Advertisement

Validity of perceived weight gain in women using long-acting reversible contraception and depot medroxyprogesterone acetate

Published:October 24, 2012DOI:https://doi.org/10.1016/j.ajog.2012.10.876

      Objective

      The objective of the study was to evaluate perceived weight gain in women using contraception and determine the validity of self-reported weight gain.

      Study Design

      We analyzed data from new contraceptive method users who self-reported a weight change at 3, 6, and 12 months after enrollment. We examined a subgroup of participants with objective weight measurements at baseline and 12 months to test the validity of self-reported weight gain.

      Results

      Thirty-four percent of participants (1407 of 4133) perceived weight gain. Compared with copper intrauterine device users, implant users (relative risk, 1.29; 95% confidence interval, 1.10–1.51) and depot medroxyprogesterone acetate users (relative risk, 1.37; 95% confidence interval, 1.14–1.64) were more likely to report perceived weight gain. Women who perceived weight gain experienced a mean weight gain of 10.3 pounds. The sensitivity and specificity of perceived weight gain were 74.6% and 84.4%, respectively.

      Conclusion

      In most women, perceived weight gain represents true weight gain. Implant and depot medroxyprogesterone acetate users are more likely to perceive weight gain among contraception users.

      Key words

      Reported weight gain is one of the main reasons that women are not satisfied with their contraceptive method.
      • Bahamondes L.
      • Pinho F.
      • de Melo N.R.
      • Oliveira E.
      • Bahamondes M.V.
      [Associated factors with discontinuation use of combined oral contraceptives].
      • Hani D.
      • Imthurn B.
      • Merki-Feld G.S.
      [Weight gain due to hormonal contraception: myth or truth?].
      In clinical studies evaluating the subdermal implant, 13.7% of users reported weight gain as an adverse effect of the method, and 2.3% reported weight gain as the reason for having their implant removed (Implanon package insert, 2006; Organon USA Inc, Roseland, NJ). However, perceived weight gain is rarely validated with objective measures.
      The literature regarding contraception and objective measurements of weight gain has focused on certain methods, mainly oral contraceptives and depot medroxyprogesterone acetate (DMPA). Currently DMPA is the only contraceptive method consistently associated with significant weight gain.
      • Berenson A.B.
      • Rahman M.
      Changes in weight, total fat, percent body fat, and central-to-peripheral fat ratio associated with injectable and oral contraceptive use.
      • Bonny A.E.
      • Ziegler J.
      • Harvey R.
      • Debanne S.M.
      • Secic M.
      • Cromer B.A.
      Weight gain in obese and nonobese adolescent girls initiating depot medroxyprogesterone, oral contraceptive pills, or no hormonal contraceptive method.
      • Pantoja M.
      • Medeiros T.
      • Baccarin M.C.
      • Morais S.S.
      • Bahamondes L.
      • Fernandes A.M.
      Variations in body mass index of users of depot-medroxyprogesterone acetate as a contraceptive.
      However, the literature fails to address how women perceive various contraceptive methods to affect their weight and whether their perception of weight gain is accurate.
      Regardless of accuracy, a woman's perception of her weight while using a contraceptive method may be just as important as the number on a scale, particularly regarding method satisfaction and continuation. In a study on self-weighing frequency and weight loss, 20% of adults at baseline reported they never weigh themselves, and 40% reported weighing themselves less than once a week.
      • Linde J.A.
      • Jeffery R.W.
      • French S.A.
      • Pronk N.P.
      • Boyle R.G.
      Self-weighing in weight gain prevention and weight loss trials.
      Therefore, many Americans rely on perceived weight as a method of weight assessment.
      The purpose of this analysis was to evaluate perceived weight gain during the first 12 months after women started a new contraceptive method and to determine the validity of self-reported weight gain. To test the accuracy of perceived weight gain, we examined a subset of women who also received height and weight measurements at baseline and 12 months after enrollment.

