sensitivitymult (version 1.0.2)

artcog: Arthritis and cognition in the elderly.

Description

The R package contains a simulated data set similar to actual data from 2009-2011 Irish Longitudinal Study of Aging (TILDA) used in Rosenbaum (2017). Additionally, in the documentation below, instructions are given for constructing the actual data set after downloading a file from ICPSR. The simulated data may be used to try the methods in this package. The actual data may be used to replicate the calculations in Rosenbaum (2017). Please be careful to distinguish the simulated data (with continuous outcomes) and the actual data (with integer outcomes), as scientific conclusions should not be based on the simulated data.

Instructions for creating the actual data are in the example section, but are not executed because you must obtain a file from ICPSR.

The simulated data were built from the actual data by calculating the trivariate within group means and the pooled within group covariance matrix. Then a data set of the same size was sampled from the trivariate Normal distribution, using the actual means and covariance matrix as the population parameters for the simulation. The simulation used the mvtnorm package. Although the data set consists of matched triples, in the simulated version, the matched sets are independent of the outcomes.

There are 219 matched triples containing one individual with arthritis (arthritis=1) and two without (arthritis=0). There are three measures of cognitive performance, words, wordsdelay and animals.

Usage

data("artcog")

Format

A data frame with 657 observations on the following 5 variables.

arthritis

1 if osteoarthritis, 0 if no arthritis

words

Individuals are read a list of words and are immediately asked to recall as many as they can.

wordsdelay

After a delay and another task, individuals are again asked to recall as many of the words as they can.

animals

Individuals are asked to name all of the animals they can think of.

mset

Indicator of the matched set: 1, 2, ..., 219.

Details

A theory that NSAIDs reduce the risk of Alzheimer's disease has often been examined by comparing elderly people with and without arthritis, reasoning that many people with arthritis have consumed NSAIDs in quantity for a long period; see McGeer et al. (1996). This comparison does not ask a person with Alzheimer's disease to recall past use of NSAIDs.

Triples were matched for age, sex, education and mother's education.

Everyone is 75 years old or older.

References

Irish Longitudinal Study of Aging. http://tilda.tcd.ie/

McGeer, P. L., Schulzer, M., and McGeer, E. G. (1996). Arthritis and anti- inflammatory agents as possible protective factors for alzheimer's disease. Neurology, 47, 425-432.

Rosenbaum, P. R. (2017) Combining planned and discovered comparisons in observational studies. Manuscript. (The artcog example is discussed in this manuscript.)

Examples

# NOT RUN {
# data(artcog) returns the simulated example.
data(artcog)
# Three correlated outcomes.
cor(artcog[,2:4])
# See documentation for principal() for use of this example.

# The code below constructs the actual data, as distinct
# from the simulated example.  The lengthy list of numbers
# assembles the 219 matched triples, or 657 = 3*219 rows,
# from the larger TILDA data set.

# }
# NOT RUN {
# Obtain from ICPSR the R data file
# ICPSR_34315-1IrishAging/34315-0001-Data.rda
# http://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/34315?q=34315

# The data should have 8504 rows and 1992 columns

d<-da34315.0001
attach(d)

wordsC<-PH118
wordsI<-PH119
wordsC[wordsC<0]<-0
wordsI[wordsI<0]<-0
words<-wordsC+wordsI

wordsdelayC<-PH712
wordsdelayC[is.na(wordsdelayC)]<-0
wordsdelayC[wordsdelayC<=-1]<-0
wordsdelayI<-PH713
wordsdelayI[is.na(wordsdelayI)]<-0
wordsdelayI[wordsdelayI<=-1]<-0
wordsdelay<-wordsdelayC+wordsdelayI

animals<-PH125

arthritis<-PH301_03
osteoA<-PH304_1
z<-rep(NA,dim(d)[1])
z[arthritis==0]<-0
z[(arthritis==1)&(osteoA==1)]<-1

detach(d)

artcog<-data.frame(z,words,wordsdelay,animals)

