AMCHP 2005 ANNUAL CONFERENCE
DELIVERING RESULTS, IMPROVING PREGNANCY & BIRTH
February 19-23, 2005
SAMANTHA GARBERS: Okay, good morning. As Danielle Mentioned, I’m a last-minute fill-in for my colleague, Terry Rosenberg, who’s fine, but whose knee is not working. So I’m muddling through her slides here, but I am familiar, she and I have worked on a paper related to looking at excess pre-pregnancy weight and excess weight gain, specifically here relating to preeclampsia and then pre-term birth. So I’d like to acknowledge, obviously, Terry, my co-author, and Mary Ann Shezan, who’s our Vice President for Research at MHRA. So I will talk a little bit about what my presentation’s going to be about. The study we have been working on looks at the vital statistics files from New York City births from 1999, 2000 and 2001. This actually ends up to be more than 330,000 births in New York City . The database contains a number of births. What we excluded was anything other than a singleton birth and we only included women who had pre-pregnancy weight data. There were about 9 percent of the births that were lacking pre-pregnancy weight and those were excluded from the analysis. In terms of the methods we used here, chi-square anova logistic regression, to look at, as I mentioned, preeclampsia and also pre-term birth is the outcomes.
Throughout the presentation we’ll be using a few definitions and instead of repeating them on each slide, I’ll just go over them. For this analysis we defined obesity as 200 pounds or more pre-pregnancy. One of the limitations of the New York City birth certificates files is that it does not contain height, so we were not able to compute the body mass index, so we did--looking at End Haynes Data, the average height of a woman is a little bit less than 5’ 4”, so using 200 pounds is a cut-off for a woman who is of average height, her BMI would be 34.3 or for 300 pounds, which is the second weight group we use for excess pre-pregnancy weight, that BMI would be 51.5, so we’re fairly confident that this definition is including a vast majority of, well, pretty much everybody who would be considered a BMI of 30 or over. For excess prenatal weight gain we used a very conservative definition, so the Institute of Medicine, the 1990 recommendations have different weight gain recommendations based on pre-pregnancy weight group.
No group including underweight women are recommended to gain more than 40 pounds, so we used a dichotomized variable, 41 pounds or over was considered excess weight gain, and that is for women of any group. Preeclampsia, this is, as it says here, it’s the physician’s diagnosis as noted on the birth certificate using a (inaudible) criteria. We did not do any sort of independent review for preeclampsia. And finally, we also used a very conservative definition for pre-term births. Sometimes people use 37 weeks, here we’re looking at pre-term birth before 34 weeks of gestation. Just to give you a sense of what the prevalence and instance of some of the variables we’re looking at here, we had 6.3 percent of the population 200 obese, 200 pounds or over, 18.4 percent had excess weight gain. We had preeclampsia of 2.1 percent and pre-term birth of 2.1 percent. I want to let you--it’s not the same women. It’s pretty unusual to get the exact same number, but that’s just--that is just a coincidence. We have a lower instance of preeclampsia than you would see in William’s, William’s Obstetrics.
Generally, most studies find about a 5 percent prevalence, but again, we’re looking only at singleton births, which is why this rate’s probably a little lower than what you might expect. One of the wonderful things about the New York City births file is the incredible diversity in terms of racial and ethnic diversity here. You can see on the slide we have about a third of the women were Hispanic, a little more than a quarter Black, 29.3 percent Non-Hispanic White and 11.7 percent Asian. Just to tell you a little bit, we have--I’ll show you on the next slide, a third of the mothers were foreign-born. This is something that we’ve been looking at in other analyses. For the Latina women, which is where we see a lot of diversity in terms of birthplace, the most common birthplaces were Dominican Republic , about 16,000 women, followed by Mexico , a little bit less than 10,000, and then Puerto Rico born, which is about 5,000 mothers in the database. So again, giving you a little bit more of a sense of the sample, we have about 9 percent were teen mothers, a nice chunk in the 20’s, also in the 30’s and almost 4 percent who were 40 or over; 44 percent were not married and let me see, one-quarter, three-quarters in terms of cutting it off less than a high school graduate or high school graduate or more. And as I mentioned in the previous slide, about a third of the mothers are foreign-born.
Also other factors that we’re going to be looking at when we do the logistic regression, 44 percent of the women, this was their first birth, doing pretty well, about 60 percent, a little bit more than 60 percent initiated prenatal care in the first trimester. We have, as you can see, a large proportion of mothers, just more than half, who had Medicaid coverage. This is because they don’t know about the states where you come from. New York State has something called PCAP, Prenatal Care Assistance Program. MHRA also does direct service provision at our MIC Centers. One of the things that we do is to work on getting women enrolled in PCAP and so this is probably a proportion higher than you would expect in other states. In terms of the prevalence of the medical risks we’re going to be looking at later, we have three-tenths present with chronic diabetes, that is diabetes that existed before pregnancy, and a little bit less than 1 percent with chronic high--high blood pressure.