      Materials and Methods

      This was a secondary analysis of the Contraceptive CHOICE Project. CHOICE is a prospective cohort study developed to promote the use of long-acting reversible contraception (LARC). The study provides contraception for 2-3 years at no cost to 9256 women in the St. Louis area in an effort to reduce unintended pregnancies in our region.
      English- or Spanish-speaking women between 14 and 45 years of age were eligible for the study if they met the following criteria: (1) resided in the study's designated recruitment region; (2) were willing to start a new method of reversible contraception or were currently not using any method; (3) did not desire to become pregnant for at least 12 months; and (4) were sexually active with a male partner or planning to become sexually active in the next 6 months. Women were excluded if they had a history of tubal sterilization or hysterectomy. Study enrollment began in August 2007 and ended in September 2011.
      All enrolled participants received contraceptive counseling at one of the designated recruitment sites, which included the university-based clinic, 2 abortion clinics, and several community-based clinics. Counseling sessions at the university-based clinic presented women with all reversible contraceptive methods and their associated effectiveness, side effects, benefits, and risks to help participants make an informed decision. Counseling sessions at the remaining clinics were based on what was usual contraceptive counseling for that clinic location. In addition, during the baseline visit, a comprehensive assessment was performed to collect demographic and reproductive history information, a screening was conducted for sexually transmitted infections, and measurement of height and weight was done. The woman's baseline body mass index (BMI) was calculated as weight (kilograms)/height (meters)2 and categorized into 1 of the following BMI groups: underweight (<18.5 kg/m2), normal (18.5-24.9 kg/m2), overweight (25-29.9 kg/m2), or obese (≥30 kg/m2).
      We performed telephone interviews using standardized survey instruments at 3 and 6 months after enrollment and every 6 months for the duration of the follow-up. Participants were compensated with a $10.00 gift card at each interview. During the follow-up surveys, participants were asked to self-report any changes in their weight. Specifically, the survey question asked, “Since we last spoke, has your weight changed by 5 pounds or more?” If the participant answered yes, she was further asked, “Did you (a) gain weight, (b) lose weight, (c) both gain and lose weight, or (d) you don't know?” This question was not included until a revised version of the survey was introduced on July 1, 2008. Therefore, only participants who answered this question are included in this analysis.
      For this analysis, we evaluated women's perceived weight change during the first 12 months of starting a new contraceptive method using data from their 3, 6, and 12 month surveys. A participant was excluded from the analysis if we could not reach a definite conclusion regarding her direction of perceived weight change over 12 months. For example, a participant was excluded if she was missing a response to the question above in any of the 3, 6, or 12 month surveys or had responded yes to weight change but was missing a response to how her weight had changed (ie, weight gain or loss). In addition, if a participant ever reported both weight gain and weight loss or she did not know the answer to the question, “How did your weight change?” in any of the 3, 6, or 12 month follow-up surveys, she was excluded from the analysis sample because we could not determine the direction (gain or loss) of her weight change. Finally, if a participant's combined responses of all 3 surveys included any combination of gain and loss, she was excluded because we could not determine her net sum of perceived weight change because we had no way of knowing which weight change was the most prominent.
      We defined 3 groups: (1) perceived weight gain; (2) perceived no weight change; and (3) perceived weight loss. Group 1 (perceived weight gain) included any participant who reported weight gain in at least one of her 3, 6, or 12 month surveys and reported no change at all of the remaining 3, 6, or 12 month survey(s). Group 2 (perceived no weight change) included participants who, at all of the 3, 6, and 12 month follow-up surveys, responded, “No, my weight has not changed by 5 pounds or more since we last spoke.” Group 3 (perceived weight loss) included any participant who reported weight loss in at least 1 of her 3, 6, or 12 month surveys and reported no change at all of the remaining 3, 6, or 12 month survey(s).
      All women enrolled in CHOICE on or after June 15, 2010, were offered eligibility screening and, if eligible, enrolled in a separate substudy to objectively assess 12 month weight change with progestin-only methods or the copper intrauterine device (IUD). Eligibility for this substudy included the following: (1) CHOICE enrollment at the university-based clinic to retain consistency in the scale used to assess weight; (2) age of 18 years or older; (3) using the levonorgestrel intrauterine system (LNG-IUS), copper T380A IUD, subdermal implant, or DMPA as the current method; (4) continued use of 1 of the previously mentioned methods for at least 11 months; and (5) a willingness to return to the university-based clinic for a 12 month visit for an objective weight assessment.
      At the time of data analysis, 415 women had enrolled and completed this substudy of objectively measured weight change. We calculated the participant's observed weight change by subtracting her recorded weight at baseline from the recorded weight at her 12 month follow-up. The observed weight change was classified as 1 of the following: a gain if the calculated weight change was a 5 pound or greater increase; no change if the calculated weight change was less than a 5 pound change in either direction; and a loss if the calculated weight change was a 5 pound or greater decrease. We chose a 5 pound change to be consistent with our survey instrument.
      The observed weight change categories of gain, no change, and loss were then compared with the perceived weight change groups (1, 2, and 3) as previously defined. To estimate sensitivity, specificity, agreement, and relative risk calculations, we combined group 2 (perceived no weight change) and group 3 (perceived weight loss) to create a dichotomous variable for perceived weight gain: yes (perceived weight gain) or no (no perceived weight gain). We created a similar variable for objective weight gain, in which the no-weight-gain group included the measured no-change and weight-loss groups.
      The CHOICE protocol and substudy described in the previous text were approved by the Washington University in St. Louis School of Medicine Human Research Protection Office. The methodic details of the Contraceptive CHOICE Project have been published in a separate report.
      • Secura G.M.
      • Allsworth J.E.
      • Madden T.
      • Mullersman J.L.
      • Peipert J.F.
      The Contraceptive CHOICE Project: reducing barriers to long-acting reversible contraception.

