Physiologic and Supraphysiologic Increasesin Lipoprotein Lipids and Apoproteins

in Late Pregnancy and PostpartumPossible Markers for the Diagnosis of "Prelipemia"

Agustin Montes, Carolyn E. Walden, Robert H. Knopp, Marian Cheung,

Margaret B. Chapman, and John J. Albers

A supraphysiologic (>95th percentile) rise in plasma lipids in pregnancy may serveas a marker for "prelipemia" in the same way that gestational diabetes is a marker forprediabetes. To qualify as prelipemic, subjects with an abnormal lipid rise antepartummust return to normal postpartum but may have other identifying characteristics. Thispaper describes the antepartum-postpartum changes of lipoprotein lipids and apo-proteins at 34 to 38 weeks of gestation and 6 and 20 weeks postpartum in 23 subjectswith physiologic and six subjects with supraphysiologic plasma lipid increases dur-ing pregnancy. These results are compared to measurements in 23 nonpregnantcontrols matched for weight, age, and race. In subjects with a physiologic antepartumlipid rise, postpartum total triglyceride and very low density lipoprotein (VLDL) lipids(cholesterol and triglyceride) and apo B returned to baseline within 6 weeks. In con-trast, low density lipoprotein (LDL) showed a slow postpartum decline in lipids andapo B with elevations remaining at 20 weeks postpartum. High density lipoprotein(HDL) cholesterol concentrations, elevated in pregnancy, remained elevated at 6weeks postpartum, but fell to baseline by 20 weeks postpartum. HDL triglyceride andapo A-l concentrations, both elevated in pregnancy, returned to baseline by 6 weekspostpartum. A supraphysiologic triglyceride rise in pregnancy was associated with aslower return of total triglycerides and VLDL to baseline, reduced HDL cholesterolante- and postpartum, atypical changes in LDL cholesterol during pregnancy andpostpartum, and evidence of hyperlipidemia among family members. Two subjectswith hypercholesterolemia in the nonpregnant state showed no marked exaggerationof total or LDL cholesterol concentrations in pregnancy. The data support the hypoth-esis that a supraphysiologic rise in plasma triglyceride concentrations in late preg-nancy may serve as a marker of prelipemia. Proof of the hypothesis requires furtherinvestigation and longer follow-up.(Arteriosclerosis 4:407-417, July/August 1984)

Glucose intolerance and hyperlipidemia are phys-iologic accompaniments of normal pregnan-

cy.12 With respect to glucose homeostasis, it is wellknown that some women manifest overt diabetesunder the stress of pregnancy but revert to normalafter delivery. The deterioration of glucose homeo-

From the Northwest Lipid Research Clinic and the Departmentof Medicine, School of Medicine, University of Washington, andthe Department of Biostatistics, School of Public Health and Com-munity Medicine, Seattle, Washington.

This work was supported by the Lipid Research Clinic ProgramContract HL 12157 and Grant HD 08960 and by the Council forTobacco Research, USA. Dr. Albers is an Established Investiga-tor of the American Heart Association.

Current address for Dr. Montes: Catedra 2, de Fisiologia, Facul-tad de Medicina, Cuidad Universitaria, Madrid 3, Spain.

Address for reprints: Robert H. Knopp, M.D., 326 Ninth Avenue,Seattle, Washington 98104.

Received August 17, 1983; revision accepted March 14, 1984.

stasis in pregnancy, known as gestational diabe-tes , 1 3 4 is considered to be a form of prediabetesbecause it predicts later progression to overt diabe-tes with a high degree of accuracy.4 It is not yetknown if a similar deterioration of lipid homeostasisoccurs in pregnancy, reverts to normal postpartum,and predicts the appearance of overt nongestationalhyperlipidemia years later. Such a condition might betermed "prelipemia" and if identified would providean important tool to study the natural history of hy-perlipidemia from its earliest stages.

This paper presents information to support thehypothesis that prelipemia is identified by a supra-physiologic lipid rise in pregnancy. The postpartumreturn of plasma lipids and lipoproteins to baseline isdescribed in normal subjects. These results are com-pared to women with supraphysiologic elevations intriglyceride or cholesterol in pregnancy but with de-

407

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

408 ARTERIOSCLEROSIS VOL 4, No 4, JULY/AUGUST 1984

monstrable normolipidemia before or after pregnan-cy. Relatives of these subjects were examined forevidence of familial hyperlipidemia. The distinctionbetween physiologic and supraphysiologic antepar-tum increases in lipoprotein lipids in pregnancy wasmade on the basis of the 95th percentile outpoints forlipoprotein triglyceride and cholesterol concentra-tions in late gestation that we have recently de-scribed.5

MethodsSubjects

Twenty-three pregnant women with a physiologicplasma lipid increase in pregnancy were recruited forthis study from the classes of the Childbirth Educa-tion Association (CEA) in Seattle and King County.Six other pregnant women with supraphysiologic lip-id increases in pregnancy were referred or foundthrough surveys. Two of these subjects, one hyper-triglyceridemic (Subject DR), and one hypercholes-terolemic (Subject AM), were identified in the originalCEA survey. All subjects gave informed written con-sent. A physiologic increase in plasma lipids in preg-nancy was defined as a total plasma triglyceride andcholesterol concentration less than the 95th percen-tile at 36 weeks of gestation (387 mg/dl for triglycer-ide and 318 mg/dl for cholesterol).5 A supraphysio-logic increase in plasma lipids in pregnancy wasdefined as a total plasma triglyceride or cholesterolconcentration at or above the 95th percentile at 36weeks of gestation. The pregnant subjects werecompared with 23 nonpregnant subjects matched forage and Quetelet Index (body mass index calculatedas weight/height2 x 1000) with the 20-week postpar-tum body mass index of the study subjects used for

matching to controls. Control subjects were drawnfrom a random sample of female Pacific NorthwestBell Company employees whose characteristicswere described previously.8"8 The nonpregnant con-trols and 20-week postpartum subjects having ante-partum physiological increases in blood lipids wereclosely matched for age, weight, height, and bodymass index (Table 1). Women with supraphysiologiclipid increases were generally heavier and had ahigher Quetelet Index.

