Final Model (with mediators)

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ gndr i.AL_cat bmxbmi ridageyr, or
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.AL_cat          _IAL_cat_0-2        (naturally coded; _IAL_cat_0 omitted)
(running logit on estimation sample)

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       67,991
Number of PSUs     =       214                Population size   =  292,944,878
                                              Subpop. no. obs   =       18,821
                                              Subpop. size      =  127,924,286
                                              Design df         =          109
                                              F(   9,    101)   =       171.42
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .9195897   .0431616    -1.79   0.077     .8379031     1.00924
      _IMJ_2 |   1.125541   .0807243     1.65   0.102     .9763994    1.297464
      _IMJ_3 |   1.256641    .187487     1.53   0.129     .9349512    1.689016
      _IMJ_4 |   1.136796   .1206675     1.21   0.230     .9211185    1.402973
        gndr |   2.678669   .1097468    24.05   0.000     2.469751    2.905259
  _IAL_cat_1 |   1.178936   .0645598     3.01   0.003     1.057679    1.314093
  _IAL_cat_2 |   1.404028   .0972246     4.90   0.000      1.22397    1.610573
      bmxbmi |   1.076922   .0039402    20.25   0.000     1.069141     1.08476
    ridageyr |   1.048122   .0022665    21.73   0.000     1.043639    1.052623
       _cons |   .0092223   .0013177   -32.80   0.000     .0069479    .0122412
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Model 2 (without mediators)

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ gndr ridageyr, or
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
(running logit on estimation sample)

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       70,190
Number of PSUs     =       214                Population size   =  306,545,053
                                              Subpop. no. obs   =       21,020
                                              Subpop. size      =  141,524,461
                                              Design df         =          109
                                              F(   6,    104)   =       210.20
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .9599054   .0403027    -0.97   0.332     .8832601    1.043202
      _IMJ_2 |   1.111107   .0745439     1.57   0.119     .9727652    1.269124
      _IMJ_3 |   1.213377   .1784316     1.32   0.191     .9066055    1.623953
      _IMJ_4 |     1.1299   .1107676     1.25   0.216      .930373    1.372216
        gndr |   2.374512    .088687    23.15   0.000     2.205085    2.556956
    ridageyr |   1.047735   .0020607    23.71   0.000     1.043659    1.051827
       _cons |   .0939611   .0077396   -28.71   0.000     .0798082    .1106239
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Gender EMM analysis

Full model (Model 3)

Stratified Analysis

Men

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1 & gndr==1): logit BP_cat i.MJ gndr i.AL_cat bmxbmi ridageyr, or
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.AL_cat          _IAL_cat_0-2        (naturally coded; _IAL_cat_0 omitted)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       69,030
Number of PSUs     =       214                Population size   =  299,205,783
                                              Subpop. no. obs   =        9,181
                                              Subpop. size      = 63,854,681.3
                                              Design df         =          109
                                              F(   8,    102)   =        46.02
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |    1.03377     .06878     0.50   0.619     .9060561    1.179486
      _IMJ_2 |   1.263003   .1145963     2.57   0.011     1.055128    1.511833
      _IMJ_3 |   1.445419   .2437275     2.18   0.031      1.03479    2.018995
      _IMJ_4 |   1.247216   .1815545     1.52   0.132     .9346373    1.664333
        gndr |          1  (omitted)
  _IAL_cat_1 |    1.20197   .0786707     2.81   0.006     1.055737    1.368458
  _IAL_cat_2 |   1.397182   .1183305     3.95   0.000     1.181281    1.652542
      bmxbmi |   1.086816   .0059293    15.26   0.000     1.075128    1.098632
    ridageyr |   1.028045   .0027995    10.16   0.000     1.022511    1.033608
       _cons |   .0369985   .0072712   -16.78   0.000     .0250624    .0546192
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Women