      Statistical Analyses

      We compared the baseline demographic and behavioral characteristics among the 3 perceived weight change groups (gain, no change, and loss) in both the CHOICE analysis sample and our substudy sample using χ2 or Fisher exact test where appropriate. We compared the mean objective weight change of the 3 perceived weight gain groups in the substudy using a 1-way ANOVA test. We calculated a kappa statistic to evaluate the agreement between objective and perceived weight gain. We calculated the sensitivity and specificity of perceived weight gain compared with the gold standard of objectively measured weight gain.
      We also calculated the relative risk (RR) of perceived weight gain by contraceptive methods and 95% confidence intervals (CIs) from Poisson regression using the SAS GENMOD procedure (SAS Institute, Cary, NC). This approach provides an unbiased estimate of the relative risk in the case of common outcomes (>10%). Confounding was defined as a greater than 10% relative change in the association between perceived weight gain and method choice with or without the potential confounding factor in the model. Confounders were included in the final model to obtain an adjusted RR of perceived weight gain among contraceptive methods. All statistical analysis was performed using SAS 9.2 software. The significance level alpha was set at 0.05.

      Results

      Of the 7977 CHOICE participants who reached 12 months at the time of data analysis, 4133 women met our inclusion criteria. Among the 3844 women excluded, 2968 had missing data regarding the weight question or the specific direction of weight change and 1146 reported both weight gain and loss. We found no significant differences in demographic and behavioral characteristics between the analysis sample and the women who were excluded because of inclusion criteria or inability to reach a definite conclusion regarding their weight change. Of the 4133 women in the analysis sample, 1407 (34.0%) perceived weight gain, 1634 (39.5%) perceived no weight change, and 1092 (26.4%) perceived weight loss.
      Table 1 displays the demographic and behavioral characteristics of the analysis sample and the 3 perceived weight change groups: gain, no change, and loss. Women who perceived weight gain were significantly more likely to be African American, parous, uninsured, and less educated. The participants in the weight-gain group were also more likely to have trouble paying for basic necessities and receive public assistance. General health was highest among those who reported no perceived weight change. Compared with women who were normal or underweight at baseline (n = 1882), women who were overweight or obese at baseline (n = 2251) were more likely to report perceived weight gain (36.7% vs 30.9%) and weight loss (30.5% vs 21.5%) and less likely to report no weight change (32.8% vs 47.6%).
      TABLE 1Baseline demographic and behavioral characteristics of study participants
      CharacteristicsAnalysis sample (n = 4133)Perceived weight gain (n = 1407)Perceived no weight change (n = 1634)Perceived weight loss (n = 1092)P value
      P value calculated using Fisher exact test. All other P values calculated using χ2 test;
      Age, y
      Significant at alpha level of .05.
      .042
       <21851 (20.6)300 (21.3)342 (20.9)209 (19.1)
       21-251624 (39.3)511 (36.3)670 (41.0)443 (40.6)
       >251658 (40.1)596 (42.4)622 (38.1)440 (40.3)
      Race
      Significant at alpha level of .05.
      < .001
       Black1989 (48.1)823 (58.5)725 (44.4)441 (40.4)
       White1810 (43.8)472 (33.5)778 (47.6)560 (51.3)
       Other334 (8.1)112 (8.0)131 (8.0)91 (8.3)
      Marital status
      Significant at alpha level of .05.