who <- c(91, 4408, 7754, 129, 4716, 8383, 135, 6066,
 8028, 280, 894, 5300, 288, 151, 667, 298, 4889, 5977,
 333, 1100, 3707, 480, 696, 8148, 568, 372,
 7578, 584, 1852, 8057, 589, 3799, 6567, 590, 7422,
 8419, 609, 2825, 8272, 669, 1197, 8471, 684, 141,
 1847, 687, 2416, 7591, 771, 5239, 6986, 782,
 4857, 7654, 850, 885, 2239, 892, 2717, 7788, 929, 248,
 4740, 975, 1965, 8242, 1036, 6459, 7973, 1059, 1541,
 5901, 1103, 6518, 8264, 1160, 4798, 7330,
 1168, 4678, 7319, 1180, 152, 2735, 1191, 3740,
 7260, 1199, 26, 5209, 1252, 2615, 3251, 1444, 4790,
 7298, 1549, 898, 7630, 1587, 4418, 7122, 1596, 5875,
 8489, 1604, 3594, 7246, 1614, 3189, 7052, 1646,
 5415, 6828, 1708, 1634, 7029, 1760, 1950, 7815, 1840,
 5860, 8334, 1843, 6054, 7331, 1849, 5617, 8046, 1854,
 2890, 7703, 1885, 5846, 7247, 1896, 4365, 7803,
 1898, 3952, 4187, 1977, 544, 940, 1987, 768, 960,
 2029, 5363, 6293, 2161, 10, 4432, 2270, 5620, 7132,
 2330, 445, 1301, 2372, 1014, 1138, 2379, 3906,
 6183, 2386, 6226, 7203, 2417, 2458, 6616, 2437, 6262,
 7178, 2442, 3840, 8024, 2443, 4955, 5834, 2455, 1969,
 5967, 2457, 6962, 7560, 2466, 986, 2895, 2498, 2461,
 5876, 2522, 1837, 4803, 2618, 7279, 7764, 2734, 4005,
 4477, 2747, 221, 3837, 2763, 4440, 7863, 2765,
 6173, 7377, 2799, 7711, 7822, 2820, 2676, 7288, 2853,
 3035, 7518, 2914, 3142, 6891, 2952, 3081, 4908, 2969,
 3077, 6837, 3013, 747, 7614, 3107, 1754, 6564,
 3178, 2242, 4377, 3192, 260, 4530, 3246, 3019, 6478,
 3313, 4710, 7271, 3389, 356, 1796, 3481, 99, 491,
 3571, 658, 1410, 3693, 4341, 7624, 3694, 522,
 7702, 3704, 6532, 7171, 3705, 4973, 7131, 3806, 2163,
 5400, 3848, 4811, 7097, 3850, 2154, 5773, 3851, 3547,
 7613, 3862, 3357, 3370, 3877, 6186, 7990, 3913,
 455, 2883, 3931, 3548, 3699, 3933, 3210, 6164, 3935,
 4712, 7813, 3940, 5598, 7826, 3964, 2129, 8005, 3997,
 49, 1537, 4000, 3915, 5392, 4044, 3014, 6130,
 4052, 5208, 7213, 4186, 1586, 4249, 4264, 7058, 7182,
 4324, 3950, 7507, 4343, 3701, 6359, 4358, 567, 1020,
 4387, 2919, 4011, 4389, 5851, 7125, 4409, 3310,
 8100, 4427, 767, 2108, 4439, 1263, 6024, 4447, 3814,
 8373, 4478, 3493, 6743, 4479, 939, 2621, 4537, 1264,
 7942, 4608, 1797, 2987, 4633, 976, 1814, 4641,
 274, 1116, 4697, 4718, 7008, 4750, 2842, 5787, 4791,
 4386, 6966, 4812, 2817, 5640, 4815, 845, 5430, 4856,
 2288, 2289, 4887, 2182, 4874, 4942, 460, 4300,
 4945, 565, 3644, 4946, 487, 3369, 4953, 4352, 6709,
 4956, 2731, 3387, 4958, 4436, 6460, 4964, 3388, 6692,
 5078, 278, 963, 5110, 842, 4842, 5166, 600,
 1530, 5199, 1775, 6210, 5204, 4993, 8477, 5210, 2646,
 5563, 5291, 2957, 7777, 5325, 4881, 7053, 5342, 4385,
 6444, 5377, 3957, 4319, 5384, 3144, 7757, 5385,
 2813, 3054, 5386, 3636, 6185, 5474, 2507, 5085, 5488,
 4278, 5675, 5584, 2606, 5359, 5599, 3180, 7037, 5605,
 2459, 5304, 5637, 2581, 3621, 5641, 2781, 4302,
 5805, 6424, 7227, 5870, 492, 5847, 5909, 1750, 5158,
 5923, 3199, 6492, 6039, 4347, 4762, 6048, 5332, 7320,
 6080, 1992, 2830, 6091, 5213, 7045, 6099, 5167,
 6511, 6135, 5177, 6944, 6172, 2983, 6455, 6176, 2319,
 3737, 6189, 5525, 7257, 6196, 4423, 6893, 6256, 2639,
 5740, 6322, 1427, 2435, 6370, 7321, 7385, 6371,
 1212, 2423, 6417, 205, 1674, 6462, 2393, 2882, 6463,
 2170, 4765, 6496, 1630, 5048, 6519, 3058, 7498, 6901,
 5237, 7508, 6984, 3819, 6548, 7042, 2961, 6445,
 7057, 5457, 7984, 7061, 5401, 6049, 7093, 502, 3847,
 7094, 4717, 6348, 7096, 825, 7844, 7099, 2188, 8337,
 7251, 7423, 7576, 7269, 2616, 6401, 7270, 2394,
 5039, 7273, 2337, 4941, 7300, 2241, 4934, 7316, 2604,
 6369, 7355, 2113, 3880, 7402, 1456, 2378, 7473, 6368,
 7243, 7592, 5583, 7892, 7615, 855, 6924, 7684,
 6412, 6822, 7852, 405, 7077, 7862, 3623, 3990, 7879,
 2447, 6334, 7913, 3927, 5299, 7930, 5289, 5844, 7983,
 297, 7772, 8006, 3869, 6930, 8009, 2729, 6480,
 8081, 4700, 6560, 8109, 62, 8061, 8130, 3351, 4381,
 8149, 2854, 6513, 8157, 5220, 7559, 8184, 1523, 4195,
 8185, 1459, 3820, 8188, 117, 1050, 8206, 2513,
 3954, 8335, 2352, 4435, 8346, 5109, 8207)

artcog<-artcog[who,]
mset<-as.numeric(gl(219,3))
artcog<-cbind(artcog,mset)

rm(z,words,wordsdelay,animals,mset)
	
# }