I added in, at the recommendation of my colleague, a couple not on these slides. We had 6.5 percent who were low birth rate, which was under 2,500 grams, you know, and 1.2 percent who were very low birth rate, under 1,500, and we have three tenths of a percent with low five-minute Apgar scores, which is four or less. These are not outcomes that we’re looking at there, but it might me something just in terms of comparability that you’d be interested in. So now I want to talk a little bit about the weight characteristics here. We saw for almost, well, for all of the risks that we’re looking at in terms of the weight characteristics and the health risks, significant differences in terms of racial and ethnic subgroups of the women. You can see here, for the Non-Hispanic Black women, almost 12 percent were 200 pounds or more pre-pregnancy. For Hispanic women, it was about half that, very small proportion of the Asian women were 200 pounds or more pre-pregnancy and for the Non-Hispanic White women similar, a little bit less than five percent. Looking at the other weight variable that we’re examining today, excess prenatal weight gain, it is--because of the large sample size, it is statistically significant according to subgroup, but you can see it’s about 20 percent, 10 percent for Asian women and then 17 percent almost for the Non-Hispanic White women.
Looking at the medical risks that we’re including and actually preeclampsia, we’re also looking at it as an outcome. You can see that the Non-Hispanic Black women were also at--had a higher proportion with preeclampsia, almost three percent, and again, following the same pattern, Hispanic women--Asian women with low percentage and also Non-Hispanic White women. Just in terms of comparison looking at the MCHA data, in 2002 there were 20.2 percent of women who gain more than 40 pounds during their pregnancy, but that of course, as I mentioned before, included all births. Here we’re looking just at singleton births, where you would expect a less weight gain, would expect. Looking at our other outcome, preterm birth, we have, again, the same pattern here. The Non-Hispanic Black women had a much higher rate of preterm birth, followed by Hispanic women, Asian women and then actually the lowest rate among the Non-Hispanic White women.
So now I’ll get into this long, long list of variables that we included in the logistic regression, and this is a little goggling. What we included were any variables that were found in other studies to be associated either with preeclampsia or with preterm birth. We included both socio-demographic variables and some health variables. Age, especially for preeclampsia, very young women or older women are at higher risk for preeclampsia. Race and ethnicity, obviously given the slides we just saw, that the medical risks and the weight variables were not evenly distributed according to--among groups of different racial and ethnic background. Marital status and educational attainment, for preeclampsia, parity is a very important variable, because women who have not had a previous live birth are also at increased risk for preeclampsia. Timing of initiation of prenatal care, you’ll see later in the regression that this ends up not being as--falls out in an interesting way.
Insurance coverage, which is Medicaid or not Medicaid, then the last three, the little asterisks, it’s not my favorite way to do it, for the chronic and pregnancy-related diabetes and then chronic and pregnancy-related high blood pressure, those were included in the preeclampsia regressions. Obviously, we didn’t include preeclampsia as a predictive for preeclampsia. That was included only in the regression looking at preterm birth, so the first regression that I’m going to go through we included chronic medical conditions of hypertension and pregnancy and pregnancy-induced in predicting preeclampsia, then we’re predicting preterm birth, we’re looking--including only preeclampsia in the model. Looking at the weight variables--yeah?
UNIDENTIFIED SPEAKER: Did you include place of maternal birth as a (inaudible)?
SAMANTHA GARBERS: It was not included in this. We have done other analyses, looking at, you know, what they call the epidemiologic paradox. It was not included in this analysis.
UNIDENTIFIED SPEAKER: Is there a difference in whether or not the mothers were born out of country among the racial groups within New York City ?
SAMANTHA GARBERS: Yes, yes—
UNIDENTIFIED SPEAKER: There were some that—
SAMANTHA GARBERS: --the Hispanic mothers, you mean—
UNIDENTIFIED SPEAKER: Explainable by that—
SAMANTHA GARBERS: Yes, that’s something that we have looked at, actually pretty significantly and I can talk about it with you later. What we find is that the Puerto Rican-born women, looking at the country of birth for foreign-born women, that we do see big difference, and that relates to other health behaviors related to preterm birth. So you can come talk to me about that, that’s something we’ve also been working on. I’m way behind on my slides here and my notes. Looking here, for the pregnancy weight, the reference group we used was 100 to 149 pounds pre-pregnancy. This is the modal group, this is the--we had the most women in this weight group. For prenatal weight gain, as I mentioned before, we just dichotomized it, so for the reference group we used a very conservative estimate of institute medicine recommendations and had only excess weight gain, 41 pounds or more.
So here, after adjusting for everything, we still have a lot of significant variables, so we broke this into two slides, first looking at the confounders, um, not surprisingly you can see--I need a little red thing, about halfway down the slide, 2.3 women who had had no previously live birth, it increased risk for preeclampsia. Also not surprising, the bottom number, an adjusted odds ratio of 7.3 for women with chronic high blood pressure, that is high blood pressure before pregnancy. Looking at some of the other significant variables in the adjusted model, Black women, Hispanic women, women with less than a high school degree and older women also it increased risk for preeclampsia even when adjusting for other factors. Looking--this is the second part, which is looking just at the weight variables. This is all in the same model, but we had so many, we had to put it on a second slide.