      .011
       Single2432 (58.9)845 (60.1)972 (59.6)615 (56.3)
       Living with partner900 (21.8)289 (20.6)345 (21.1)266 (24.4)
       Married551 (13.3)169 (12.0)234 (14.3)148 (13.5)
       Separated/divorced/widowed247 (6.0)103 (7.3)81 (5.0)63 (5.8)
      Insurance
      Significant at alpha level of .05.
      < .001
       None1668 (40.6)641 (45.8)621 (38.2)406 (37.3)
       Private1873 (45.5)535 (38.3)806 (49.6)532 (48.9)
       Public571 (13.9)222 (15.9)199 (12.2)150 (13.8)
      Education
      Significant at alpha level of .05.
      < .001
       High school diploma/GED or less1363 (33.0)546 (38.8)490 (30.0)327 (30.0)
       Some college1662 (40.2)572 (40.7)639 (39.1)451 (41.3)
       College degree or graduate school1107 (26.8)288 (20.5)505 (30.9)314 (28.7)
      Trouble paying for basic expenses
      Significant at alpha level of .05.
      < .001
       No2608 (63.2)771 (54.9)1138 (69.8)699 (64.0)
       Yes1519 (36.8)633 (45.1)493 (30.2)393 (36.0)
      Receiving public assistance
      Significant at alpha level of .05.
      < .001
       No2680 (64.9)816 (58.1)1136 (69.6)728 (66.7)
       Yes1448 (35.1)589 (41.9)496 (30.4)363 (33.3)
      Parity
      Significant at alpha level of .05.
      < .001
       Nulliparous2030 (49.1)608 (43.2)867 (53.1)555 (50.8)
       Parous2103 (50.9)799 (56.8)767 (46.9)537 (49.2)
      Baseline BMI, kg/m2
      Significant at alpha level of .05.
      < .001
       Underweight (<18.5)211 (5.1)45 (3.2)125 (7.7)41 (3.8)
       Normal (18.5-24.9)1671 (40.4)537 (38.2)770 (47.1)364 (33.3)
       Overweight (25-29.9)1040 (25.2)376 (26.7)360 (22.0)304 (27.8)
       Obese (≥30)1211 (29.3)449 (31.9)379 (23.2)383 (35.1)
      Baseline method
      Significant at alpha level of .05.
      < .001
       LNG-IUS1892 (45.8)650 (46.2)705 (43.2)537 (49.2)
       Copper IUD532 (12.9)156 (11.1)212 (13.0)164 (15.0)
       Implant651 (15.8)266 (18.9)233 (14.3)152 (13.9)
       DMPA282 (6.8)130 (9.3)102 (6.2)50 (4.6)
       Pills, patch, ring774 (18.7)204 (14.5)381 (23.3)189 (17.3)
      General health (baseline)
      Significant at alpha level of .05.
      < .001
       Excellent1199 (29.1)352 (25.1)545 (33.4)302 (27.7)
       Very good1681 (40.7)549 (39.2)659 (40.4)473 (43.3)
       Good1016 (24.6)399 (28.4)365 (22.3)252 (23.1)
       Fair215 (5.2)94 (6.7)61 (3.7)60 (5.5)
       Poor15 (0.4)8 (0.6)3 (0.2)4 (0.4)
      General health (12 months)
      Significant at alpha level of .05.
      < .001
       Excellent1012 (24.6)267 (19.0)475 (29.2)270 (24.8)
       Very good1651 (40.1)522 (37.2)689 (42.4)440 (40.3)
       Good1115 (27.1)433 (30.9)383 (23.6)299 (27.4)
       Fair322 (7.8)167 (11.9)75 (4.6)80 (7.3)
       Poor19 (0.4)14 (1.0)3 (0.2)2 (0.2)
      Data are n (percentage) unless otherwise specified. Some n values do not sum to total number in column heading, which accounts for missing responses.
      BMI, body mass index; DMPA, depot medroxyprogesterone acetate; GED, general education development test; IUD, intrauterine device; LNG-IUS, levonorgestrel intrauterine system.
      Nault. Perceived weight gain in women. Am J Obstet Gynecol 2013.
      a P value calculated using Fisher exact test. All other P values calculated using χ2 test;
      b Significant at alpha level of .05.
      Table 2 presents the demographic and behavioral characteristics for the subgroup of participants who received an objective measurement of weight change at baseline and 12 months. Of the 415 women who completed this assessment, 281 participants met the inclusion criteria. We found no differences when we compared excluded participants with those included in the analysis sample. Black race, lower socioeconomic status, higher baseline BMI, and use of DMPA or the implant were significantly associated with perceived weight gain.