Length of gestation was estimated from the firstday of the last menstrual period. Pregnant subjectswere studied within 2 weeks of the antepartum targetweek and within 1 week of the postpartum targetweek. All subjects were free-living and consumingtheir customary diets; there was no systematic at-tempt to restrict calories in these subjects. All fastingglucose concentrations were normal in the pregnantsubjects (mean 72.6 ± 1.6 mg/dl, range 60-87) andin nonpregnant subjects (91.5 ± 7.6 mg/dl, range79-108). No medications except multivitamins weretaken during pregnancy. All newborn infants werehealthy and of normal weight (mean 3581 g ± 87.6,range 2727-3949 g), except one who was anence-phalic. The lipid and apoprotein measurements inthe mother of this infant were typical. Of the womenwith physiological hyperlipidemia, 19 women breast-fed postpartum and four did not. None took oral con-traceptives during the postpartum study period. Oneother woman (Subject EW) was studied before, aswell as during and after, pregnancy. Her results areconsidered separately. She was normoglycemic anddelivered a normal infant weighing 3580 g. The lacta-tional status of the women with supraphysiologic lipidincreases are noted in the tables or text. None usedoral contraception postpartum.

Table 1. Characteristics of Pregnant Subjects with Physiologic or Supraphysiologic LipidIncreases and Nonpregnant Normal Controls

Physiologic increaseAntepartum week 34-38Postpartum week 6Postpartum week 20

Nonpregnant controls

Supraphysiologic increaseTriglyceride

D.R.F.S.Y.F.C.B.

CholesterolM.R.N.C.

No.

23——

23

Age(yrs)

28.0 ±4.2——

27.8±4.2

26322727

3430

Weight(kg)

70.6±7.162.2±7.760.6±8.7

60.3 ±8.1

64.5*84.580.469.6

64.555.2

Height(cm)

167.6 ±6.4——

166.5±6.0

162.6172.7162.6165.0

154.9165.1

Body massindex

(kg/cm2 x1000)

2.52 ±0.302.22 + 0.222.16±0.29

2.17±0.23

0.244*0.2830.3040.256

0.2690.203

Values are means ± SD.*Prepregnant body weight.

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

PREGNANCY HYPERLIPIDEMIA AND PRELIPEMIA Montes et al. 409

Blood Sampling

Blood samples in pregnant and postpartum wom-en were obtained by a research nurse in each sub-ject's home to minimize the effects of stress on fueland lipoprotein homeostasis. Nonpregnant subjectscame to the Northwest Lipid Research Clinic forblood drawing. All subjects had fasted for 12 to 14hours before the sampling. About 35 ml of venousblood were obtained from subjects who were sitting.The blood for lipid analysis contained 1.5 mg EDTAper ml blood and was kept on ice until centrifuged inthe laboratory. The blood for glucose estimation wascollected in tubes containing 2.5 mg sodium fluorideand 2.0 mg potassium oxalate per ml blood.

Lipoprotein and Apoproteln Analyses

Plasma samples were overlain with 0.9% salineand ultracentrifuged in a 40.3 rotor at 40,000 rpm at17° C for 18 hours.9 The top 2 ml were analyzed asvery low density lipoprotein (VLDL). The bottom frac-tion was brought to a measured 5 ml with saline andanalyzed as the d > 1.006 fraction. Another 3 ml ofwhole plasma was treated with heparin and MnCLgto precipitate VLDL and low density lipoprotein(LDL), allowing measurement of high density lipo-protein (HDL) lipids in the supernatant.9'10 The apo Bcontent of the supernatant fraction never exceeded 2mg/dl by immunoassay. LDL lipid concentrationswere determined as the difference between the d <1.006 infranatant and HDL. Triglyceride and cho-lesterol were measured using the AutoAnalyzer IImethod.9

Apolipoprotein B (apo B) was measured in totalplasma and VLDL by a double antibody radioimmu-

noassay as previously described.11 LDL apo B wascalculated as the difference between total plasmaand VLDL apo B. Apolipoprotein A-l was measuredon whole plasma by a radial immunodiffusion meth-od after tetramethylurea delipidation.12

The ratios of triglyceride/cholesterol and lipid/apo-protein in each lipoprotein fraction are presented asindices of lipoprotein composition.

Statistical Analyses

The mean, standard deviation, and median de-scribed lipoprotein lipids, apoproteins and ratios forthe nonpregnant controls and the antepartum andpostpartum subjects. Nonparametric statisticalmethods were used for significance testing becauseof the small sample sizes. Wilcoxon's matched pairstest13 was performed to determine if significant differ-ences existed between antepartum and postpartummeasurements and between 6- and 20-week post-partum values. The Wilcoxon rank sum test13 wasused to detect differences between controls and an-tepartum, 6-week, and 20-week postpartum sub-jects. Mean values and 95% confidence intervalswere used to graphically illustrate antepartum-to-postpartum changes.

ResultsPhysiologic Hyperllpldemla In Pregnancy

Table 2 presents the antepartum-postpartum lipo-protein observations in subjects with physiologichyperlipidemia in pregnancy and matched nonpreg-nant controls. Selected lipoprotein lipid and apopro-tein changes are illustrated in Figure 1. Ratios of

Table 2. Plasma Lipoprotein Lipid and Apoproteln Concentrations (mg/dl) in 23 Women In LatePregnancy and Postpartum and in 23 Age- and Body-Mass-Matched Nonpregnant Controls

Pregnant

Weeks 34-38

Mean±SD Median Mean

Week

± S D

Postpartum

6 Week

Median Mean ± so

20

Median

Nonpregnant

Mean±sD

controls

Median

TotalTriglycerideCholesterol

222 ±60251 ±32

223*244*

71 ±23205 + 23

66t204*t

66±18190±28

62t187*t*

VLDLTriglyceride 107 ±41 100*Cholesterol 22 ± 9 20*ApoB 20±11 23*

LDLTriglycerideCholesterolApoB

29±18 23t7±4 6*f4±3 4t

HDLTriglycerideCholesterolApoA-1

72 ±21161 ±3984±23

29 ±964 ± 9

164+16

70151*81*

26*65*

167*

27 + 10124±21104±17

8±364±12

127±22

26*t127*t103*f

9*t62*

130t

28±16 21t6±5 5't4±2 4t

23±7 21*t120 ±24 120't88 ±16 90*tt

8±456±11

132±16

10t

131f

59±19171+26

33±1411+67 + 6

14±10104±2361 ±10

12±656±12

128±23

58165

31106

1310362

1153

124

'Significantly different from controls at p s 0.05.tSignificantly different from antepartum at p < 0.05.^Significantly different from 6 week postpartum at p < 0.05.