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1 & gndr==0): logit BP_cat i.MJ gndr i.AL_cat bmxbmi ridageyr, or
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.AL_cat          _IAL_cat_0-2        (naturally coded; _IAL_cat_0 omitted)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       69,151
Number of PSUs     =       214                Population size   =  300,284,148
                                              Subpop. no. obs   =        9,640
                                              Subpop. size      = 64,069,604.7
                                              Design df         =          109
                                              F(   8,    102)   =        86.72
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8397025    .057369    -2.56   0.012     .7333614    .9614637
      _IMJ_2 |    1.02052   .1209216     0.17   0.864     .8069195    1.290663
      _IMJ_3 |   .8498809   .2407074    -0.57   0.567     .4848084    1.489862
      _IMJ_4 |   .9591782   .1709363    -0.23   0.816     .6737561    1.365513
        gndr |          1  (omitted)
  _IAL_cat_1 |   1.160256   .0894539     1.93   0.056     .9958428    1.351814
  _IAL_cat_2 |   1.401877   .1502714     3.15   0.002     1.133555    1.733712
      bmxbmi |   1.074915   .0044394    17.49   0.000     1.066152     1.08375
    ridageyr |   1.071599   .0034429    21.52   0.000     1.064797    1.078444
       _cons |   .0041514   .0008711   -26.14   0.000     .0027388    .0062924
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Interaction Analysis

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr i.AL_cat bmxbmi ridageyr, or
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
i.AL_cat          _IAL_cat_0-2        (naturally coded; _IAL_cat_0 omitted)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       67,991
Number of PSUs     =       214                Population size   =  292,944,878
                                              Subpop. no. obs   =       18,821
                                              Subpop. size      =  127,924,286
                                              Design df         =          109
                                              F(  13,     97)   =       113.80
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8408813    .055042    -2.65   0.009       .73857    .9573653
      _IMJ_2 |    .900777   .0927807    -1.01   0.313     .7344439     1.10478
      _IMJ_3 |   .7677384   .2118636    -0.96   0.340      .444304    1.326619
      _IMJ_4 |   .8710944   .1504346    -0.80   0.426     .6186083    1.226633
        gndr |   2.314181   .1481713    13.10   0.000      2.03838    2.627299
 _IMJXgndr_1 |   1.207781   .1173943     1.94   0.055     .9961488    1.464376
 _IMJXgndr_2 |   1.483855   .1925096     3.04   0.003     1.147414    1.918946
 _IMJXgndr_3 |   2.089399    .626913     2.46   0.016     1.152804    3.786928
 _IMJXgndr_4 |   1.547986   .3697322     1.83   0.070     .9642232    2.485173
        gndr |          1  (omitted)
  _IAL_cat_1 |   1.187764   .0650172     3.14   0.002     1.065647    1.323876
  _IAL_cat_2 |   1.409399   .0979624     4.94   0.000      1.22802    1.617567
      bmxbmi |   1.077271   .0039565    20.27   0.000     1.069457    1.085141
    ridageyr |   1.048078   .0022851    21.54   0.000     1.043559    1.052617
       _cons |    .009751   .0014435   -31.28   0.000     .0072716     .013076
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

previous use lincom

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr i.AL_cat bmxbmi ridageyr, or

# long form
lincom (1*_IMJ_1 + 0*_IMJ_2 + 0*_IMJ_3 + 0*_IMJ_4 + 1*gndr + 1*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4) - (1*_IMJ_1 + 0*_IMJ_2 + 0*_IMJ_3 + 0*_IMJ_4 + 0*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4)

# reduced form
lincom gndr + _IMJXgndr_1
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
i.AL_cat          _IAL_cat_0-2        (naturally coded; _IAL_cat_0 omitted)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       67,991
Number of PSUs     =       214                Population size   =  292,944,878
                                              Subpop. no. obs   =       18,821
                                              Subpop. size      =  127,924,286
                                              Design df         =          109
                                              F(  13,     97)   =       113.80
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8408813    .055042    -2.65   0.009       .73857    .9573653
      _IMJ_2 |    .900777   .0927807    -1.01   0.313     .7344439     1.10478
      _IMJ_3 |   .7677384   .2118636    -0.96   0.340      .444304    1.326619
      _IMJ_4 |   .8710944   .1504346    -0.80   0.426     .6186083    1.226633
        gndr |   2.314181   .1481713    13.10   0.000      2.03838    2.627299
 _IMJXgndr_1 |   1.207781   .1173943     1.94   0.055     .9961488    1.464376
 _IMJXgndr_2 |   1.483855   .1925096     3.04   0.003     1.147414    1.918946
 _IMJXgndr_3 |   2.089399    .626913     2.46   0.016     1.152804    3.786928
 _IMJXgndr_4 |   1.547986   .3697322     1.83   0.070     .9642232    2.485173
        gndr |          1  (omitted)
  _IAL_cat_1 |   1.187764   .0650172     3.14   0.002     1.065647    1.323876
  _IAL_cat_2 |   1.409399   .0979624     4.94   0.000      1.22802    1.617567
      bmxbmi |   1.077271   .0039565    20.27   0.000     1.069457    1.085141
    ridageyr |   1.048078   .0022851    21.54   0.000     1.043559    1.052617
       _cons |    .009751   .0014435   -31.28   0.000     .0072716     .013076
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_1 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.027841   .0659409    15.59   0.000     .8971484    1.158534
------------------------------------------------------------------------------