      TABLE 2Baseline demographic and behavioral characteristics of women with objective weight measurements at baseline and 12 months
      CharacteristicsSubstudy analysis sample (n = 281)Perceived weight gain (n = 111)Perceived no weight change (n = 100)Perceived weight loss (n = 70)P value
      Age, y.836
       <2137 (13.2)15 (13.5)15 (15.0)7 (10.0)
       21-25110 (39.1)45 (40.5)39 (39.0)26 (37.1)
       >25134 (47.7)51 (46.0)46 (46.0)37 (52.9)
      Race
      Significant at alpha level of .05;
      < .001
      P value was calculated using Fisher exact test. All other P values were calculated using χ2 test;
       Black141 (50.2)75 (67.6)40 (40.0)26 (37.2)
       White128 (45.5)35 (31.5)53 (53.0)40 (57.1)
       Other12 (4.3)1 (0.9)7 (7.0)4 (5.7)
      Marital status.413
      P value was calculated using Fisher exact test. All other P values were calculated using χ2 test;
       Single156 (55.5)64 (57.7)57 (57.0)35 (50.0)
       Living with partner66 (23.5)28 (25.2)17 (17.0)21 (30.0)
       Married40 (14.2)12 (10.8)18 (18.0)10 (14.3)
       Separated/divorced/widowed19 (6.8)7 (6.3)8 (8.0)4 (5.7)
      Insurance
      Significant at alpha level of .05;
      .011
       None90 (32.0)44 (39.6)28 (28.0)18 (25.7)
       Private151 (53.7)51 (46.0)64 (64.0)36 (51.4)
       Public40 (14.2)16 (14.4)8 (8.0)16 (22.9)
      Education.266
       High school diploma/GED or less65 (23.1)30 (27.1)19 (19.0)16 (22.9)
       Some college110 (39.2)44 (39.6)35 (35.0)31 (44.3)
       College degree or graduate school106 (37.7)37 (33.3)46 (46.0)23 (32.9)
      Trouble paying for basic expenses
      Significant at alpha level of .05;
      .004
       No176 (62.9)56 (50.9)70 (70.0)50 (71.4)
       Yes104 (37.1)54 (49.1)30 (30.0)20 (28.6)
      Receiving public assistance
      Significant at alpha level of .05;
      .001
       No180 (64.1)60 (54.1)78 (78.0)42 (60.0)
       Yes101 (35.9)51 (45.9)22 (22.0)28 (40.0)
      Parity
      Significant at alpha level of .05;
      .018
       Nulliparous129 (45.9)46 (41.4)57 (57.0)26 (37.1)
       Parous152 (54.1)65 (58.6)43 (43.0)44 (62.9)
      Baseline BMI by CDC classifications, kg/m2
      Significant at alpha level of .05;
      .035
      P value was calculated using Fisher exact test. All other P values were calculated using χ2 test;
       Underweight (<18.5)9 (3.2)2 (1.8)5 (5.0)2 (2.9)
       Normal (18.5-24.9)121 (43.0)39 (35.2)55 (55.0)27 (38.6)
       Overweight (25-29.9)64 (22.8)28 (25.2)19 (19.0)17 (24.3)
       Obese (≥30)87 (31.0)42 (37.8)21 (21.0)24 (34.3)
      Baseline method
      Significant at alpha level of .05;
      .004
      P value was calculated using Fisher exact test. All other P values were calculated using χ2 test;
       LNG-IUS99 (35.2)38 (34.2)29 (29.0)32 (45.7)
       Copper-IUD73 (26.0)18 (16.2)35 (35.0)20 (28.6)
       Implant78 (27.8)37 (33.4)26 (26.0)15 (21.4)
       DMPA31 (11.0)18 (16.2)10 (10.0)3 (4.3)
      General health.598
      P value was calculated using Fisher exact test. All other P values were calculated using χ2 test;
       Excellent77 (27.4)34 (30.6)25 (25.0)18 (25.7)
       Very good123 (43.8)50 (45.1)45 (45.0)28 (40.0)
       Good73 (26.0)23 (20.7)28 (28.0)22 (31.4)
       Fair7 (2.5)4 (3.6)1 (1.0)2 (2.9)
       Poor1 (0.3)0 (0.0)1 (1.0)0 (0.0)
      General health at 12 months.381
      P value was calculated using Fisher exact test. All other P values were calculated using χ2 test;
       Excellent67 (24.0)21 (19.1)28 (28.0)18 (26.1)
       Very good124 (44.4)48 (43.6)42 (42.0)34 (49.3)
       Good71 (25.5)34 (30.9)25 (25.0)12 (17.4)
       Fair16 (5.7)7 (6.4)4 (4.0)5 (7.2)
       Poor1 (0.4)0 (0.0)1 (1.0)0 (0.0)
      Objective weight