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

410 ARTERIOSCLEROSIS VOL 4, No 4, JULY/AUGUST 1984

5 0 -

Mean ; gSf. CI for lipid" Mean ; 95% CI tor

apo proteinMean t 95*/. C l for

nonpregnant controls

Triglycerlde

ApoB

LDL:150-

2* 100 H

N

\- V ApoB- • ?

HDL:150-

100-

5 0 -

Weeks. 36<-Antepartum

Apo A i

T

Cholesterol

6 20Postpartum

Figure 1. Mean ± SE of VLDL, LDL, and HDL lipids andapoproteins in late gestation and at 6 and 20 weeks post-partum. VLDL triglyceride and apo B reached a baselinelevel by 6 weeks postpartum. LDL cholesterol and apo Bdeclined slowly postpartum but were still elevated at 20weeks postpartum. If the rate of decline is extrapolated tothe mean ± SE nonpregnant control level (hatched area),LDL cholesterol would reach baseline by 75 weeks, apo Bby 42 weeks, and triglyceride (data not shown) by 50weeks. HDL apo A-l reached nonpregnant control levelsby 6 weeks while cholesterol reached control levels at 20weeks.

lipoprotein lipids and apoproteins are presented inTable 3. Total triglyceride concentrations in late ges-tation were significantly elevated (3.5-fold) abovenonpregnant control values, but were only 8 and 4mg/dl higher at 6 and 20 weeks postpartum, respec-tively, a nonsignificant elevation. In contrast, wholeplasma cholesterol concentrations in pregnancywere 50% higher than control values and remained25% and 13% higher at 6 and 20 weeks postpartum,respectively.

VLDL triglyceride, cholesterol, and apo B concen-trations in pregnancy were all significantly higherthan postpartum and control observations (Table 2).Postpartum concentrations were slightly below con-trol levels, significantly for cholesterol. The relation-ships of VLDL constituents expressed as ratios werenot significantly different comparing antepartum andpostpartum observations (Table 3). However, thetriglyceride/cholesterol ratio was higher at all ofthese times compared to controls.

In LDL in pregnancy, triglyceride and cholesterolwere 438% and 47% above control levels, respec-tively. Postpartum LDL triglyceride and cholesterolconcentrations declined gradually, but remained sig-nificantly above control levels at 20 weeks, by 61.5%and 16.5%, respectively. LDL apo B was significantlyelevated in pregnancy, 38% above controls, and waseven higher postpartum by 70% and 44% abovecontrols at 6 and 20 weeks, respectively. Triglycerideenrichment of LDL relative to cholesterol and apo Bwas particularly marked in pregnancy and declinedpostpartum (Table 3). The triglyceride/cholesterolratio remained significantly greater than controls at20 weeks postpartum while the triglyceride/apo Bratio was similar to controls.

In HDL, triglyceride was increased 2.4-fold in preg-nancy and fell to a level slightly below controls at 6and 20 weeks postpartum. Cholesterol was 23%above the control level in pregnancy, 17% abovecontrols at 6 weeks postpartum, and equal to thecontrol level at 20 weeks postpartum. Apo A-l con-centrations were elevated significantly in pregnancy,35% above control, but were at the same levels ascontrols at 6 and 20 weeks postpartum. As a result ofthese changes, HDL was triglyceride-rich and cho-lesterol-poor in pregnancy (Table 3), cholesterol-richat 6 weeks postpartum, and triglyceride-poor at 20weeks postpartum.

Observations were made on total plasma triglycer-ide, VLDL, LDL, and HDL cholesterol and glucosebefore pregnancy as well as during pregnancy andpostpartum in Subject EW (Figure 2). EW was takingOrthonovum oral contraception until 18 days beforethe pregnancy measurement, which was 1 monthbefore conception. A gradual fall in plasma glucosecontrasted with a rise in all lipid constituents. Post-partum triglyceride fell to below prepregnancy levelsby 6 weeks. HDL cholesterol reached the prepreg-nancy control level by 20 weeks, while LDL choles-terol at 6 weeks (139 mg/dl) and 20 weeks (115 mg/dl) was still above the prepregnancy level (99 mg/dl).

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

PREGNANCY HYPERLIPIDEMIA AND PRELIPEMIA Montes et al. 411

Table 3. LIpoprotein Lipid and Apoproteln Ratios In 23 Women in Late Pregnancy and Postpartum and In 23Age- and Body-Mass-Matched Nonpregnant Controls

Pregnant

Weeks 34-38

Mean±SD Median Mean

Week

± S D

Postpartum

6 Week

Median Mean±sD

20

Median

Nonpregnant

Mean±sD

controls

Median

VLDLTG/cholTG/apo BChol/apo B

LDLTG/cholTG/apo BChol/apo B

HDLTG/cholTG/apo A-1Chol/apo A-1

5.10±1.10 5.00*7.79 ±5.07 5.211.57 ±0.98 0.99

4.88 ±2.40 4.00*9.30±9.08 7.002.31+3.07 1.60

0.46±0.100.91 ±0.302.02 + 0.65

0.44*0.87*1.93

0.46 ±0.15 0.50*0.18±0.05 0.16*0.39 ±0.07 0.40*

0.23 ±0.080.26 + 0.081.21 ±0.25

0.14±0.060.07 ±0.030.51 ±0.11

0.22*t0.25f1.14*t

0.13*t0.07t0.48*t

6.14±3.767.56 ±4.011.50 ±1.02

0.20 ±0.070.26 ±0.081.38 ±0.29

0.15 ±0.080.07 + 0.030.42 ±0.08

5.67*6.181.15

0.20*t0.25f1.43*t

0.16*t0.07*t

3.69 ±3.4810.88± 14.863.06 ±3.32

0.14 + 0.080.23±0.171.72 ±0.28

0.21 ±0.090.09 ±0.040.44 + 0.06

2.994.111.50

0.120.221.74

0.200.090.44

'Significantly different from controls at p s 0.05.tSignificantly different from antepartum at p == 0.05.^Significantly different from 6 weeks postpartum at p ^ 0.05.