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_1 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.027841   .0659409    15.59   0.000     .8971484    1.158534
------------------------------------------------------------------------------

light use lincom

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr i.AL_cat bmxbmi ridageyr, or

# long form
lincom (0*_IMJ_1 + 1*_IMJ_2 + 0*_IMJ_3 + 0*_IMJ_4 + 1*gndr + 0*_IMJXgndr_1 + 1*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4) - (0*_IMJ_1 + 1*_IMJ_2 + 0*_IMJ_3 + 0*_IMJ_4 + 0*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4)

# reduced form
lincom gndr + _IMJXgndr_2
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
i.AL_cat          _IAL_cat_0-2        (naturally coded; _IAL_cat_0 omitted)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       67,991
Number of PSUs     =       214                Population size   =  292,944,878
                                              Subpop. no. obs   =       18,821
                                              Subpop. size      =  127,924,286
                                              Design df         =          109
                                              F(  13,     97)   =       113.80
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8408813    .055042    -2.65   0.009       .73857    .9573653
      _IMJ_2 |    .900777   .0927807    -1.01   0.313     .7344439     1.10478
      _IMJ_3 |   .7677384   .2118636    -0.96   0.340      .444304    1.326619
      _IMJ_4 |   .8710944   .1504346    -0.80   0.426     .6186083    1.226633
        gndr |   2.314181   .1481713    13.10   0.000      2.03838    2.627299
 _IMJXgndr_1 |   1.207781   .1173943     1.94   0.055     .9961488    1.464376
 _IMJXgndr_2 |   1.483855   .1925096     3.04   0.003     1.147414    1.918946
 _IMJXgndr_3 |   2.089399    .626913     2.46   0.016     1.152804    3.786928
 _IMJXgndr_4 |   1.547986   .3697322     1.83   0.070     .9642232    2.485173
        gndr |          1  (omitted)
  _IAL_cat_1 |   1.187764   .0650172     3.14   0.002     1.065647    1.323876
  _IAL_cat_2 |   1.409399   .0979624     4.94   0.000      1.22802    1.617567
      bmxbmi |   1.077271   .0039565    20.27   0.000     1.069457    1.085141
    ridageyr |   1.048078   .0022851    21.54   0.000     1.043559    1.052617
       _cons |    .009751   .0014435   -31.28   0.000     .0072716     .013076
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_2 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.233699   .1234798     9.99   0.000     .9889662    1.478432
------------------------------------------------------------------------------

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_2 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.233699   .1234798     9.99   0.000     .9889662    1.478432
------------------------------------------------------------------------------

moderate use lincom

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr i.AL_cat bmxbmi ridageyr, or

# long form
lincom (0*_IMJ_1 + 0*_IMJ_2 + 1*_IMJ_3 + 0*_IMJ_4 + 1*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 1*_IMJXgndr_3 + 0*_IMJXgndr_4) - (0*_IMJ_1 + 0*_IMJ_2 + 1*_IMJ_3 + 0*_IMJ_4 + 0*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4)

