancy interventions have the advantage of capitalizing on the ongoing contact that occurs as part of standard obstetrical care. However, most women see providers monthly or less frequently early in pregnancy and then increase frequency of contact towards the end of pregnancy. This pattern may be sub optimal in allowing interventions to reduce excessive GWG, in fact, the best time for intervention may be preconceptionally [1]. In nonpregnant populations, greater success early in a weight control program is a strong and consistent predictor of later weight control success [211–213]. Thus, future research is needed to examine whether intervening earlier in pregnancy yields better outcomes and to identify the most effective frequency of contact (weekly, monthly). Also, the feasibility and cost–benefit of weekly meetings in the context of US healthcare requires study.
In conclusion, research in weight control literature over the past 30 years has shown that eating and physical activity habits can be changed to promote successful weight control in nonpregnant populations. The growing literature on pregnancy suggests that it is possible to intervene to reduce GWG, but many challenges remain. It is likely that the strategies shown to be effective for weight control outside of pregnancy may also promote better weight control and prevent excessive weight gain during pregnancy. However, more research is needed to study the most effective methods to prevent excessive GWG and the short- (conception through to 1 year) and long-term (adolescence through to adulthood) impact of altering GWG on child outcomes including duration of gestation, neurocognitive deficits, weight and body composition, and development of obesity and related comorbidities [1]. In addition to improving the effectiveness of interventions aimed at individuals, it is likely that excessive GWG is a function of the same neighborhood and environmental factors that cause obesity in children and other adults. The time is ripe for rigorous testing of the safety and effects of more intensive and multilevel weight control programs to prevent excessive GWG and, thus, promote optimal health for women and their offspring.
Executive summary
Moderate caloric restriction (low calorie diet) is the bedrock of an effective weight control program. Varying compositions of fat, protein and carbohydrates in the diet all show clinically meaningful weight loss when combined with caloric restriction.
Use of meal replacements and structured meal plans appears to improve adherence to calorie goals. Several studies, reviews and meta-analyses have consistently demonstrated that partial meal replacement improves short- and long-term dietary adherence to a low calorie diet and offers an effective, affordable and convenient option to promote adherence, weight control and improvements in metabolic risk factors.
High physical activity (60–90 min/day) and lifestyle (10,000 steps/day) goals are integral to effective weight loss programs. Randomized trials comparing diet alone, exercise alone and the combination of diet plus exercise have consistently demonstrated that long–term weight losses are better with the combination of diet plus exercise.
Daily self-weighing of the body is a significant predictor of weight loss, less weight regain and avoidance of initial weight gain in adults. Frequent self-weighing is not associated with adverse psychological outcomes.
Daily monitoring of food intake and exercise is often considered the single most important component of a behavioral weight control program. A number of clinical trials have found daily self–monitoring of intake and activity to be significantly correlated with short- and long-term weight loss outcomes.
Behavior therapy typically includes a combination of strategies including stimulus control, reinforcement, self-monitoring, problem solving and goal setting. A meta-analysis often studies comparing behavior therapy with control conditions found that behavioral intervention resulted in significantly better weight losses in all the studies identified.
Continued patient-provider contact (typically weekly for 6 months and every 2 weeks thereafter) seems to provide patients with the support and motivation to continue to practice weight control behaviors, such as keeping food records and exercising regularly.
A total of 20 published intervention studies have examined effects of interventions to prevent excessive gestational weight gain in adult women; 12 were randomized controlled trials and eight were nonrandomized comparisons; nine reported positive effects on gestational weight gain (GWG); four reported effects in certain subgroups; and seven reported nonsignificant effects.
Every study that used calorie goals also reported a significant positive effect on reducing GWG. Overweight/obese women were generally provided goals of 18–25 kcal/kg of early pregnancy weight; higher calorie goals were prescribed for normal weight women based on 36 kcal/kg and lower goals for morbidly obese women based on 15 kcal/kg.
Studies that included daily diet monitoring uniformly reported positive effects on GWG.
Every study that found a positive effect on GWG included weight monitoring as part of the treatment package.
Studies with more frequent contact during pregnancy appeared to have better outcomes.
Nearly all studies reported use of some behavioral strategies, typically self-monitoring with feedback. Studies that included more extensive behavior therapy with calorie goals found positive effects.
Including moderate levels of physical activity did not appear to yield positive effects on GWG, unless combined with a strong dietary plan.
Interventions to prevent excessive gestational weight gain appear most effective when they more closely mimic effective behavioral lifestyle programs in nonpregnant populations.
There remain several lines for future investigation, including identification of optimal calorie goals, frequency of self-weighing and patient–provider contact, and a dose of physical activity during pregnancy.
Footnotes
This research was supported in part by NIH grant DK071667; the NIH was not involved in the design and conduct of the study; collection, management, analysis and interpretation of the data; or preparation, review or approval of the manuscript. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.