80-

60-

I6O-

80-

mg/dl

240-

I6O

80-

O-1

Glucose

Triglycerkje

Cholesterol -VLDL

LDL

8 16 24Pregnancy

32 40

Weeks

6 20Posfpartum

Figure 2. Plasma glucose, triglyceride, and cholesterol and lipoprotein cholesterolconcentrations in subject EW before, during, and after pregnancy. Glucose declinedwhile all lipids increased throughout pregnancy. Postpartum, triglyceride and VLDLcholesterol fell to slightly below prepregnancy levels by 6 weeks of gestation and HDLcholesterol fell to normal by 20 weeks. LDL cholesterol declined more slowly andremained above prepregnancy levels at 20 weeks postpartum.

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

412 ARTERIOSCLEROSIS VOL 4, No 4, JULY/AUGUST 1984

Table 4. Hypertrlglycerldemla In Pregnancy: Subject D.R.

Whole plasma (mg/dl)TriglycerideCholesterol

VLDL (mg/dl)TriglycerideCholesterolApo B

LDL (mg/dl)TriglycerideCholesterolApo B

HDL (mg/dl)TriglycerideCholesterolApo A-1

Pregnant

Week 36

600329

3827944

164240181

1014

158

(387)*(318)

(246)(47)

(113)(218)121

(21)(42)95

Week 6

211169

13123

9

3411499

71295

(157)*(265)

(122)(29)

(36)(177)

(4)(44)

Postpartum

Week 20

102163

69157

24109

636

116

(127)*(225)

(83)(28)

(35)(157)

(5)(35)

Week 40

115154

7117

2597

1540

(127)*(225)

(83)(28)

(35)(157)

(5)(35)

D.R.'s infant's birth weight was 3239 g. D.R. elected not to breast feed postpartum or use oralcontraceptives. See text for family study lipids.

'Denotes 95th percentiles for total, VLDL, and LDL lipids and 5th percentile for HDL lipids. Values for36 weeks of gestation and 6 weeks postpartum are based on a study of 553 pregnant women (seereference 5); values are from reference 6 for 20- to 29-year-old nonpregnant, nonhormone-taking sub-jects (Weeks 20 and 40).

Table 5. Summary of Endogenous Hypertrlglycerldemic Subjects Studied In Pregnancy

Total triglyceride (mg/dl)(95th percentile)D.R.F.S.Y.F.C.B.

Total cholesterol (mg/dl)(95th percentile)D.R.F.S.Y.F.C.B.

LDL cholesterol (mg/dl)(95th percentile)D.R.F.S.Y.F.C.B.

HDL cholesterol (mg/dl)(95th percentile)D.R.F.S.Y.F.C.B.

Pre-pregnant

(127)*———

116

(225)———

223

(157)——

160

(35)

——40

Pregnant

Week 36

(387)t600434716

1195

(318)329184257274

(218)240

706061

(42)14404338

Week 6

(157)t211179221392

(265)169212267260

(177)114152187163

(44)12343636

Postpartum

20

(127)*102124151—

(225)163169205

—

(157)109124124—

(35)362932—

40

(127)*115——

570

(225)154——

232

(157)97———

(35)40——34

100

(127)*———

173

(225)———

213

(157)———

139

(35)———39

*Percentiles from a population of nonpregnant subjects (see reference 6).tPercentiles from a population study of pregnant and postpartum subjects (see reference 5).

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

PREGNANCY HYPERLIPIDEMIA AND PRELIPEMIA Montes et al. 413

£• £• 6

/ /

20 40 60

Weeks Postpartum

100

Figure 3. Plasma total triglyceride and LDL and HDL cholesterol concentrationsantepartum, at 36 weeks of gestation, and at various times postpartum in four hyper-triglyceridemic subjects in pregnancy compared with the 95th or 5th percentiles of arandom population of pregnant women or nonpregnant women aged 20 to 29 years.Total triglyceride for subject YF (aged 33) is above the 95th percentile for women aged20-29 at 20 weeks postpartum, but below the 95th percentile for women aged 30-39.

Supraphysiologic Hyperlpidemia in Pregnancy

The results are considered by the individual sub-ject in the text and are summarized in Tables 4 to 6and Figure 3. Subject DR was studied in detail (Table4 and Figure 3). She was markedly hypertriglyceride-mic and slightly hypercholesterolemic in late gesta-tion as compared to 95th percentile reference val-ues.5 The VLDL triglyceride level of 382 mg/dl waswell above the 95th percentile in pregnancy (246 mg/dl). The LDL triglyceride and cholesterol concentra-tions both exceeded the 95th percentile at 36 weeksof gestation, but the relative amounts of her triglycer-ide, cholesterol, and apo B in LDL remained approxi-mately the same as in the 23 normal subjects (Table2). Her HDL triglyceride and cholesterol concentra-tions were markedly reduced in pregnancy, while theapo A-l concentrations remained at a level typical forpregnancy (158 mg/dl).