# reduced form
lincom gndr + _IMJXgndr_3
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
i.AL_cat          _IAL_cat_0-2        (naturally coded; _IAL_cat_0 omitted)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       67,991
Number of PSUs     =       214                Population size   =  292,944,878
                                              Subpop. no. obs   =       18,821
                                              Subpop. size      =  127,924,286
                                              Design df         =          109
                                              F(  13,     97)   =       113.80
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8408813    .055042    -2.65   0.009       .73857    .9573653
      _IMJ_2 |    .900777   .0927807    -1.01   0.313     .7344439     1.10478
      _IMJ_3 |   .7677384   .2118636    -0.96   0.340      .444304    1.326619
      _IMJ_4 |   .8710944   .1504346    -0.80   0.426     .6186083    1.226633
        gndr |   2.314181   .1481713    13.10   0.000      2.03838    2.627299
 _IMJXgndr_1 |   1.207781   .1173943     1.94   0.055     .9961488    1.464376
 _IMJXgndr_2 |   1.483855   .1925096     3.04   0.003     1.147414    1.918946
 _IMJXgndr_3 |   2.089399    .626913     2.46   0.016     1.152804    3.786928
 _IMJXgndr_4 |   1.547986   .3697322     1.83   0.070     .9642232    2.485173
        gndr |          1  (omitted)
  _IAL_cat_1 |   1.187764   .0650172     3.14   0.002     1.065647    1.323876
  _IAL_cat_2 |   1.409399   .0979624     4.94   0.000      1.22802    1.617567
      bmxbmi |   1.077271   .0039565    20.27   0.000     1.069457    1.085141
    ridageyr |   1.048078   .0022851    21.54   0.000     1.043559    1.052617
       _cons |    .009751   .0014435   -31.28   0.000     .0072716     .013076
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_3 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.575932   .3023817     5.21   0.000     .9766215    2.175243
------------------------------------------------------------------------------

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_3 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.575932   .3023817     5.21   0.000     .9766215    2.175243
------------------------------------------------------------------------------

heavy use lincom

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr i.AL_cat bmxbmi ridageyr, or

# long form
lincom (0*_IMJ_1 + 0*_IMJ_2 + 0*_IMJ_3 + 1*_IMJ_4 + 1*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 1*_IMJXgndr_4) - (0*_IMJ_1 + 0*_IMJ_2 + 0*_IMJ_3 + 1*_IMJ_4 + 0*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4)

# reduced form
lincom gndr + _IMJXgndr_4
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
i.AL_cat          _IAL_cat_0-2        (naturally coded; _IAL_cat_0 omitted)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       67,991
Number of PSUs     =       214                Population size   =  292,944,878
                                              Subpop. no. obs   =       18,821
                                              Subpop. size      =  127,924,286
                                              Design df         =          109
                                              F(  13,     97)   =       113.80
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8408813    .055042    -2.65   0.009       .73857    .9573653
      _IMJ_2 |    .900777   .0927807    -1.01   0.313     .7344439     1.10478
      _IMJ_3 |   .7677384   .2118636    -0.96   0.340      .444304    1.326619
      _IMJ_4 |   .8710944   .1504346    -0.80   0.426     .6186083    1.226633
        gndr |   2.314181   .1481713    13.10   0.000      2.03838    2.627299
 _IMJXgndr_1 |   1.207781   .1173943     1.94   0.055     .9961488    1.464376
 _IMJXgndr_2 |   1.483855   .1925096     3.04   0.003     1.147414    1.918946
 _IMJXgndr_3 |   2.089399    .626913     2.46   0.016     1.152804    3.786928
 _IMJXgndr_4 |   1.547986   .3697322     1.83   0.070     .9642232    2.485173
        gndr |          1  (omitted)
  _IAL_cat_1 |   1.187764   .0650172     3.14   0.002     1.065647    1.323876
  _IAL_cat_2 |   1.409399   .0979624     4.94   0.000      1.22802    1.617567
      bmxbmi |   1.077271   .0039565    20.27   0.000     1.069457    1.085141
    ridageyr |   1.048078   .0022851    21.54   0.000     1.043559    1.052617
       _cons |    .009751   .0014435   -31.28   0.000     .0072716     .013076
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_4 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.276011    .222936     5.72   0.000      .834159    1.717863
------------------------------------------------------------------------------

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_4 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.276011    .222936     5.72   0.000      .834159    1.717863
------------------------------------------------------------------------------