      Significant at alpha level of .05;
      < .001
      P value was calculated using Fisher exact test. All other P values were calculated using χ2 test;
       Gain114 (40.6)85 (76.6)26 (26.0)3 (4.3)
       No change99 (35.2)21 (18.9)56 (56.0)22 (31.4)
       Loss68 (24.2)5 (4.5)18 (18.0)45 (64.3)
      Mean weight change, lb
      Significant at alpha level of .05;
      2.2 ± 12.210.3 ± 9.91.5 ± 8.1–9.5 ± 10.5< .001
      P value was calculated using 1-way analysis of variance.
      Data are n (percentage) or mean ± SD unless otherwise specified. Some n values do not sum to total number in column heading, which accounts for missing responses.
      BMI, body mass index; CDC, Centers for Disease Control and Prevention; DMPA, depot medroxyprogesterone acetate; GED, general education development test; IUD, intrauterine device; LNG-IUS, levonorgestrel intrauterine system.
      Nault. Perceived weight gain in women. Am J Obstet Gynecol 2013.
      a Significant at alpha level of .05;
      b P value was calculated using Fisher exact test. All other P values were calculated using χ2 test;
      c P value was calculated using 1-way analysis of variance.
      The mean weight change over 12 months for the total subset of women with objective weight change assessment was a 2.2 pound increase. The 3 perceived weight change groups (gain, no change, and loss) experienced a mean weight change of 10.3 pounds gained, 1.5 pounds gained, and 9.5 pounds lost, respectively (P < .001). Of the 114 women who objectively gained 5 pounds or more during the 12 month period, 85 perceived weight gain. Conversely, among the 167 who did not gain 5 or more pounds, 141 perceived no weight gain. Therefore, the sensitivity and specificity of perceived weight gain was 74.6% (85 of 114) and 84.4% (141 of 167), respectively. Additionally, we calculated a kappa statistic to measure the agreement between perceived weight gain and actual weight gain. The tests demonstrated moderate to good agreement (kappa = 0.59, P < .0001). The positive predictive value for perceived weight gain was 77%.
      The Figure presents perceived weight change by contraceptive method. Each bar represents the total number of women using the specified method at baseline. The subdermal implant and DMPA showed the most perceived weight gain of all reversible methods analyzed. Forty-one percent of the implant users and 46% of the DMPA users perceived a weight gain of 5 pounds or more. Table 3 shows the crude and adjusted relative risk of perceived weight gain for each contraceptive method compared with the copper IUD. After adjusting for race, the implant (RR, 1.29; 95% CI, 1.10–1.51) and DMPA (RR, 1.37; 95% CI, 1.14–1.64) users were significantly more likely to perceive a weight gain compared with copper IUD users. The LNG-IUS, the pill, the patch, and the ring users were no more likely to perceive weight gain compared with the copper IUD users.
      Figure thumbnail gr1
      FIGUREPercent of women reporting weight change by baseline contraceptive method
      Nault. Perceived weight gain in women. Am J Obstet Gynecol 2013.
      TABLE 3Risk of perceived weight gain and contraceptive method
      MethodRelative risk (95% CI)
      CrudeAdjusted
      Model adjusted for race.
      LNG-IUS1.17 (1.01–1.35)1.13 (0.98–1.30)
      Copper-IUDReferenceReference
      Implant1.39 (1.18–1.63)1.29 (1.10–1.51)
      DMPA1.57 (1.31–1.88)1.37 (1.14–1.64)
      Pills, patch, ring0.90 (0.75–1.07)0.88 (0.74–1.05)
      CI, confidence interval; DMPA, depot medroxyprogesterone acetate; IUD, intrauterine device; LNG-IUS, levonorgestrel intrauterine system.
      Nault. Perceived weight gain in women. Am J Obstet Gynecol 2013.
      a Model adjusted for race.