At 6 weeks postpartum, DR's plasma triglycerideconcentrations remained above the 95th percentile,but at 20 and 40 weeks postpartum they fell withinthe 95th percentile of triglyceride for nonpregnant,nonhormone-using subjects.6 A similar slow postpar-tum decline was seen for VLDL triglyceride. LDL tri-

glycerides had dropped more abruptly by 6 weekspostpartum, while HDL triglyceride declined slowly.LDL cholesterol showed a continuing declinethrough 40 weeks postpartum but was within the95th percentile by 6 weeks postpartum. HDL choles-terol was still low at 6 weeks postpartum, rose by 20weeks postpartum, but even at 40 weeks postpartumwas only marginally above the 5th percentile for HDLcholesterol in nonpregnant women aged 20-29. Incontrast, apo A-l concentrations at 6 and 20 weekspostpartum were only slightly below the mean fornonpregnant, nonhormone-treated normal subjects(125 mg/dl).12

Two years later, subject DR was observed in asecond pregnancy at 36 weeks of gestation. Totalplasma triglyceride was 400 mg/dl (>95th percen-tile), LDL cholesterol was 167 mg/dl (<95th percen-tile), and HDL cholesterol was 48 mg/dl (<10th per-centile).

To investigate the possibility that DR's pathophys-iological response to pregnancy is due to a familialtrait, lipoprotein lipids were determined in her father(aged 61) and mother (aged 62). Results are pre-sented in mg/dl. The father had a total triglyceride of

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

414 ARTERIOSCLEROSIS VOL 4, No 4, JULY/AUGUST 1984

Table 6. Hypercholesterolemla In Pregnancy:Subject A.M.

Pregnant

Week 36

Postpartum

Week 6 Week 20

Whole plasmaTriglycerideCholesterol

VLDL (mg/dl)TriglycerideCholesterolApo B

LDL (mg/dl)TriglycerideCholesterolApo B

HDL (mg/dl)TriglycerideCholesterolApo A-1

(mg/dl)226 (387)326 (318)

87 (246)15 (47)20

98 (113)230 (218)116

23 (21)60 (42)

164

118 (157)250 (265)

40 (122)9 (29)

11

50 (36)167 (177)184

13 (4)59 (44)

150

105 (160)254 (240)

20 (112)4 (28)4

68 (28)181 (159)

103

11 (4)52 (33)

150

A.M.'s infant's birth weight was 2789 g. A.M. breast fedfor 8 weeks postpartum.

Parentheses denote 95th percentiles for total, VLDL,and LDL lipids and 5th percentile for HDL lipids. Values for36 weeks of gestation and 6 weeks postpartum are basedon a study of 553 pregnant women (see reference 5);values for age 30-39 nonpregnant, nonhormone-takingsubjects (Week 20) are from reference 6.

242 (>90th age-specific percentile), a total-choles-terol of 257 (>75th percentile), an LDL cholesterol of181 (~90th percenile), and an HDL cholesterol of 37(~15th percentile). The mother's lipoproteins wereclose to the median values for her age and sex: totaltriglyceride, 149; total cholesterol, 227; LDL choles-terol, 175; while her HDL cholesterol, 72 mg, wasnear the 75th percentile. Thus, there is evidence forelevated total triglyceride and LDL cholesterol in afirst degree relative of DR.

Three other subjects with supraphysiologic triglyc-eride increases in pregnancy are compared to DR inTable 5 and are also illustrated in Figure 3. One ofthese women, subject CB, was the mother of twohypertriglyceridemic children with fasting chylomi-cronemia and plasma triglyceride levels of 6460 and796 mg/dl at ages 3 and 1, respectively, on an unre-stricted diet. CB's parents had triglycerides at the75th and 90th percentiles and cholesterol at the 75thpercentile.68 CB therefore appears to be an obligateheterozygote for the Type I phenotype. YF's brother(aged 29) had total and LDL cholesterol concentra-tions (237 and 163 mg/dl, respectively) at or abovethe age- and sex-specific 95th percentiles (222 and165 mg/dl) and a total triglyceride concentration (135mg/dl) >90th percentile. YF's father had averagelipoprotein lipid concentrations. Thus, there is an as-sociation of hyperlipidemia in a first degree relativeof YF. The relatives of FS were not available forstudy.

All four subjects were below the 95th percentilevalues for total plasma triglycerides either before

pregnancy or 20 weeks postpartum, but were mark-edly above the 95th percentile in late gestation. Allsubjects showed a delay in the postpartum decline oftotal triglyceride concentrations compared with nor-mal subjects, in that at 6 weeks postpartum, totaltriglyceride concentrations were consistently abovethe 95th percentile (Table 5 and Figure 3). Similartrends were observed in VLDL triglyceride and cho-lesterol (data not shown). Disparate responses wereseen among the four subjects with respect to totaland LDL cholesterol ante- and postpartum. DR wasthe only subject showing a combined elevation oftriglyceride and cholesterol in pregnancy. All othershad abnormal increases only in total and VLDL trigly-ceride. Another distinction was the LDL cholesterolfall at 6 weeks postpartum in DR, but a rise in theother three. The 20-week postpartum or prepreg-nancy LDL cholesterol concentrations in these threesubjects were also higher than the pregnancy levels.HDL cholesterol concentrations were not as low insubjects FS, YF, and CB as in subject DR, but werenear or below the 5th percentiles both ante- andpostpartum.

Hypercholesterolemic subject AM is presented inTable 6 and in Figure 3. She was not on a cholester-ol-restricted diet at the time of the study. Total plas-ma cholesterol concentrations were similar at 6 and20 weeks postpartum, but were near or above the95th percentile value. VLDL lipids and apo B wereconsistently normal (compare the apo B results withthose in Table 2). The LDL cholesterol concentrationwas above the 95th percentile at 36 weeks of gesta-tion,5 fell postpartum to slightly below the 95th per-centile at 6 weeks, but exceeded the 95th percentileat 20 weeks.6 Compared with 6 and 20 weeks post-partum, the LDL cholesterol rise due to pregnancywas 49 to 63 mg/dl. The apo B concentration wasabout 40 mg/dl above the average at 36 weeks ofgestation and 20 weeks postpartum, but increased to80 mg/dl above the average at 6 weeks postpartum,an exaggeration of the pattern observed in normalsubjects at 6 weeks postpartum (Table 2). HDL lipidswere fairly typical in this patient; however, apo A-l didnot fall postpartum as much as expected.