Model 2 EMM analysis (Without mediators)

Stratified Analysis

Men

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1 & gndr==1): logit BP_cat i.MJ gndr ridageyr, or
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       70,190
Number of PSUs     =       214                Population size   =  306,545,053
                                              Subpop. no. obs   =       10,341
                                              Subpop. size      = 71,193,951.1
                                              Design df         =          109
                                              F(   5,    105)   =        27.14
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   1.072846    .065836     1.15   0.254     .9499846    1.211598
      _IMJ_2 |   1.175496   .1036334     1.83   0.069     .9870414    1.399931
      _IMJ_3 |   1.216172   .2048764     1.16   0.248     .8709469    1.698237
      _IMJ_4 |   1.136323   .1415161     1.03   0.307     .8877778    1.454452
        gndr |          1  (omitted)
    ridageyr |   1.027805   .0025803    10.92   0.000     1.022703    1.032931
       _cons |   .4421024   .0485597    -7.43   0.000      .355614    .5496256
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Women

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1 & gndr==0): logit BP_cat i.MJ gndr ridageyr, or
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       70,190
Number of PSUs     =       214                Population size   =  306,545,053
                                              Subpop. no. obs   =       10,679
                                              Subpop. size      = 70,330,509.5
                                              Design df         =          109
                                              F(   5,    105)   =       127.66
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8788361   .0518984    -2.19   0.031     .7817666    .9879585
      _IMJ_2 |   1.062862   .1053095     0.62   0.540     .8733574    1.293485
      _IMJ_3 |    1.11219   .2761763     0.43   0.669     .6798883    1.819366
      _IMJ_4 |   1.113811   .1969892     0.61   0.543     .7844718    1.581414
        gndr |          1  (omitted)
    ridageyr |   1.071201   .0029653    24.85   0.000      1.06534    1.077095
       _cons |     .03911   .0048993   -25.88   0.000     .0305113    .0501318
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Interaction Analysis

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr ridageyr, or
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       70,190
Number of PSUs     =       214                Population size   =  306,545,053
                                              Subpop. no. obs   =       21,020
                                              Subpop. size      =  141,524,461
                                              Design df         =          109
                                              F(  10,    100)   =       124.87
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8837478   .0491521    -2.22   0.028     .7915073    .9867379
      _IMJ_2 |   .9450544    .085931    -0.62   0.536     .7892065    1.131678
      _IMJ_3 |    1.01972   .2408252     0.08   0.934     .6385536    1.628412
      _IMJ_4 |   1.013036   .1680923     0.08   0.938     .7291217    1.407504
        gndr |   2.114103   .1225442    12.92   0.000     1.884657    2.371483
 _IMJXgndr_1 |   1.190736   .1030301     2.02   0.046     1.003083    1.413493
 _IMJXgndr_2 |   1.343683   .1634726     2.43   0.017     1.055788    1.710081
 _IMJXgndr_3 |   1.340381   .3702171     1.06   0.291     .7753261    2.317245
 _IMJXgndr_4 |   1.221547   .2665687     0.92   0.361     .7926362    1.882551
        gndr |          1  (omitted)
    ridageyr |   1.047645   .0020781    23.46   0.000     1.043534    1.051771
       _cons |   .0993138   .0088017   -26.06   0.000     .0833153    .1183843
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

previous use lincom

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr ridageyr, or

# long form
lincom (1*_IMJ_1 + 0*_IMJ_2 + 0*_IMJ_3 + 0*_IMJ_4 + 1*gndr + 1*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4) - (1*_IMJ_1 + 0*_IMJ_2 + 0*_IMJ_3 + 0*_IMJ_4 + 0*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4)