As you can see, even when you’re taking into account existing medical risks, including diabetes and high blood pressure, women who had excess pre-pregnancy weight, either 200 pounds or more or an even greater effect for women who are 300 pounds or more, were significantly in the increased risk for preeclampsia. Also, even controlling for the pre-pregnancy weight and the medical conditions, women who gained more then the recommended amount of weight during their pregnancy were also at increased risk for preeclampsia. And you can see because it’s such a large sample that we’re working with that it’s pretty tight confidence intervals, but a substantial additional risk. We then did a--looked at a second outcome, which is preterm birth, and found some things the same, but also a number of things that are different.
The first three, again, the first three adjusted odds ratio women who are 35 years or more, Black women and Hispanic women at increased risk for preterm birth even when you’re adjusting for other factors. What we found were some other interesting things. Women who are not married were at increased risk for preterm birth. Again, an effect of education. I mentioned earlier, the prenatal care, we didn’t see any affect in terms of when women initiated prenatal care. What we did see was that women who had no prenatal care whatsoever had a significant increased risk for preterm birth, even when you adjust for other factors that might be associated with late entry into prenatal care. No insur--here we’re seeing an effect of no insurance, self pay, these are largely undocumented women in New York City . And most importantly, you can see a very significant risk for preeclampsia associated with preterm birth, an odds ratio of 6.2. And using the same format as before, now looking at the weight variables predicting preterm birth, we see a very small effect for women who are 200 to 299 pounds.
We don’t see a significant added risk when you adjust for all the other factors, for women who are 300 pounds or more. And maybe more surprisingly, there appears to be a protective effect for women who had excess weight gain during pregnancy. What is important to keep mind is that we adjusted for preeclampsia in this model, so if you had excess pre-pregnancy weight and excess pregnancy weight gain and didn’t get preeclampsia, there might be a protective effect that you would see, but this is, as we said, an adjusted model. One of the things you would expect to see is a macrosomic infant, a large for gestational age, which is moving in the opposite direction. What we really, in terms of recommendations or approaches, what to do with the data that we see here, one of the patterns that we noticed were women who were at a double risk almost.
As you can see here, this is looking at weight gain, excess weight gain, the prevalence of excess weight gain by pre-pregnancy weight. And you can see that women who are in the largest weight groups, 16, almost 17 percent and 16 percent still gained 41 pounds or more during pregnancy. This is certainly something in terms of an approach of what to do with this data, helping women reduce their risk either in pre-pregnancy weight or in terms of managing weight during pregnancy. Even more surprising, it’s pretty unusual to get a graph that’s this clearly spelled out and this is something that has been found in other analyses. There’s a study in Stockholm and then another one in Missouri . What we see here is a very clear incremental increase in weight with each successive pregnancy, so this is looking at the pre-pregnancy weight for women who had no prior birth, it’s 137 pounds.
On average for women who are delivering their second child, I’m doing the math here, that’s almost six pounds average weight gain. And you see for each pregnancy, at least in our New York City data, keeping on those extra pounds and that’s the mean pre-pregnancy weight, so for each of these women there are women who have retained less weight, but there are also many who have retained more weight. And this is something when we’re seeing, well, a mean weight of 152 pounds, getting very close to 200 pounds where you do see an added risk for both preeclampsia--well, for preeclampsia, but not for preterm birth. So for our conclusions, obesity and excess prenatal weight gain do increase the risk for preeclampsia for all women, even when you’re adjusting for other factors. The greatest increased risk for preeclampsia is among women who are weighing 300 pounds or more pre-pregnancy. In our data that includes 700 women weighing 300 pounds or more before they’re pregnant.
Even without women who have chronic diseases including diabetes or high blood pressure, obesity increased the risk of preeclampsia. And then preeclampsia, in term, as you mentioned, it was a 6.2 adjusted odds ratio, so preeclampsia increases greatly the risk of preterm birth. So it appears that the negative affects of obesity and excess weight gain are mediated through preeclampsia, so we’re not seeing a direct association with the preterm birth, but certainly they’re increasing the risks of preeclampsia. So as I mentioned in terms of recommendations, nutrition and exercise counseling are needed. Pre-pregnancy, if you can work with mothers to reduce their weight before they get pregnant.
We have, since we are, as an organization, involved in direct service provision, what we call “Teachable Moments”, especially if women have a short interval between pregnancies, that the post-partum period is an ideal time to talk with women about maintaining a healthy weight in‑between pregnancies and also, well, this is going down to the third bullet, working with women to help avoid the incremental pregnancy increase with each excessive pregnancy. And then also working, because we did see an affect for weight gain in addition to pre-pregnancy weight, working with women to have moderate appropriate weight gains during pregnancy. And this is Terry’s email, but you can reach both of us. I’m open for questions. There may be things that I can’t answer, since I’m a late fill-in, but certainly Terry and I can both work with you if you have questions. Thank you.