A familial heterozygous hypercholesterolemicwoman with tendonous xanthomas (subject NC) wasstudied at 37 weeks of gestation and 6 and 20 weekspostpartum. The patient was on a cholesterol-re-stricted and polyunsaturate-supplemented dietthroughout. Total cholesterol was 614 mg/dl and to-tal triglyceride, 232 mg/dl at 37 weeks of gestation.Postpartum, 6- and 20-week total cholesterol con-centrations were 554 mg/dl and 530 mg/dl with cho-lesterol concentrations in VLDL of 20 mg/dl and 24mg/dl; LDL, of 484 mg/dl and 444 mg/dl; and HDL, of60 mg/dl and 62 mg/dl. In NC the total cholesterolrise in pregnancy (comparing 36 weeks of gestationand 20 weeks postpartum) was 84 mg/dl, only slight-ly higher than the total cholesterol increment ob-served in 23 normal women over the same interval(61 mg/dl). The conclusion from hypecholesterole-

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

PREGNANCY HYPERLIPIDEMIA AND PRELIPEMIA Montes et al. 415

mic subjects AM and NC is that plasma cholesterol inhypercholesterolemic subjects is not disproportion-ately exaggerated in pregnancy in contrast to plasmatriglyceride in hypertriglyceridemics.

Discussion

Women with physiologic or supraphysiologic in-creases in plasma triglyceride or cholesterol in preg-nancy were studied at 34 to 38 weeks of gestationand at 6 and 20 weeks or longer postpartum to an-swer the following questions: How rapidly do thechanges in plasma lipoproteins return to normalpostpartum in subjects who have a physiologic lipidincrease antepartum? Do subjects with supraphysio-logic increases antepartum return to normal postpar-tum in the same manner and do they return to normalat all? Are there identifying characteristics in lipopro-tein concentrations or family history that would sug-gest that the supraphysiologic lipid rise in pregnancyis a pathophysiologic process indicative of a subclini-cal or prelipemic trait that may eventually progress toovert hyperlipidemia?

In subjects with a physiologic triglyceride rise inpregnancy, plasma total triglyceride and all VLDLconstituents reached a baseline level by 6 weekspostpartum essentially identical to nonpregnantcon In subjects with a physiologic triglyceride risein pregnancy, plasma total triglyceride and all VLDLconstituents reached a baseline level by 6 weekspostpartum essentially identical to nonpregnant con-trol subjects. All subjects with a supraphysiologicaltriglyceride rise in pregnancy had normal total plas-ma triglyceride concentrations (<95th percentile) ei-ther before pregnancy or by 20 weeks postpartum,but the rate of return to normal postpartum wasslower than in normal subjects with plasma triglycer-ide still elevated at 6 weeks postpartum. It is notknown if the excessive triglyceride rise antepartum(and to a lesser degree 6 weeks postpartum) is dueto an increased rate of VLDL entry, diminished clear-ance, or both, and if the mechanism is the same foreach individual. Evidence from animal models doesindicate that the physiologic hypertriglyceridemia ofpregnancy is due to increased VLDL entry into thecirculation rather than to a primary impairment ofVLDL triglyceride clearance,14"17 with the increase inVLDL secretion due to an estrogen effect.18 Interest-ingly, pregnancy does not appear to cause a dis-proportionate cholesterol increase in total plasmaor LDL in subjects with hypercholesterolemia inthe nonpregnant state, which indicates that preg-nancy may not be useful in identifying latent hyper-cholesterolemia.

A slow postpartum return to the nonpregnantbaseline was observed in the constituents of LDLeven in subjects with physiologic hyperlipidemia inpregnancy. In subjects with a physiologic lipid riseantepartum, the postpartum return to baseline wasprolonged beyond 20 weeks postpartum and may beas long as 1 to 1.5 years if one assumes a linear rate

of decline extrapolated from 6 and 20 weeks (seeFigure 2). Changes in LDL composition also persist-ed to at least 20 weeks postpartum, such as an in-crease in the LDL triglyceride/cholesterol ratio whichin this instance mimicked the preceding change inpregnancy. Another interesting feature is that LDLhad relatively more lipid and less apo B antepartumand less lipid and more apo B postpartum. There isas yet no information on the mechanism of the ante-partum and postpartum LDL elevations and changesin composition. It is tempting to speculate that thepersistent postpartum changes in LDL may reflectthe regression of maternal fat stores and other tis-sues with mobilization of cholesterol and an en-hanced recycling of esterified cholesterol into LDL.

A contrasting pattern is seen in LDL constituentsof subjects with excessive triglyceride increases inpregnancy. In three subjects, LDL cholesterol wasreduced in pregnancy, was markedly increased 6weeks postpartum, and then gradually fell. The re-ciprocal relationship observed between VLDL andLDL is reminiscent of that seen originally by Leesand Wilson19 and may be a feature of familial com-bined hyperlipidemia,20 a highly atherogenic condi-tion.21 On the other hand, subject DR possibly repre-sents another form of hypetriglyceridemia, with thesimultaneous supraphysiologic rise in LDL andVLDL in pregnancy also suggesting a combined dis-order of triglyceride and cholesterol metabolism. Theimportant point is that differing patterns of lipoproteinabnormalities appear to exist among pregnant wom-en with supraphysiologic triglyceride increases inpregnancy and these may reflect differing patho-physiologic and genetic mechanisms. Pure hyper-cholesterolemia as exemplified by subjects AM andNC contrasted with hypertriglyceridemia in that totaland LDL cholesterol were not unduly exaggeratedduring pregnancy.