# reduced form
lincom gndr + _IMJXgndr_1
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       70,190
Number of PSUs     =       214                Population size   =  306,545,053
                                              Subpop. no. obs   =       21,020
                                              Subpop. size      =  141,524,461
                                              Design df         =          109
                                              F(  10,    100)   =       124.87
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8837478   .0491521    -2.22   0.028     .7915073    .9867379
      _IMJ_2 |   .9450544    .085931    -0.62   0.536     .7892065    1.131678
      _IMJ_3 |    1.01972   .2408252     0.08   0.934     .6385536    1.628412
      _IMJ_4 |   1.013036   .1680923     0.08   0.938     .7291217    1.407504
        gndr |   2.114103   .1225442    12.92   0.000     1.884657    2.371483
 _IMJXgndr_1 |   1.190736   .1030301     2.02   0.046     1.003083    1.413493
 _IMJXgndr_2 |   1.343683   .1634726     2.43   0.017     1.055788    1.710081
 _IMJXgndr_3 |   1.340381   .3702171     1.06   0.291     .7753261    2.317245
 _IMJXgndr_4 |   1.221547   .2665687     0.92   0.361     .7926362    1.882551
        gndr |          1  (omitted)
    ridageyr |   1.047645   .0020781    23.46   0.000     1.043534    1.051771
       _cons |   .0993138   .0088017   -26.06   0.000     .0833153    .1183843
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_1 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |    .923202   .0592154    15.59   0.000      .805839    1.040565
------------------------------------------------------------------------------

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_1 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |    .923202   .0592154    15.59   0.000      .805839    1.040565
------------------------------------------------------------------------------

light use lincom

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr ridageyr, or

# long form
lincom (0*_IMJ_1 + 1*_IMJ_2 + 0*_IMJ_3 + 0*_IMJ_4 + 1*gndr + 0*_IMJXgndr_1 + 1*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4) - (0*_IMJ_1 + 1*_IMJ_2 + 0*_IMJ_3 + 0*_IMJ_4 + 0*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4)

# reduced form
lincom gndr + _IMJXgndr_2
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       70,190
Number of PSUs     =       214                Population size   =  306,545,053
                                              Subpop. no. obs   =       21,020
                                              Subpop. size      =  141,524,461
                                              Design df         =          109
                                              F(  10,    100)   =       124.87
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8837478   .0491521    -2.22   0.028     .7915073    .9867379
      _IMJ_2 |   .9450544    .085931    -0.62   0.536     .7892065    1.131678
      _IMJ_3 |    1.01972   .2408252     0.08   0.934     .6385536    1.628412
      _IMJ_4 |   1.013036   .1680923     0.08   0.938     .7291217    1.407504
        gndr |   2.114103   .1225442    12.92   0.000     1.884657    2.371483
 _IMJXgndr_1 |   1.190736   .1030301     2.02   0.046     1.003083    1.413493
 _IMJXgndr_2 |   1.343683   .1634726     2.43   0.017     1.055788    1.710081
 _IMJXgndr_3 |   1.340381   .3702171     1.06   0.291     .7753261    2.317245
 _IMJXgndr_4 |   1.221547   .2665687     0.92   0.361     .7926362    1.882551
        gndr |          1  (omitted)
    ridageyr |   1.047645   .0020781    23.46   0.000     1.043534    1.051771
       _cons |   .0993138   .0088017   -26.06   0.000     .0833153    .1183843
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_2 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.044045   .1114583     9.37   0.000     .8231381    1.264952
------------------------------------------------------------------------------

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_2 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.044045   .1114583     9.37   0.000     .8231381    1.264952
------------------------------------------------------------------------------

moderate use lincom

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr ridageyr, or

# long form
lincom (0*_IMJ_1 + 0*_IMJ_2 + 1*_IMJ_3 + 0*_IMJ_4 + 1*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 1*_IMJXgndr_3 + 0*_IMJXgndr_4) - (0*_IMJ_1 + 0*_IMJ_2 + 1*_IMJ_3 + 0*_IMJ_4 + 0*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4)