      Comment

      We found perceived weight gain to be a reasonable measure of objective weight gain. Although not a perfect measure, most women who self-reported weight gain (77%) were objectively found to have gained 5 pounds or more over 12 months. In addition, participants who perceived weight gain were found to have gained an average of 8.8 pounds more over 12 months than the group that perceived no change (P < .001). Our results suggest that self-reported weight change may be a reasonable proxy for true weight gain and may be a practical way for clinicians and epidemiologists to monitor patients' weight changes. Pronk et al
      • Pronk N.P.
      • Crain A.L.
      • Vanwormer J.J.
      • Martinson B.C.
      • Boucher J.L.
      • Cosentino D.L.
      The use of telehealth technology in assessing the accuracy of self-reported weight and the impact of a daily immediate-feedback intervention among obese employees.
      also noted that self-reported weight is an acceptable alternative, given circumstances in which it is difficult or inefficient to obtain measured weight.
      Currently more than half of reproductive-age women in the United States have a BMI greater than 25 kg/m2, which, according to the Centers for Disease Control and Prevention's standard weight status categories, classifies them as overweight or obese.
      • Rahman M.
      • Berenson A.B.
      Self-perception of weight gain among multiethnic reproductive-age women.
      Similarly, 54.5% of women in our analysis had a BMI greater than 25 kg/m2 at baseline, placing them into the overweight or obese BMI categories. More than one third of the overweight and obese women in our analysis (37%) perceived weight gain. This finding is alarming; in addition to already being at an unhealthy weight, many subsequently reported a weight gain of at least 5 pounds over 12 months. Our results indicate that this is true weight gain for most women.
      Because of the many adverse physical and mental health effects that commonly accompany obesity, it is important for clinicians to be aware of their patients' weight gains. It may be useful, as well as simple and cost effective, to have women report perceived weight changes to their primary care provider at regular intervals. This perceived change may be a trigger for an objective assessment. If validated, the clinician may suggest weight loss strategies or consider screening for diseases associated with obesity, such as diabetes or hypertension.
      We were interested in perceived weight gain and its association with certain contraceptive methods. We found that almost half of implant and DMPA users (40.9% and 46.1%, respectively) perceived weight gain. As stated in previous text, DMPA is the 1 method that has been consistently associated with significant weight gain in previous studies.
      • Berenson A.B.
      • Rahman M.
      Changes in weight, total fat, percent body fat, and central-to-peripheral fat ratio associated with injectable and oral contraceptive use.
      • Bonny A.E.
      • Ziegler J.
      • Harvey R.
      • Debanne S.M.
      • Secic M.
      • Cromer B.A.
      Weight gain in obese and nonobese adolescent girls initiating depot medroxyprogesterone, oral contraceptive pills, or no hormonal contraceptive method.
      • Pantoja M.
      • Medeiros T.
      • Baccarin M.C.
      • Morais S.S.
      • Bahamondes L.
      • Fernandes A.M.
      Variations in body mass index of users of depot-medroxyprogesterone acetate as a contraceptive.
      To our knowledge, there are few studies addressing weight gain or perceived weight gain with the use of the etonogestrel implant. Future studies should determine whether the subdermal implant is associated with the objective evidence of weight gain. Understanding that some contraceptive methods have higher rates of perceived weight gain could prove helpful to clinicians when counseling their patients. It also may be helpful for clinicians to understand that certain patient characteristics (eg, race, social economic status, education, parity, baseline BMI, etc) are more likely to be associated with perceived weight gain.
      Strengths of our analysis include a relatively large sample size and a diverse sample in terms of race and socioeconomic status. Our population reflects the St. Louis population but may not be generalizable to other US populations. As a note of caution, the perceived weight gain group in Table 1 (n = 1407) is 10 times as large as the objective weight gain group in Table 2 (n = 111). Thus, associations in the perceived weight gain group were more likely to be statistically significant (because of the larger samples size and power), whereas few associations were statistically significant in the objective weight gain group because of the smaller sample size.
      One limitation of our study includes the definition of perceived weight gain and loss. Because the survey question asked about weight change in intervals (since the participant's last survey) and not since the woman started her method, it was challenging to clearly define self-reported weight change over 12 months. Women who reported both weight gain and weight loss during the 12 months were excluded from the analysis, even though their net weight change may have been greater than 5 pounds in either direction. In addition, participants were asked whether they perceived a weight change of 5 pounds or more. Thus, a participant may have perceived a weight gain, but if it was less than 5 pounds, she reported no weight change. If the woman perceived a 3 or 4 pound weight gain at each of the surveys, she may have perceived an overall 9-12 pound weight gain over 12 months, yet her responses placed her in the perceived no-weight-change group.
      Another limitation of our study is that CHOICE is an observational study, not a randomized trial; thus, there is still the potential for residual bias despite our efforts to control for confounding variables. Finally, there is reason to believe that media may have influenced participants' reports and their associations of weight gain with certain contraceptive methods. For example, DMPA has been widely reported to be associated with weight gain and many women are aware of this association; therefore, women using DMPA may have been more likely to report weight gain.
      In conclusion, self-reported weight change is easy to obtain and in most women represents true weight gain. The perception of weight gain is clinically important because it may affect a woman's satisfaction with her contraceptive method or influence a woman's decision to continue the use of the method. Future studies should consider interventions that can promote healthy weight control, especially in women at high risk for weight gain, and should assess the relationship between perceived weight gain and contraceptive continuation and satisfaction.