In subjects with a physiologic lipid rise antepartum,HDL cholesterol and apo A-l both increased in preg-nancy. Postpartum, apo A-l returned to baseline first(within 6 weeks), followed by cholesterol (within 20weeks). A consistent marker in the subjects with asupraphysiologic triglyceride rise antepartum wasthe low HDL cholesterol antepartum which did notincrease postpartum even when triglycerides had re-turned to within the 95th percentile. This observationsuggests that the reduction in HDL may be the morefundamental abnormality and is a clue that the hy-pertriglyceridemia of pregnancy is a true pathophy-siologic disorder consistent with an underlying preli-pemic trait. It is noteworthy that the HDL cholesterolreduction was less marked in the three subjects withchanges in LDL reciprocal to VLDL than in the onlyhypertriglyceridemic subject, DR, in whom bothVLDL triglyceride and LDL cholesterol increased toan abnormal degree in pregnancy. An HDL choles-terol reduction intermediate between normal and en-dogenous hypertriglyceridemia is characteristic offamilial combined hyperiipidemia20 and combinedhyperiipidemias in general.22-23

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

416 ARTERIOSCLEROSIS VOL 4, No 4, JULY/AUGUST 1984

A final argument that a supraphysiologic rise inplasma triglyceride antepartum is a pathophysiologicprocess signifying an underlying prelipemic trait isbased on the fact that all three subjects for whomrelatives were available show evidence of hyperlip-idemia among family members. The differenceamong the subjects' responses to pregnancy, as wellas the available information among relatives, sug-gests that the designation of prelipemia could repre-sent several distinct and possibly genetic underlyingdisorders of lipoprotein metabolism. Further investi-gations of larger numbers of prelipemic women within-depth family studies are necessary to resolve thisquestion. In addition, long-term follow-up will be nec-essary to verify the hypothesis that the hyperlipide-mic stress of pregnancy can unmask a prelipemictrait. Nonetheless, it is noteworthy that subject CBhas remained hyperlipidemic for almost 2 yearssince delivery, her only normal value having beenobtained before the pregnancy we observed.

While the postpartum observations made at spe-cific times and the distinctions between subjects withphysiologic and supraphysiologic lipid increases inpregnancy are entirely new, our antepartum obser-vations confirm several previously described fea-tures of hyperlipidemia in pregnancy. These includean increase in total and VLDL apo B,24 little or nochange in VLDL composition,2 2425 an increase intriglyceride content of LDL and HDL,22425 and anincrease in apo A-l.26 Thus, conclusions drawn fromthis paper can be considered to be generally repre-sentative of the body of information on hyperlipide-mia in pregnancy.

The possibility that the hyperlipidemia of preg-nancy might confer added arteriosclerosis risk hasrecently been raised.2728 While it is true that an in-crease in LDL cholesterol could enhance arterioscle-rosis, usually this change is associated with an apoB-rich29 and triglyceride-poor LDL,30 which is the op-posite of that seen in pregnancy. In addition, HDLcholesterol and apo A-l concentrations were in-creased, which may facilitate cholesterol removalfrom tissues. In light of these results, it is perhaps notsurprising that the epidemiologic evidence31 thatarteriosclerosis is more common in multiparouswomen is inconclusive. It is more likely that the ob-served lipoprotein changes serve a physiologicalpurpose to support growth of maternal and fetal tis-sues and production of large quantities of steroidhormones.5'16 The extent to which the postpartumlipoprotein changes are atherogenic is deserving offurther study.

In summary, evidence is presented to support thehypothesis that the hyperlipidemic stress of preg-nancy can unmask a latent hypertriglyceridemia, byusing as a basis for recognition a triglyceride rise inpregnancy exceeding the 95th percentile. Womenwith such a supraphysiologic triglyceride rise ante-partum may return to normal levels postpartum moreslowly than normal, have LDL lipids that change in anatypical manner antepartum and postpartum, have

HDL cholesterol concentrations that are persistentlylow antepartum and postpartum, and have hyperlipi-demic family members. In contrast, hypercholestero-lemia is not greatly exaggerated in pregnancy. Long-term follow-up studies of women with geneticallywell-characterized disorders of lipoprotein metabo-lism are required to determine if an abnormal lipopro-tein response in pregnancy can identify prelipemicsubjects and distinguish among the major disordersof lipoprotein metabolism. Identification of prelipe-mia will provide an opportunity to study prospectivelythe natural progression, potential for atherosclero-sis, and possible treatment of hyperlipidemia fromearly adulthood.

AcknowledgmentsThe authors express their thanks to the staff of the Northwest

Liptd Research Clinic Core Laboratory who performed analyses;to Susan Irvine who carefully obtained the samples; to the gra-cious women of the Childbirth Education Association who volun-teered for participation in the study; and to Donna Gola and LoriLebon for excellent secretarial assistance.

References1. O'Sulllvan JB, Mahan CM. Criteria for the oral glucose toler-

ance test in pregnancy. Diabetes 1964;13:278-2852. Knopp RH, Warth MR, Carroll CJ. Lipid metabolism in preg-

nancy. I. Changes in lipoprotein triglyceride and cholesterol innormal pregnancy and the effects of diabetes mellitus. J Re-prod Med 1973;10:95-101

3. Dandrow RV, O'Sulllvan JB. Obstetric hazards of gestation-al diabetes. Am J Obstet Gynecol 1966;96:1144-1147

4. O'Sulllvan JB. Gestational diabetes: Unsuspected, a symp-tomatic diabetes in pregnancy. N Engl J Med 1961;264:1082-1085

5. Knopp RH, Bergelln RO, Wahl PW, Walden CE, ChapmanM, Irvine S. Population-based lipoprotein lipid reference val-ues for pregnant women compared to nonpregnant womenclassified by sex hormone use. Am J Obstet Gynecol1982;143:626-637

6. Wahl PW, Walden CE, Albere JJ, et al. Distribution of lipo-protein triglyceride and lipoprotein cholesterol in an adult pop-ulation by age, sex, and hormone use: The Pacific NorthwestBell Telephone Company Health Survey. Atherosclerosis1981;39:111-124

7. Knopp RH, Warnlck GR, Walden CE, el al. Relationship ofgender, sex hormone use, and age with lipoprotein cholester-ol/triglyceride ratios in an adult population: The Pacific North-west Bell Telephone Company Health Survey. Atherosclero-sis 1981:39:133-146

8. Hoover JJ, Walden CE, Bergelln RO, et al. Cholesterol andtriglyceride distributions in an adult employee population: ThePacific Northwest Bell Telephone Company Health Survey.Lipids 1980:15:895-903