# reduced form
lincom gndr + _IMJXgndr_3
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       70,190
Number of PSUs     =       214                Population size   =  306,545,053
                                              Subpop. no. obs   =       21,020
                                              Subpop. size      =  141,524,461
                                              Design df         =          109
                                              F(  10,    100)   =       124.87
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8837478   .0491521    -2.22   0.028     .7915073    .9867379
      _IMJ_2 |   .9450544    .085931    -0.62   0.536     .7892065    1.131678
      _IMJ_3 |    1.01972   .2408252     0.08   0.934     .6385536    1.628412
      _IMJ_4 |   1.013036   .1680923     0.08   0.938     .7291217    1.407504
        gndr |   2.114103   .1225442    12.92   0.000     1.884657    2.371483
 _IMJXgndr_1 |   1.190736   .1030301     2.02   0.046     1.003083    1.413493
 _IMJXgndr_2 |   1.343683   .1634726     2.43   0.017     1.055788    1.710081
 _IMJXgndr_3 |   1.340381   .3702171     1.06   0.291     .7753261    2.317245
 _IMJXgndr_4 |   1.221547   .2665687     0.92   0.361     .7926362    1.882551
        gndr |          1  (omitted)
    ridageyr |   1.047645   .0020781    23.46   0.000     1.043534    1.051771
       _cons |   .0993138   .0088017   -26.06   0.000     .0833153    .1183843
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_3 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.041584   .2720195     3.83   0.000     .5024505    1.580718
------------------------------------------------------------------------------

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_3 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |   1.041584   .2720195     3.83   0.000     .5024505    1.580718
------------------------------------------------------------------------------

heavy use lincom

use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

xi: svy,subpop(if include==1): logit BP_cat i.MJ*gndr gndr ridageyr, or

# long form
lincom (0*_IMJ_1 + 0*_IMJ_2 + 0*_IMJ_3 + 1*_IMJ_4 + 1*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 1*_IMJXgndr_4) - (0*_IMJ_1 + 0*_IMJ_2 + 0*_IMJ_3 + 1*_IMJ_4 + 0*gndr + 0*_IMJXgndr_1 + 0*_IMJXgndr_2 + 0*_IMJXgndr_3 + 0*_IMJXgndr_4)

# reduced form
lincom gndr + _IMJXgndr_4
      pweight: wtmec12yr
          VCE: linearized
  Single unit: missing
     Strata 1: sdmvstra
         SU 1: sdmvpsu
        FPC 1: <zero>

i.MJ              _IMJ_0-4            (naturally coded; _IMJ_0 omitted)
i.MJ*gndr         _IMJXgndr_#         (coded as above)
(running logit on estimation sample)

note: gndr omitted because of collinearity

Survey: Logistic regression

Number of strata   =       105                Number of obs     =       70,190
Number of PSUs     =       214                Population size   =  306,545,053
                                              Subpop. no. obs   =       21,020
                                              Subpop. size      =  141,524,461
                                              Design df         =          109
                                              F(  10,    100)   =       124.87
                                              Prob > F          =       0.0000

------------------------------------------------------------------------------
             |             Linearized
      BP_cat | Odds Ratio   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      _IMJ_1 |   .8837478   .0491521    -2.22   0.028     .7915073    .9867379
      _IMJ_2 |   .9450544    .085931    -0.62   0.536     .7892065    1.131678
      _IMJ_3 |    1.01972   .2408252     0.08   0.934     .6385536    1.628412
      _IMJ_4 |   1.013036   .1680923     0.08   0.938     .7291217    1.407504
        gndr |   2.114103   .1225442    12.92   0.000     1.884657    2.371483
 _IMJXgndr_1 |   1.190736   .1030301     2.02   0.046     1.003083    1.413493
 _IMJXgndr_2 |   1.343683   .1634726     2.43   0.017     1.055788    1.710081
 _IMJXgndr_3 |   1.340381   .3702171     1.06   0.291     .7753261    2.317245
 _IMJXgndr_4 |   1.221547   .2665687     0.92   0.361     .7926362    1.882551
        gndr |          1  (omitted)
    ridageyr |   1.047645   .0020781    23.46   0.000     1.043534    1.051771
       _cons |   .0993138   .0088017   -26.06   0.000     .0833153    .1183843
------------------------------------------------------------------------------
Note: _cons estimates baseline odds.

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_4 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |    .948749   .2022165     4.69   0.000     .5479625    1.349535
------------------------------------------------------------------------------

Unknown #command

 ( 1)  [BP_cat]gndr + [BP_cat]_IMJXgndr_4 = 0

------------------------------------------------------------------------------
      BP_cat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         (1) |    .948749   .2022165     4.69   0.000     .5479625    1.349535
------------------------------------------------------------------------------

Age

To be calculated later