      References

        • Bahamondes L.
        • Pinho F.
        • de Melo N.R.
        • Oliveira E.
        • Bahamondes M.V.
        [Associated factors with discontinuation use of combined oral contraceptives].
        Rev Bras Ginecol Obstet. 2011; 33: 303-309
        • Hani D.
        • Imthurn B.
        • Merki-Feld G.S.
        [Weight gain due to hormonal contraception: myth or truth?].
        Gynakol Geburtshilfliche Rundsch. 2009; 49: 87-93
        • Berenson A.B.
        • Rahman M.
        Changes in weight, total fat, percent body fat, and central-to-peripheral fat ratio associated with injectable and oral contraceptive use.
        Am J Obstet Gynecol. 2009; 200: 329.e1-329.e8
        • Bonny A.E.
        • Ziegler J.
        • Harvey R.
        • Debanne S.M.
        • Secic M.
        • Cromer B.A.
        Weight gain in obese and nonobese adolescent girls initiating depot medroxyprogesterone, oral contraceptive pills, or no hormonal contraceptive method.
        Arch Pediatr Adolesc Med. 2006; 160: 40-45
        • Pantoja M.
        • Medeiros T.
        • Baccarin M.C.
        • Morais S.S.
        • Bahamondes L.
        • Fernandes A.M.
        Variations in body mass index of users of depot-medroxyprogesterone acetate as a contraceptive.
        Contraception. 2010; 81: 107-111
        • Linde J.A.
        • Jeffery R.W.
        • French S.A.
        • Pronk N.P.
        • Boyle R.G.
        Self-weighing in weight gain prevention and weight loss trials.
        Ann Behav Med. 2005; 30: 210-216
        • Secura G.M.
        • Allsworth J.E.
        • Madden T.
        • Mullersman J.L.
        • Peipert J.F.
        The Contraceptive CHOICE Project: reducing barriers to long-acting reversible contraception.
        Am J Obstet Gynecol. 2010; 203: 115.e1-115.e7
        • Pronk N.P.
        • Crain A.L.
        • Vanwormer J.J.
        • Martinson B.C.
        • Boucher J.L.
        • Cosentino D.L.
        The use of telehealth technology in assessing the accuracy of self-reported weight and the impact of a daily immediate-feedback intervention among obese employees.
        Int J Telemed Appl. 2011; 2011 (909248)
        • Rahman M.
        • Berenson A.B.
        Self-perception of weight gain among multiethnic reproductive-age women.
        J Womens Health (Larchmt). 2012; 21: 340-346