9. Lipid Research Clinics Program. Manual of laboratory op-erations, Vol 1: Upid and lipoprotein analysis. (DHEW pub.no. (NIH) 75-628. Washington, DC: U.S. Government Print-ing Office, 1974

10. Hatch FT, Lees RS. Practical methods for plasma lipoproteinanalysis. In: Paoletti R, Kritchevsky R, eds. Advanced UpidResearch, vol 6. 1968:1-68

11. Albers JJ, Cabana VG, Hazzard WR. Immunoassay of hu-man plasma apolipoprotein B. Metabolism 1975:24:1339-1351

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

PREGNANCY HYPERLIPIDEMIA AND PRELIPEMIA Montes et al. 417

12. Albers JJ, Wahl PW, Cabana VG, Hazzard WR, Hoover JJ.Quantitation of apolipoprotein A-l of human plasma high den-sity lipoprotein. Metabolism 1976;25:633-644

13. Siegel S. Nonparametric statistics for the behavioral sci-ences. New York: McGraw-Hill, 1956:520

14. Humphrey JL, Childs MT, Montes A, Knopp RH. Lipid me-tabolism in pregnancy. VII. Kinetics of chylomicron triglycer-ide removal in fed pregnant rat. Am J Physiol 1980;239:E81-87

15. Childs MT, Tollefson J, Knopp RH, Applebaum-BowdenD. Lipid metabolism in pregnancy. VIII. Effects of dietary fatvs. carbohydrate on lipoprotein and hepatic lipids and tissuetriglyceride lipases. Metabolism 1981;30:27-35

16. Knopp RH, Montes A, Childs M, Li J. Metabolic adjust-ments in normal and diabetic pregnancy. Clin Obstet Gynecol

17. Childs MT, Knopp RH. Role of diet in the hyperiipidemia ofpregnancy, [abstr] Fed Proc 1978;37:348

18. Chan L, Jackson RL, O'Malley B, Means AR. Synthesis ofvery low density lipoproteins in the cockerel. J Clin Invest1976;58:368-379

19. Wilson DE, Lees RS. Metabolic relationships among theplasma lipoproteins: Reciprocal changes in the concentra-tions of very low, and low density lipoproteins in man. J ClinInvest 1972;51:1051-1057

20. Chait A, Albers J J, Brunzell JD. Very low density lipoproteinover production in genetic forms of hypertriglyceridemia. EurJ Clin Invest 1979;10:17-22

21. Goldstein JL, Schrott HG, Hazzard WR, Bierman EL, Mo-tulsky AG. Hyperiipidemia in coronary heart disease. II. Ge-netic analysis of lipid levels in 176 families and delineation ofa new inherited disorder, combined hyperiipidemia. J ClinInvest 1973:52:1544-1568

22. Walden CE, Wahl PW, Knopp RH, et al. Hyperiipidemia in

the Pacific Northwest Bell Telephone Company Health Sur-vey: Part I: Lipoprotein cholesterol and triglyceride concentra-tions. Arteriosclerosis 1983:3:117-124

23. Walden CE, Knopp RH, Wahl PW, et al. Hyperiipidemia inthe Pacific Northwest Bell Telephone Company Health Sur-vey: Part II: Lipoprotein lipid interrelationships. Arteriosclero-sis 1983:3:125-131

24. Hillman L, Schonfeld G, Miller JP, Wolff G. Apoliproteins inhuman pregnancy. Metabolism 1975:24:943-952

25. Warth MR, Arky RA, Knopp RH. Lipid metabolism in preg-nancy. III. Altered lipid composition in intermediate, very low,low and high density lipoprotein fractions. J Clin EndocrinolMetab 1975;41:649-655

26. Schonfeld G, Pf leger B. The structure of human high densitylipoprotein and the levels of apoprotein Al in plasma as deter-mined by radioimmunoassay. J Clin Invest 1974;54:236-246

27. Glueck CJ, Christopher C, Tsang RC, Mellies MJ. Choles-terol-free diet and the physiologic hyperiipidemia of preg-nancy in familial hypercholesterolemia. Metabolism 1980;29:949-955

28. McMurry MP, Connor WE, Goplerud CP. The effects ofdietary cholesterol upon the hypercholesterolemia of preg-nancy. Metabolism 1981;30:869-879

29. Sniderman A, Shapiro S, Marpole D, Skinner B, Teng B,Kwiterovich P. Association of coronary atherosclerosis withhyperapobetalipoproteinemia. Proc Natl Acad Sci USA 1980;77:604-608

30. Tall AR, Small DM, Atkinson D, Rudel LL. Studies on thestructure of low density lipoproteins from Macaca fascicularisfed an atherogenic diet. J Clin Invest 1978:62:1354-1363

31. Bengtsson C, Rybo G, Westerberg H. Number of pregnan-cies, use of contraceptives, and menopausal age in womenwith ischemic heart disease, compared to a population sam-ple of women. Acta Med Scand (Suppl) 1973:549:75-81

Index Terms, lipoproteins • apoproteins • pregnancy • hyperiipidemia • prelipemia

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from

A Montes, C E Walden, R H Knopp, M Cheung, M B Chapman and J J Alberspregnancy and postpartum. Possible markers for the diagnosis of "prelipemia".

Physiologic and supraphysiologic increases in lipoprotein lipids and apoproteins in late

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1984 American Heart Association, Inc. All rights reserved.

Avenue, Dallas, TX 75231is published by the American Heart Association, 7272 GreenvilleArteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.4.4.4071984;4:407-417Arterioscler Thromb Vasc Biol. 

http://atvb.ahajournals.org/content/4/4/407World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

  http://www.lww.com/reprints

Information about reprints can be found online at: Reprints: 

document.Permissions and Rights Question and AnswerFurther information about this process is available in theis being requested is located, click Request Permissions in the middle column of the Web page under Services.Clearance Center, not the Editorial Office. Once the online version of the published article for which permission

can be obtained via RightsLink, a service of the CopyrightArteriosclerosis, Thrombosis, and Vascular Biology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on February 7, 2018http://atvb.ahajournals.org/

Dow

nloaded from